Tia-920a Ballot Draft - Telecommunications Industry Association

Transcript

TR41.3.3-08-05-004 Telecommunications Telephone Terminal Equipment Transmission Requirements for Wideband Digital Telephones SP-3-4705-RV1 (to become ANSI/TIA-920-A) Draft 10 (Apr 25, 2008) Editor: Tom Harley [email protected] Changes Note: DISABLE CHANGE TRACKING BEFORE SAVING. Draft 10 Changes made during New Orleans meeting Draft 9 Changes made during Albuquerque NM meeting Draft 7 & 8 Changes made during Ottawa, ON meeting Draft 5 & 6 (Tom & Allen Wo & Roger ): 1. Changes made during St. Petersburg, FL meeting 2. Changes agreed to at FL but deferred, including duplicating handset updates into headset section. 3. Acknowledgements page updated. Draft 4 (Tom): 1. Changed handsfree to speakerphone. 2. Scattered G711 references deleted. 3. Changed speakerphone RLR=16 dB target to 2 dB by adding 14 dB correction. 4. Changed handset frequency responses for send and receive, as agreed. 5. Updated references. 6. TCLw handset 52 dB. 7. Specified 10N for high leak position of handset, added “artificial ear/mouth vs ear/mouth simulator” and “preferred ear simulator” sections to mirror 810B updates. 8. Reordered Section 4 and 5 to be more like 810B. 9. Rewrote handset Rx frequency response requirement. Draft 3 (Roger): 1. In 2.1 Scope adds ITU-T reference, deletes “comparison of different products”. Note added that when comparing different products, following identical procedures is important. 2. In 3. Normative References adds IETF RFC 1890 RTP reference for L16-256. 3. In 4.1 Codec L16-256 definition refined. 4. In 6.1 Vocoder Mandatory Requirements, G711 reference deleted. Draft 2 (Roger): 1. Tweaked first sentence Scope. 2. In 6.1 deleted “mandatory G.711 codec for narrowband operation” from 1st sentence and deleted 2nd paragraph about also meeting TIA-810-A requirements. Draft 1 (Roger): 3. Imported the 810-B changes into the handset section, clause 7 and some of Annex A. Deleted: 9 Deleted: Feb Deleted: 1 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) TABLE OF CONTENTS 1. INTRODUCTION ..................................................................................................................... 1 2. OVERVIEW .............................................................................................................................. 2 2.1. 2.2. 2.3. 2.4. 2.5. 2.6. SCOPE .................................................................................................................................. 2 LIMITS OF APPLICABILITY ................................................................................................ 2 CATEGORIES OF CRITERIA ............................................................................................... 2 FCC PART 68 ....................................................................................................................... 3 ENVIRONMENTAL .............................................................................................................. 3 SAFETY ................................................................................................................................ 3 3. NORMATIVE REFERENCES ................................................................................................ 4 4. DEFINITIONS, ABBREVIATIONS AND ACRONYMS ..................................................... 6 4.1. CODEC ................................................................................................................................. 6 4.2. EAR REFERENCE POINT (ERP).......................................................................................... 6 4.3. ARTIFICIAL EAR/MOUTH VS. EAR/MOUTH SIMULATOR .............................................. 6 4.4. HATS POSITION .................................................................................................................. 6 4.5. NOMINAL VOLUME CONTROL SETTING ......................................................................... 6 4.6. REFERENCE VOLUME CONTROL SETTING...................................................................... 6 4.7. PREFERRED EAR SIMULATOR .......................................................................................... 6 4.8. STANDARD TEST POSITION .............................................................................................. 6 4.9. RECOMMENDED TEST POSITIONS (RTP)......................................................................... 7 4.10. MOUTH REFERENCE POINT (MRP) .................................................................................. 7 4.11. SESSION DESCRIPTION PROTOCOL (SDP)....................................................................... 7 4.12. REFERENCE CODEC ........................................................................................................... 7 4.13. DIRECT DIGITAL PROCESSING ......................................................................................... 8 4.14. SOUND PRESSURE LEVELS ............................................................................................... 9 4.15. ELECTRIC POWER AND NOISE LEVELS ........................................................................... 9 4.16. 50TP ..................................................................................................................................... 9 4.17. ABBREVIATIONS AND ACRONYMS .................................................................................. 9 4.18. TEST SIGNALS .................................................................................................................. 10 1.1.1. Choice of Test Signal.................................................................................................. 10 1.1.2. Frequency Tolerance of Test Signals and Analysis .................................................... 10 4.19. TESTING MODE ................................................................................................................. 11 4.20. PRECAUTIONS................................................................................................................... 11 5. GENERAL TECHNICAL REQUIREMENTS..................................................................... 12 5.1. 6. VOICE CODING MANDATORY REQUIREMENTS ........................................................... 12 1.1.3. Transmission Format of L16-256 Codec .................................................................... 12 1.1.4. Overload Point ............................................................................................................ 13 1.1.5. Quiet Code and Full Scale Code ................................................................................. 13 1.1.6. 0 dBm0 (Digital Milliwatt) ......................................................................................... 13 HANDSET TECHNICAL REQUIREMENTS ..................................................................... 14 i SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 6.1. HANDSET FREQUENCY RESPONSE .................................................................................14 1.1.7. Handset Send Frequency Response .............................................................................14 1.1.8. Handset Receive Frequency Response ........................................................................16 6.2. HANDSET WIDEBAND LOUDNESS RATINGS AND RECEIVE VOLUME CONTROL .....18 1.1.9. Handset Wideband Send Loudness Rating (SLR).......................................................19 1.1.10. Handset Wideband Receive Loudness Rating (RLR).................................................19 1.1.11. Handset Receive Volume Control Range ...................................................................19 1.1.12. Magnetic Field for Hearing Aid Coupling..................................................................20 1.1.13. Handset Talker Sidetone (STMR) ..............................................................................20 1.1.14. Handset Sidetone Delay..............................................................................................21 6.3. HANDSET NOISE ...............................................................................................................21 1.1.15. Handset Send Noise....................................................................................................21 1.1.16. Handset Send Single Frequency Interference.............................................................21 1.1.17. Handset Receive Noise ...............................................................................................22 1.1.18. Handset Receive Single Frequency Interference ........................................................22 6.4. HANDSET RECEIVE COMFORT NOISE (ADVISORY) .....................................................22 1.1.19. General........................................................................................................................22 1.1.20. Measurement Method .................................................................................................22 1.1.21. Requirement................................................................................................................23 6.5. HANDSET DISTORTION AND NOISE ...............................................................................23 1.1.22. Handset Send Distortion and Noise............................................................................23 1.1.23. Handset Receive Distortion and Noise .......................................................................24 6.6. WEIGHTED TERMINAL COUPLING LOSS (TCLW).........................................................25 1.1.24. Measurement Method .................................................................................................25 1.1.25. Requirements ..............................................................................................................26 6.7. STABILITY LOSS ...............................................................................................................27 1.1.26. Measurement Method .................................................................................................28 1.1.27. Requirement................................................................................................................28 6.8. LONG DURATION MAXIMUM ACOUSTIC PRESSURE (STEADY STATE INPUT) .........28 1.1.28. General........................................................................................................................28 1.1.29. Measurement Method .................................................................................................28 1.1.30. Requirements ..............................................................................................................28 6.9. SHORT DURATION MAXIMUM ACOUSTIC PRESSURE (PEAK) ....................................28 1.1.31. General........................................................................................................................28 1.1.32. Measurement Method .................................................................................................29 1.1.33. Requirements ..............................................................................................................29 6.10. VOIP TELEPHONE DELAY ................................................................................................29 1.1.34. Requirement................................................................................................................29 1.1.35. Handset Send Delay....................................................................................................29 1.1.36. Handset Receive Delay...............................................................................................30 7. HEADSET TECHNICAL REQUIREMENTS......................................................................31 7.1. HEADSET FREQUENCY RESPONSE .................................................................................31 1.1.37. Headset Send Frequency Response ............................................................................31 1.1.38. Headset Receive Frequency Response .......................................................................34 7.2. HEADSET WIDEBAND LOUDNESS RATINGS .................................................................37 1.1.39. Headset Wideband Send Loudness Rating (SLR) ......................................................37 ii SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 1.1.40. Headset Wideband Receive Loudness Rating (RLR) ................................................ 37 1.1.41. Headset Talker Sidetone ............................................................................................ 38 7.2.2. HEADSET SIDETONE DELAY ........................................................................................... 38 7.3. HEADSET NOISE ............................................................................................................... 38 1.1.42. Headset Send Noise.................................................................................................... 38 1.1.43. 7.3.2 Headset Send Single Frequency Interference................................................. 39 1.1.44. ........................................................................................................................................ 39 1.1.45. 7.3.3 Headset Receive Noise................................................................................... 39 1.1.46. Headset Receive Single Frequency Interference........................................................ 40 7.4. HEADSET DISTORTION AND NOISE ............................................................................... 40 1.1.47. Headset Send Distortion and Noise............................................................................ 40 1.1.48. Headset Receive Distortion and Noise....................................................................... 41 7.5. WEIGHTED TERMINAL COUPLING LOSS (TCLW) ........................................................ 42 1.1.49. Measurement Method................................................................................................. 42 1.1.50. Requirements.............................................................................................................. 43 7.6. HEADSET LONG DURATION MAXIMUM ACOUSTIC PRESSURE (STEADY STATE)... 44 1.1.51. Requirements...............................................................Error! Bookmark not defined. 7.7. SHORT DURATION MAXIMUM ACOUSTIC PRESSURE (PEAK).................................... 45 1.1.52. Requirements.............................................................................................................. 45 8. SPEAKERPHONE TECHNICAL REQUIREMENTS (ADVISORY) .............................. 46 8.1. SPEAKERPHONE FREQUENCY RESPONSE ..................................................................... 46 1.1.53. Speakerphone Send Frequency Response .................................................................. 46 1.1.54. Speakerphone Receive Frequency Response ............................................................. 48 8.2. SPEAKERPHONE WIDEBAND LOUDNESS RATINGS AND RECEIVE VOLUME CONTROL ........................................................................................................................................ 49 1.1.55. Speakerphone Wideband Send Loudness Rating (SLR)............................................ 49 1.1.56. Speakerphone Wideband Receive Loudness Rating (RLR)....................................... 50 1.1.57. Speakerphone Receive Volume Control .................................................................... 50 8.3. SPEAKERPHONE NOISE ................................................................................................... 50 1.1.58. Speakerphone Send Noise.......................................................................................... 50 1.1.59. Speakerphone Send Single Frequency Interference................................................... 50 1.1.60. Speakerphone Receive Noise ..................................................................................... 51 1.1.61. Speakerphone Receive Single Frequency Interference .............................................. 51 8.4. SPEAKERPHONE DISTORTION AND NOISE .................................................................... 51 1.1.62. Speakerphone Send Distortion and Noise.................................................................. 52 1.1.63. Speakerphone Receive Distortion and Noise ............................................................. 52 8.5. WEIGHTED TERMINAL COUPLING LOSS (TCLW) ........................................................ 54 1.1.64. Measurement Method................................................................................................. 54 1.1.65. Requirements.............................................................................................................. 54 8.6. STABILITY LOSS ............................................................................................................... 54 1.1.66. Measurement Method................................................................................................. 54 1.1.67. Requirement ............................................................................................................... 54 ANNEX A (NORMATIVE) – CALCULATION OF LOUDNESS RATINGS ............................. 55 ANNEX B (INFORMATIVE) – MEASUREMENT AND LEVEL CONVERSIONS ................. 58 ANNEX C (INFORMATIVE) – R40 PREFERRED FREQUENCIES ......................................... 60 iii SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) ANNEX D (NORMATIVE) – DRP TO ERP TRANSFER FUNCTION .......................................61 iv SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) FOREWORD (This foreword is not part of this standard.) This document is a TIA Telecommunications standard produced by Working Group TR-41.3.3 of Committee TR-41. This standard was developed in accordance with TIA procedural guidelines, and represents the consensus position of the Working Group and its parent Subcommittee TR-41.3, which served as the formulating group. This standard is based on TIA-920. The TR-41.3.3 VoIP and PCM Transmission Performance Working Group acknowledge the contribution made by the following individuals in the development of this standard. Name: Representing: Al Baum Tom Harley Roger Britt Uniden Texas Instruments Nortel Miguel De Araujo Nortel Michael Chen VTech Juan Corona Steve Whitesell Joachim Pomy James Bress John Bareham Glenn Hess Amar Ray Steve Graham Allen Woo Kirit Patel Bob Young Ron Magnuson VTech VTech Avaya/ETSI AST Technology Labs Consultant MWM Acoustics Embarq Plantronics Plantronics Cisco Systems Consultant Consultant Chair Editor Emeritus Chair Copyrighted parts of ITU-T Appendix I to Recommendation G.113 and Recommendation P.79 are used with permission of the ITU. The ITU owns the copyright for the ITU Recommendations. Copyrighted parts of ISO 3 are used with permission of the ISO. The ISO owns the copyright for the ISO Standards. Suggestions for improvement of this standard are welcome. They should be sent to: Telecommunications Industry Association Engineering Department Suite 300 250 Wilson Boulevard Arlington, VA 22201 ( http://www.tiaonline.org ) v SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 1. Introduction This revision of TIA-920 establishes handset, headset and speakerphone telephone audio performance requirements for wideband digital telephones regardless of protocol or digital format. A number of improvements and corrections have been made, particularly related to the use of improved ear simulators. This standard addresses wideband performance, where wideband is defined as the frequency range between 150 and 6800 Hz. Requirements for conventional narrowband telephony, in the frequency range between 300 and 3400 Hz are defined in ANSI/TIA-810-B. 1 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 2. Overview 2.1. Scope This standard establishes voice performance requirements for devices that function as a wideband digital telephone. Transmission may be over any digital interface including wired, or wireless, Local or Wide Area Networks, Firewire/IEEE1394, Universal Serial Bus (USB), public ISDN or digital over twisted pair wire. This includes TDM-based and packet-based (e.g. VoIP) telephones. These telephones may be connected through modems, Voice Gateways, wireless access points, PBXs, or personal computer-based telephones. Examples include, but are not limited to: ISDN telephones, digital proprietary telephones, VoIP telephones (corded and cordless), softphones (such as a laptop computers), IEEE 802.11 telephones, USB telephones, USB devices, DECT telephones, Bluetooth® Telephones and Bluetooth devices. For telephone systems that incorporate a Universal Serial Bus (USB) type interface or a Bluetooth type interface to a host (such as a laptop computer), it may be desirable for the USB or Bluetooth device to meet the requirements of relevant clauses of this standard, where the host device is assumed to have a 0 dB loss plan, in its default state. It may be desirable for the device to provide gain adjustment for both the send and receive channels. When connected to a host device, the full system shall then meet all of the associated requirements of this standard. A USB or Bluetooth device may have a handset, headset or speakerphone configuration. Technical requirements are specified for handset, headset and speakerphone modes of operation regardless of the technology used to couple the handset or headset to the telephone. The test measurement methods in this standard reference procedures in IEEE Standards 269, 269a and 1329 where applicable, as well as the appropriate ITU-T Recommendations. Several performance measurement procedures are established, each of which yields standardized measurement data that may be used for the determination of compliance with this standard. Although this document may reference specific procedures or test equipment the intent is not to be all-inclusive. Any measurement procedure and equipment that can result in an identical measurement is considered valid. NOTE - If the main purpose for testing to this standard is comparison testing of different products, rather than compliance testing, then it is important that identical test procedures and equipment be used when testing the different products. While the procedures may call out specific test points within the requirements, the full range of the requirements take precedence. 2.2. Limits of Applicability This standard is not intended to describe specific requirements for the following types of digital voice terminal equipment: telephones with carbon transmitters, ISDN terminal adapters, cellular voice terminals (cell phones), and group audio terminals. 2.3. Categories of Criteria Mandatory requirements are designated by the word "shall". Advisory requirements are designated by the word "should," or "may," or "desirable" which are used interchangeably in this standard. Advisory criteria represent product goals or are included in an effort to ensure universal product compatibility. Where both a mandatory and an advisory level are specified for the same criterion, the advisory level 1 Bluetooth is a registered trademark of the Bluetooth SIF. This standard and TIA do not endorse Bluetooth products or services. 2 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) represents a goal currently identifiable as having distinct compatibility or performance advantages toward which future designs should strive. 2.4. FCC Part 68 This standard is intended to be in conformity with Part 68 of the Federal Communications Commission (FCC) Rules and Regulations, but is not limited to the scope of those rules. In the event that Part 68 requirements are more stringent than those contained in this standard, the provisions of Part 68 apply. 2.5. Environmental This standard does not contain environmental requirements. Environmental requirements can be found in ANSI/TIA/EIA-571-B. 2.6. Safety This standard does not contain safety requirements. Compliance with the applicable UL and CSA safety standards may be required in certain locations. 3 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 3. Normative References The following standards contain provisions, which, through reference in this text, constitute provisions of this Standard. At the time of publication, the editions indicated were valid. All standards are subject to revision, and parties to agreements based on this Standard are encouraged to investigate the possibility of applying the most recent editions of the standards indicated below, or their successors. ANSI and TIA maintain registers of currently valid national standards published by them. [1] ANSI/EIA/TIA-571-B-1999, Environmental Considerations. [2] ANSI/EIA/TIA-810-B-2007, Transmission Requirements for Narrowband Voice over IP and Voice over PCM Digital Wireline Telephones. ( http://www.tiaonline.org/standards/ip/ ) [3] ANSI S1.4-1990, Sound Level Meters. [4] ASTM D 2240-2002, Standard Test Method for Rubber Property – Durometer Hardness [5] IEEE Standard 269-2002 & 269a-2006, Standard Methods for Measuring Transmission Performance of Analog and Digital Telephone Sets, Handsets, and Headsets. [6] IEEE Standard 1329-1999, Standard Method for Measuring Transmission Performance of HandsFree Telephone Sets. [7] IETF RFC 1890 AVT Profiles. [8] IETF RFC 1890 RTP Profile for Audio and Video Conferences with minimal control. [9] ISO 3: 1973 Preferred numbers - Series of preferred numbers. [10] ITU-T Recommendation G.122 (1993), Influence of national systems on stability and talker echo in international connections. [11] ITU-T Recommendation G.131 (2003), Talker Echo and Its Control. [12] ITU-T Recommendation G.711 (1988), Pulse code Modulation (PCM) of voice frequencies. [13] ITU-T Recommendation O.41 (1994), Psophometer for use on telephone-type circuits. [14] ITU-T Recommendation P.10 (1998), Vocabulary of terms of telephone transmission quality and telephone sets. [15] ITU-T Recommendation P.51 (1996), Artificial mouth. [16] ITU-T Recommendation P.57 (2005), Artificial ears. [17] ITU-T Recommendation P.58 (1996), Head and torso simulator for telephonometry. [18] ITU-T Recommendation P.64 (1999), Determination of sensitivity/frequency characteristics of local telephone systems. [19] ITU-T Recommendation P.79 (1999), Calculation of loudness ratings for telephone sets. [20] ITU-T Recommendation P.79, Annex G (2001), Wideband loudness rating algorithm 4 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) [21] ITU-T Recommendation P.311 (1998), Transmission characteristics for wideband (150-7000 Hz) digital handset telephones. [22] ITU-T Recommendation P.341 (1998), Transmission characteristics for wideband (150-7000 Hz) digital handsfree telephony terminals. [23] ITU-T Recommendation P.360 (1998), Efficiency of devices for preventing the occurrence of excessive acoustic pressure by telephone receivers. [24] ITU-T Recommendation P.501 (2000), Test signals for use in telephonometry. [25] ITU-T Recommendation P.1010 (2004), Fundamental voice transmission objectives for VoIP terminals and gateways 5 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 4. Definitions, Abbreviations and Acronyms For the purposes of this Standard, the following definitions apply. 4.1. Codec A codec is a combination of an analog-to-digital encoder and a digital-to-analog decoder operating in opposite directions of transmission in the same equipment. An example of such a codec is the reference codec, which is described later. In this document if the context does not indicate the reference codec, then the term codec refers to the digital voice signal coder and encoder. For the purposes of this standard L16-256 is the linear 16 bit codec as defined by RFC 1890 at 256 kilobits per second (16 bits per sample and 16,000 samples per second). 4.2. Ear Reference Point (ERP) A virtual point for geometric reference located at the entrance to the listener's ear, traditionally used for calculating telephonometric loudness ratings. 4.3. Artificial Ear/Mouth vs. Ear/Mouth Simulator This standard uses the terms “ear simulator” and “mouth simulator” synonymously with the terms “artificial ear” (ITU-T P.57) and “artificial mouth” (ITU-T P.51), respectively, to harmonize with IEEE Std 269 and 269a. 4.4. HATS Position The HATS (head and torso simulator) position (ITU-T P.64 Annex D and Annex E) is the correct handset position for measuring sensitivity and frequency response characteristics. The HATS position has been shown to be essentially identical to the LRGP (loudness rating guard-ring position) position, except for the mouth simulator direction, which has been corrected with a 19 degree downwards rotation to more closely match real talkers. For handsets with omnidirectional microphones, measurements on the two heads may differ slightly, typically less than 1 dB. For handsets with directional or noise-canceling microphones, the differences will be larger, and the HATS position will give the more realistic results. Some equipment may use the term “LRGP-H” for the HATS position. 4.5. Nominal Volume Control Setting The nominal volume control setting is the receive volume control setting that results in the RLR closest to the nominal RLR value. All tests shall be performed with the receive volume control set to the nominal volume control setting, unless otherwise specified. For handsets the RLR is measured with the receiver in the high leak position. 4.6. Reference Volume Control Setting The reference volume control setting is the quietest volume control setting that complies with the mandatory low leak RLR requirement in 6.2.3. 4.7. Preferred Ear Simulator The preferred ear simulator is the Type 3.3. For alternative ear simulators, see relevant sections of IEEE Std 269. 4.8. Standard Test Position The Standard Test Position consists of a high leak position and a low leak position. The high leak position is defined as the Type 3.3 artificial ear with the receiver contacting the pinna with a force of 10 N. 6 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) The low leak position is defined as the Type 3.3 artificial ear with the receiver contacting the pinna with a force of 18 N. 4.9. Recommended Test Positions (RTP) RTP for a handset may be defined (using coordinates as defined in ITU-T P.64, Annex E) by following these steps: 1. Find the Ear-Cap Reference Point (ECRP) on the handset. Unless otherwise specified by the manufacturer, the ECRP is the intersection of the external ear-cap reference plane with a normal axis through the effective acoustic center of the sound outlet ports. Generally, the acoustic center of the sound outlet ports is at the center of their distribution. For some handsets, the ear cap reference plane has to be estimated. For example, a tangent to a curved surface at the effective acoustic center. 2. Line up the handset ECRP at the Ear simulator ERP on the positioning device. The ear cap reference plane shall be identical to the reference plane of the positioning device. When the positioning device is set to ERP, then the ear cap reference plane, the reference plane of the positioning device and the ERP plane are identical. 3. Translation: Move the handset ECRP in ear cap reference plane relative to ERP along the ye and/or ze axis. If no coordinates are given, leave the ECRP centered on ERP, equivalent to (0, 0) coordinates. The ye -axis is defined along the length of the handset with positive ye being in a direction towards the microphone from ECRP. Positive ze axis is in a downward direction towards the floor. 4. Rotation: Adjust the angle(s) of the handset positioner about the ERP of the coordinate system. 5. If no coordinates are given, use angles consistent with the HATS position. 6. Application of force: Adjust pressure or distance along the axis of motion of the positioning device. This axis is defined by a line that passes through the ERP of the left and right ears. It is parallel to the Ym axis. If no force or position is given, use 6N. 7. RTP can then be defined as the combination of the above translation, rotation and force specifications. The manufacturer of the device under test is responsible for providing this data. 4.10. Mouth Reference Point (MRP) The mouth reference point is located on axis and 25 mm in front of the lip plane of a mouth simulator. 4.11. Session Description Protocol (SDP) Session Description Protocol is a standard way of defining dynamically an RTP media payload (media format). RFC stands for Request for Comment and is used by the Internet Engineering Task Force (IETF) to define IP protocol standards. 4.12. Reference Codec A reference codec is used for testing digital telephone terminals with analog test equipment. Error! Reference source not found. shows the basic test setup using a reference codec. A codec that approaches an ideal codec and has superior, well-defined, characteristics qualifies as a reference codec. 7 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) When a 0.775 volt rms analog signal is applied to the coder input, a 0 dBm0 digital code is present at the digital reference. When a 0 dBm0 digital code is applied to the decoder, a 0.775 volt rms analog signal appears at the decoder output. At the digital reference point, 0 dBm0 is 3.17 dB below digital full scale. Although power levels are referenced to 600 ohms, the reference codec does not require physical 600 ohm source and termination resistors. The coder input impedance is high relative to the generator and the decoder output impedance is low relative to the measuring voltmeter. The interface block, shown in Error! Reference source not found. and Error! Reference source not found., passes the voice channel digital bit stream to the terminal without modification. There is no gain or loss in the send or receive direction due to the interface. If the interface does change the digital voice stream then the terminal and interface shall be considered jointly as the terminal. An example of this is a receive volume control implemented in a PBX or gateway. Figure 1 – Digital Telephone Set Test Arrangement with Reference Codec Digital Reference Point (Junction j) Send pM Mouth Sound Pressure at MRP Digital Set Decoder v Coder GEN Interface pE Ear Sound Pressure at ERP vRCV Receive 4.13. vSEND Reference Codec Direct Digital Processing Direct digital generation of the receive signal and analysis of the send signal may be used in place of the reference codec as shown in Error! Reference source not found.. Although this method is preferred, the test methodology usually refers to Error! Reference source not found., the reference codec method, for the sake of clarity. 8 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Figure 2 – Digital Telephone Set Test Arrangement using Direct Digital Generation and Analysis Digital Reference Point (Junction j) Send Digital Analysis pM Mouth Sound Pressure at MRP Digital Set Interface Digital Generation pE Ear Sound Pressure at ERP Receive 4.14. Sound Pressure Levels Sound pressure level is a value expressed as a ratio of the pressure of a sound to a reference pressure. The following sound level units are used in this standard: dBPa: The sound pressure level, in decibels, of a sound is 20 times the logarithm to the base 10 of the ratio of the pressure of this sound to the reference pressure of 1 Pascal (Pa). Note: 1 Pa = 1 N/m2. dBSPL: The sound pressure level, in decibels, of a sound is 20 times the logarithm to the base 10 of the ratio of the pressure of this sound to the reference pressure of 2 X 10-5 N/m2 (0 dBPa corresponds to 94 dBSPL). dBA: The A-weighted sound level is the sound pressure level in dBSPL, weighted by use of metering characteristics and A-weighting specified in ANSI S1.4. 4.15. Electric Power and Noise Levels The following electric power and noise level units are used in this standard: dBm0: 4.16. The absolute power level at a digital reference point of the same signal that would be measured as the absolute power level, in dBm, if the reference point was analog. The absolute power in dBm is defined as 10 log (power in mW / 1 mW). When the impedance is 600 ohm resistive, dBm can be referred to a voltage of 0.775 volts, which results in a reference active power of 1 mW. Note that 0 dBm0 is not the maximum digital code. For Mu law and L16-256 wideband codecs, 0 dBm0 is 3.17 dB below digital full scale. 50TP The acoustic test point 50 cm from the front center of the speakerphone telephone and 30 cm above the test table. 4.17. Abbreviations and Acronyms Abbreviations and acronyms, other than in common usage, which appear in this standard, are defined below. AGC Automatic Gain Control 9 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) CSS DRP EFR ERP FFT HATS ISDN LRGP L16-256 MRP OLR PBX PCM RLR RFC RTP RTP SDP SLR STMR TCLT TCLw VAD VoIP 4.18. Composite Source Signal Drum Reference Point Enhanced Full Rate Ear Reference Point Fast Fourier Transform Head and Torso Simulator Integrated Services Digital Network Loudness Rating Guard-ring Position Linear, Sixteen Bit, 256 kbit/s Codec Mouth Reference Point Overall Loudness Rating Private Branch Exchange Pulse Code Modulation Receive Loudness Rating Request for Comment (used by the IETF to define IP Protocol) Real Time Protocol Recommended Test Position Session Description Protocol Send Loudness Rating Sidetone Masking Rating Temporally weighted Terminal Coupling Loss Weighted Terminal Coupling Loss Voice Activity Detector Voice over Internet Protocol Test Signals 4.18.1. Choice of Test Signal IEEE Std. 269 & 269a, IEEE Std. 1329 and ITU-T Recommendation P.501 have recommendations on which test signals are appropriate. In particular, see Annex F, “Test Signals” and Annex G, “Analysis Methods” of IEEE Std. 269. The bandwidth of the test signal used shall nominally cover 100 Hz to 8000 Hz unless otherwise specified. The test signal used should be stated. The test signal levels specified in this standard shall be used. Test signal levels that differ from those specified in this standard may also be required. Algorithmic processes, such as Echo Control, VAD and AGC, may influence the test results or require test signals other than sinusoidal. Speakerphone modes of operation almost always employ algorithms that are sensitive to the temporal and spectral characteristics of the signals. For these devices the use of sinusoidal test signals may not be appropriate and should be used with caution. Unless explicitly stated otherwise in this document, for testing to determine TIA-920A compliance tones shall not be used to measure the following. Send, Receive, and Sidetone Spectral Response, Send, Receive, and Sidetone Loudness, TCLw and Stability. 4.18.2. Frequency Tolerance of Test Signals and Analysis Test signals shall have a frequency within 3% of the specified value. The range shall be within 3% of the specified test range. This is to accommodate different generation and analysis methods. 10 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Analysis using sine waves may be done at R40 preferred frequencies or at 1/12th octave band center frequencies. Analysis using other test signals shall be done in 1/12th octave bands unless otherwise specified. R40 analysis shall be done from 100 Hz to 8000 Hz. One-twelfth octave analysis shall be done from 92 Hz to 7286 Hz. Wherever a frequency range of 100 to 8000 Hz is specified, a range of 92 Hz to 7286 Hz may be used. 4.19. Testing mode Technical requirements apply to the telephone operating in the PCM 256 kbit/s mode. The telephone must be able to be put into the L16-256 codec mode regardless of which other codecs are supported. 4.20. Precautions Coding, decoding, packetization and other signal processing may introduce significant delays that must be accounted for by the measurement system. For example, when measuring the frequency response of a device using a stepped sine signal it is possible for the generator and measuring device to track incorrectly. The measuring device might be measuring the level of delayed frequency, fn, while the generator has progressed to fn+1. Refer to IEEE Std. 269 & 269a for additional precautions regarding test signal usage. Telephones using nonlinear voice signal processing may require subjective testing to validate or supplement objective measurement. 11 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 5. General Technical Requirements 5.1. Voice Coding Mandatory Requirements A telephone supporting wideband operation shall support the L16-256 codec. Note: Existing wideband conferencing systems commonly use the G.722 and G.722.1 codecs. Interoperability with existing systems requires additional codecs either within the telephone itself or in other system equipment such as voice gateways. If the telephone uses additional vocoders, the manufacturer must ensure that their implementation passes the standard test vectors associated with that codec. For bit exact vocoders it is important to ensure that vector testing has been performed and found to be compliant with the associated ITU requirement. Technical requirements of this standard apply only to linear PCM L16-256 codec at 256 kbit/s. Unless specified otherwise: • • Test methods are given in IEEE Std 269. When sine wave stimulus is used the frequency tolerance is 3%, and even submultiples of the sampling frequency (normally 16000 Hz) must not be used. Algorithmic processes, such as Echo Control, VAD and AGC, may influence the test results or require test signals other than sine waves. IEEE Standards 269, 269 allows several types of test signals. The test signal used should be stated. The test signal levels shall be equivalent to the test signal levels specified in this standard. Test signal levels that differ from those specified in this standard may also be required. Note that the use of inappropriate test signals may result in erroneous test results. Packet voice latency may introduce significant delay that must be accounted for by the test equipment. Telephones using nonlinear voice signal processing may require subjective testing. NOTE: For telephones where tandem codecs (other than L16-256) are used (e.g. cordless interface to a telephone set), the codec may affect the test results and some voice transmission technologies may be unable to meet specified noise and/or distortion requirements. Such devices need further investigation. 5.1.1. Transmission Format of L16-256 Codec L16-256 denotes uncompressed 16-bit linear PCM coding of wideband speech sampled at 16 kHz having a bit rate of 256 kbit/s. The L16-256 coder shall use 16-bit signed representation with 65535 equally divided steps between minimum and maximum signal level from -32768 to 32767 given in Table 1. The value will be represented in two’s complement notation and transmitted in network byte order with the most significant byte first. Table 1 – Codec PCM Codes Reference Digital representation 0 dBm0 See sub clause 6.1.4 12 Level Relative to 0 dBm0 (dB) 0 RMS Analog Voltage (V) 0.775 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 3.17 +Full Scale ± 32767 32767 is x7FFF hex ⎯ ⎯ -Full Scale -32768 is x8000 hex. ⎯ ⎯ ⎯ ⎯ Full Scale Quiet Code x0000 hex Note: All values except the Quiet Code are sinusoidal. 1.116 5.1.2. Overload Point Digital full scale is +3.17 dB above the 0 dBm level. 5.1.3. Quiet Code and Full Scale Code Quiet code is the digital code representing the smallest encoded analog level. Full scale code is the digital code representing the largest encoded analog level. 5.1.4. 0 dBm0 (Digital Milliwatt) The 1 kHz 0 dBm0 Sine wave is represented by the following values: Sample [ 0] Sample [ 1] Sample [ 2] Sample [ 3] Sample [ 4] Sample [ 5] Sample [ 6] Sample [ 7] Sample [ 8] Sample [ 9] Sample [10] Sample [11] Sample [12] Sample [13] Sample [14] Sample [15] = 0 = 8705 = 16085 = 21016 = 22748 = 21016 = 16085 = 8705 = 0 = -8705 = -16085 = -21016 = -22748 = -21016 = -16085 = -8705 The values are calculated for a 0 dBm0 level which is 3.17 dB below the digitally encoded peak of + 32767. When converting from A-Law, Mu-Law or other encoding formats the precise code representation may differ somewhat due to processing errors. Care should be taken with conversion algorithms to minimize distortion. 13 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 6. Handset Technical Requirements All tests shall be performed with a Type 3.3 ear simulator with the handset in the HATS position. The Type 3.3 ear simulator shall comply with the specifications given in ITU-T Recommendation P.57. The Type 3.3 shall have a hardness of 35 ±6 degrees Shore-OO, as measured according to ASTM 2240. (ITU-T Recommendation P.57-2002 originally specified a hardness of 55 ±10 degrees ShoreOO for Type 3.3.) Unless otherwise specified, all tests shall be performed in the Standard Test Position with the receiver located at the high leak position. A manufacturer may specify Recommended Test Positions for the high leak and the low leak positions. The RTP may specify the handset position with respect to ERP, or other aspects of the test position intended to simulate actual use. If the CPE is tested at the RTP, the RTP shall be documented and used for all handset tests. All tests shall be preformed with the receive volume control set to the nominal volume control setting, unless otherwise specified. 6.1. Handset Frequency Response 6.1.1. Handset Send Frequency Response The send frequency response is the overall response of the transducer, send amplifier, and the codec send filter. Send sensitivity is the ratio of the voltage output of the reference codec, or digital bit stream equivalent, to the sound pressure at the Mouth Reference Point (MRP) for each frequency or frequency band (Fi) as shown in the equation below: SMJ = 20 log (VSEND / PM) dB rel 1 V / Pa Equation [1] Where SMJ PM VSEND Send Sensitivity, Mouth to Junction, at Fi. Sound pressure at the MRP at Fi. RMS output voltage of the reference codec at Fi. 6.1.1.1. Measurement Method Measurements should be done in ISO 1/12th octave bands or R40 intervals or smaller, over a minimum range of 100 Hz through 8000 Hz using the measurement set-up shown in Figure 3. Direct digital processing may be employed as explained in clause 4.13. The test signal level shall be −4.7 dBPa at the MRP. 14 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Figure 3 – Handset Send Frequency Response Measurement Method vSEND HATS Decoder Measuring Amplifier Digital Interface GEN Set Send pM v Coder Mouth Simulator Quiet Room Reference Codec 6.1.1.2. Requirement The send frequency response shall fall between the upper and lower limits given in Table 2 and shown in Figure 4. The limit curves shall be determined by straight lines joining successive coordinates given in the table, where frequency response is plotted on a linear dB scale against frequency on a logarithmic scale. Note: The frequency response mask is a floating or “best fit” mask. Figure 4 includes the nominal send frequency response characteristic to illustrate the design intent of the limits. Table 2 – Co-ordinates of Handset Send Response Limits Limit Curve Frequency (Hz) Send Response Limit (dB) [arbitrary level] Upper Limit 100 140 1000 2000 5000 8000 -1 +3 +3 +8 +8 +3 Lower Limit 200 200 250 1000 3000 5000 6500 6500 - infinity -6 -3 -3 -1 -1 -6 - infinity 15 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Figure 4 – Handset Send Frequency Response Mask 10 Arbitrary Level (dB) 5 0 -5 -10 -15 -20 100 1000 Frequency (Hz) 10000 6.1.2. Handset Receive Frequency Response Receive frequency response is the ratio of the sound pressure measured in the ear simulator to the voltage input to the reference codec, or digital bit stream equivalent, for each frequency or frequency band (Fi) as shown in the equation below: = 20 log (PE / VRCV) dB rel 1 Pa / V SJE Equation [2] Where SJE PE VRCV Receive Sensitivity, Junction to Ear, at Fi. ERP Sound pressure measured by ear simulator at Fi. Measurement data are converted from the Drum Reference Point, DRP, to the ERP. RMS Input voltage to the reference codec, or digital bit stream equivalent at Fi. 6.1.2.1. Measurement Method The receive frequency response shall be measured with the receiver at the high leak position. The receive frequency response is measured using the measurement set-up shown in Figure 5. Direct digital processing may be employed as explained in clause 4.13. Measurements should be done in 1/12th octave bands, or R40 intervals over a range of 100 Hz through 3350 Hz. Measurements should be done in 1/3rd octave bands for the R40 frequencies of 4000, 5000, 6300, 8000 Hz. The test signal level shall be -18.2 dBV (-16 dBm0), or digital bit stream equivalent. The frequency response measured with the ear simulator must be transformed to the ear reference point (ERP). Note: It is useful to look at 1/12th octave resolution all the way up to 8000 kHz, in order to better understand variability in the receive side, low pass characteristics, etc. 16 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Figure 5 – Handset Receive Frequency Response Measurement Method Receive p E Ear Simulator Decoder Digital Sound Pressure Measuring Amplifier Interface Set Coder GEN vRCV Quiet Room Reference Codec 6.1.2.2. Requirement The receive frequency response requirements between 100 Hz and 8000 Hz (referenced to the ERP) are as follows: 1. With the receiver at the high leak position, the receive frequency response: Shall fall within the mandatory limits in a. Table 3 (shown in Figure 6). Should fall within the desired limits in b. Table 3 (shown in Figure 6). The limit curves shall be determined by straight lines joining successive co-ordinates given in the table, where frequency response is plotted on a linear dB scale against frequency on a logarithmic scale. The frequency response mask is a floating or “best fit” mask. See Error! Reference source not found. for the details on the derivation of the nominal frequency response characteristics. Table 3 – Co-ordinates of Handset Receive Response Limits 17 Formatted: Normal, Left Formatted: Normal, Left Formatted: Normal, Left Deleted: ¶ SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Limit Curve Upper Limit Lower Limit Frequency (Hz) Desirable Receive Response Limit (dB) [arbitrary level] Mandatory Receive Response Limit (dB) [arbitrary level] 100 130 700 1000 1400 2000 4000 8000 200 200 800 4000 6800 6800 +1 +4 +1 +4 +4 Deleted: +4 +9 +9 +9 +8 - infinity -10 -4 -4 -10 -infinity +9 +8 - infinity -10 -4 -4 -10 -infinity Deleted: ¶ ¶ Figure 6 – Handset Receive Frequency Response Mask 15 Arbitrary Level (dB) 10 5 0 -5 -10 -15 100 1000 Frequency (Hz) 10000 Note: The lower mask above has no relative maximum around 3 kHz. A prior intention for this wideband handset receive lower mask to have a “3 kHz bump” is evident from the TIA-810B Appendix. The TIA-920A working group’s new intent is to consistently remove the 3 kHz bump from the narrow band in the future TIA-810C standard. Deleted: ¶ Deleted: Note; The above response resulted from a contribution from Steve Graham. But the lower mask “bump” starting at 2 kHz (that harmonizes the narrow band response) was not unanimously accepted. The following response mask is submitted for discussion by Steve Graham:¶ H e a d s e t R e c e iv e F re q u e n y R e s p o n s e T e m p la t e 15 6.2. Handset Wideband Loudness Ratings and Receive Volume Control Arbitrary Level (dB) 10 5 0 -5 -1 0 -1 5 -2 0 -2 5 100 1000 F re q u e n c y (H z) 18 1000 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) The loudness rating algorithm is defined in ITU-T Recommendation P.79 and summarized in Annex A. Loudness ratings are calculated from a telephone’s send, receive and sidetone frequency response measurement data. They provide single number metrics, which describe how loud the telephone will sound to a user. An important characteristic of P.79 loudness ratings is that the louder the telephone, the more negative the loudness rating. 6.2.1. Handset Wideband Send Loudness Rating (SLR) The SLR is the loudness loss in the send direction from the acoustic signal at the mouth reference point to the send signal at the digital reference point. Refer to Annex A and ITU-T Recommendation P.79. 6.2.1.1. Measurement Method The SLR shall be calculated from the send frequency response measurement (clause 6.1.1) using equations Error! Reference source not found. and Error! Reference source not found. in Error! Reference source not found. and frequency bands 1 to 20, Table 14. 6.2.1.2. Requirement The terminal shall be designed to have an SLR value of 8 dB, with a tolerance of ±4.0 dB. 6.2.2. Handset Wideband Receive Loudness Rating (RLR) The RLR is defined in Annex A. 6.2.2.1. Measurement Method The RLR shall be calculated from the receive frequency response measurement (clause Error! Reference source not found.) using equations Error! Reference source not found. and Error! Reference source not found. in Error! Reference source not found. and frequency bands 1 to 20, in Table 14. 6.2.2.2. Requirement The RLR values measured with the receiver at the high leak position shall have an RLR value of 2 dB, with a tolerance of -4.0/+8.0 dB and should have a nominal RLR value of 2 dB, with a tolerance of ±4.0 dB. 6.2.3. Handset Receive Volume Control Range The current regulatory volume control requirements are specified in 47 CFR Part 68.317. NOTE: The wideband RLR measurements in this document use the HATS (Head and Torso Simulator) while the current 47 CFR Part 68.317 references narrowband ROLR measurements in ANSI/EIA/TIA-579-1991, which specifies the Type 1 ear (ITU-T Recommendation P.57). 6.2.3.1. Measurement Method The RLR shall be calculated from the receive frequency response measured with the receiver at the low leak position. The measurement shall be done with the volume control at either the Reference Volume Control Setting, or the manufacturer’s defined reference volume setting, and the maximum setting. Use equations Error! Reference source not found. and Error! Reference source not found. in Error! Reference source not found. and frequency bands 1 to 20, Table 14. Measure the receive distortion and noise (see clause Error! Reference source not found.) with a 1004 Hz sine wave at -16 dBm0 input at the maximum volume control setting. 19 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 6.2.3.2. Requirement The RLR values measured with the receiver at the low leak position shall have a nominal RLR value of 2 dB, with a tolerance of ±4.0 dB at the Reference Volume Control Setting or the manufacturer’s defined reference volume setting. With the receiver at the low leak position the RLR at the maximum volume control setting shall be at least 12 dB louder than the RLR at the Reference Volume Control Setting, or the manufacturer defined reference volume setting (also measured at the low leak position). If the RLR at the maximum volume control setting is more than 18 dB louder than the RLR at the Reference Volume Control Setting or the manufacturer’s defined reference volume setting, then the CPE shall automatically reset to either the Nominal Volume Control Setting, Reference Volume Control Setting, or the manufacturer’s defined reference volume setting, after ending the call. To ensure that there is no significant clipping, the receive signal to total distortion and noise ratio at the maximum volume control setting shall be greater than 20 dB with a 1004 Hz, -16 dBm0 input. (See clause 6.5.2 for test method.) NOTE: Some special purpose CPE provide high receive gain for hearing impaired users. These CPE are intended to provide the highest gain for below normal input signal levels, and they might fail the distortion requirement at the maximum volume control setting when measured with the specified test signal level. 6.2.4. Magnetic Field for Hearing Aid Coupling This standard does not contain Magnetic Field for Hearing Aid Coupling requirements. The current regulatory hearing aid compatibility magnetic output requirements are specified in 47 CFR Part 68.316. NOTE: Part 68.316 does not provide suitable references for testing Digital Telephones. Suitable test procedures are currently in TSB-31-C-1, Part 68 Rationale and Measurement Guidelines." 6.2.5. Handset Talker Sidetone (STMR) The sidetone masking rating (STMR) of a digital telephone set is the loudness of the path from the mouth to the ear of the same headset. STMR is calculated from the ratio of the acoustic output signal from the receiver at the ear reference point (ERP) to the acoustic input signal at the mouth reference point (MRP) over the specified frequency band. It’s desirable for the STMR to be constant over the receive volume control range. 6.2.5.1. Measurement Method The test signal level at the MRP shall be -4.7 dBPa. For each frequency given in Table 14, bands 1 to 20, the sound pressure in the artificial ear shall be measured. The frequency response measured with the ear simulator must be transformed to the ear reference point (ERP). Refer to Annex D. The STMR shall be calculated using equation Error! Reference source not found. of Error! Reference source not found.. Telephone sets with adjustable receive levels shall be tested at the minimum, nominal and maximum settings. 6.2.5.2. Requirement The value of STMR shall be within the range of 18 dB ± 6 dB, for any adjustable receive level. 20 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 6.2.6. Handset Sidetone Delay In a digital telephone, sidetone echo occurs when significant delay is introduced into the speech path between the handset microphone and the handset receiver by the sidetone feedback algorithm. Ideally, the sidetone signal should be a real-time signal. Sidetone delay less than 5 ms is generally perceived as normal sidetone. Sidetone delay between 5 and 10 ms is generally perceived as unnatural sidetone, with an uncomfortable hollow characteristic. Sidetone delay greater than 10 ms is generally perceived as a distinct talker echo signal. Since the sidetone level could be as loud, or louder than a talker echo signal, sidetone delay greater than 5 ms is undesirable. 6.2.6.1. Measurement Method See the method described in IEEE Standards 269, 269. 6.2.6.2. Requirement Sidetone delay shall be less than 5 ms. Sidetone delay should be less than 1 ms. 6.3. Handset Noise 6.3.1. Handset Send Noise 6.3.1.1. General The send noise of a digital telephone is the 5 second average noise level at the digital transmit output with the telephone transmitter isolated from sound input and mechanical disturbances. 6.3.1.2. Measurement Method In a quiet environment (ambient noise less than 30 dBA), free of mechanical disturbances, measure the A-weighted, 5 second average, noise level at the digital interface output or the reference codec decoder output over the frequency range of 100 to 8000 Hz. 6.3.1.3. Requirement The overall send noise shall be less than -68 dBm0, A-weighted. 6.3.2. Handset Send Single Frequency Interference 6.3.2.1. General Narrow-band noise, including single frequency interference, is an impairment that can be perceived as a tone depending on its level relative to the overall weighted noise level. 6.3.2.2. Measurement Method In a quiet environment (ambient noise less than 30 dBA), free of mechanical disturbances, measure the A-weighted noise level at VSEND with a selective voltmeter or spectrum analyzer with an effective bandwidth of not more than 31 Hz, over the frequency range of 100 to 8000 Hz. If FFT analysis is used, then “Flat Top” windowing shall be employed. 6.3.2.3. Requirement The A-weighted send single frequency interference shall be less than -78 dBm0 in each band. 21 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 6.3.3. Handset Receive Noise 6.3.3.1. General The receive noise of a digital telephone is the 5 second average noise level measured at the output of the telephone receiver with the digital telephone receiving the digital quiet code. Receive noise measurement results must be transformed from the DRP of the ear simulator to the ERP. If a single wideband measurement is made the transfer function must be realized using a minimum phase, parametric filter (or equivalent). Refer to IEEE Std. 269. 6.3.3.2. Measurement Method A signal corresponding to a decoder quiet code is applied at the digital interface. The A-weighted noise level is measured in the artificial ear over the frequency range of 100 to 8500 Hz. The ambient noise for this measurement shall not exceed 30 dBA. 6.3.3.3. Requirement The receive noise shall be less than 40 dBA. 6.3.4. Handset Receive Single Frequency Interference 6.3.4.1. General Narrow-band noise, including single frequency interference, is an impairment that can be perceived as a tone depending on its level relative to the overall weighted noise level. This test measures the weighted noise level characteristics in narrow bands of not more than 31 Hz, which can then be compared to the overall weighted receive noise level. Narrow-band noise is measured at the output of the telephone receiver with the digital telephone receiving the digital quiet code. 6.3.4.2. Measurement Method A signal corresponding to a decoder quiet code is applied at the digital interface. The A-weighted noise level is measured in the artificial ear with a selective voltmeter or spectrum analyzer, with an effective bandwidth of not more then 31 Hz, over the frequency range of 100 to 8500 Hz. If FFT analysis is used, then “Flat Top” windowing shall be employed. The ambient noise for this measurement shall not exceed 30 dBA. 6.3.4.3. Requirement The receive single frequency interference shall be 10 dB quieter than the A-weighted receive noise, and shall be below 30 dBA. 6.4. Handset Receive Comfort Noise (Advisory) If comfort noise is introduced to replace actual background noise the level should roughly match the loudness of the original background noise. There is more likely to be annoyance if the comfort noise is greater than the original noise than if it is less than the original noise. 6.4.1. General The receive comfort noise of a digital telephone is the short-term average background noise level measured at the output of the telephone receiver with the terminal receiving either a silence indication packet or no packets for some non-transient period of time. 6.4.2. Measurement Method The digital interface is sent a quiet code ⎯ the code that represents silence for the coder format. 22 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Disable comfort noise generation and any echo canceller on the terminal. Apply a white noise test signal to the terminal and adjust the signal amplitude such that the receive noise level measured at the terminal is 48 dBA. The test signal level is assigned the level of ‘N dB’. Enable comfort noise generation on the terminal. The following test sequence must be followed for all test noise levels of ‘M dB’, which will range from N-10 to N+10 dB. 1. The digital interface is sent the quiet code for 10 seconds. 2. Apply 300-3400 Hz band-limited white noise of level M dB to the terminal for 60 seconds. 3. Stop sending data to the terminal; this should cause the comfort noise generator to trigger and apply comfort noise to the terminal receiver. Wait 10 seconds. 4. During the next 10 seconds the acoustic noise level at the receiver is measured. 5. Steps 1-4 are repeated for varying M in 5 dB increments. 6.4.3. Requirement For a test signal level of M = N, verify that the measured receive noise level is 48 dBA +0.5/-3.0. For all input noise levels M in the range of N-10 to N+10 the receive noise level measured must be within +0.5/-3.0 dB of the expected acoustic receive noise level for that input. This expected receive noise level for any given M and N would be 48 dBA - (N-M). Note; Some committee members feel this comfort noise section should be removed, or suggestions made for “black box” testing. Commercial phones often have no means to disable CNG and AEC. 6.5. Handset Distortion and Noise The distortion and noise requirements only apply to linear 16 bit PCM at 256 kbit/s. 6.5.1. Handset Send Distortion and Noise 6.5.1.1. Method of Measurement The highest signal levels employed for this test may exceed the published specifications for the mouth simulator. At high test levels, short duty cycles may be required to prevent overheating of the mouth simulator. Prior to testing the telephone, the output level and distortion of the mouth simulator should be verified at the maximum sound pressure level used for each frequency. This need not be verified before each test. The distortion of the mouth simulator should be at least 10 dB less than the maximum allowable telephone distortion for each frequency and level. The mouth verification should be done at the MRP. Apply a sine wave signal at the MRP, at the levels given in Table 4 at the following frequencies: 160, 315, 502, 803, 1004, 2008 and 3150 Hz. The ratio of the signal-to-total distortion and noise power of the digitally encoded signal output is measured. The test frequency tolerance is 3%, but even submultiples of the sampling frequency must not be used. Note: In cases where the sound pressure exceeds +6 dBPa, the linearity of the artificial mouth should be checked, as it exceeds the limits of ITU-T Recommendation P.51. 23 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 6.5.1.2. Requirement The ratio of signal-to-total distortion and noise (SDN) of the digitally encoded signal output shall be above the limits given in Table 4, with A-weighting applied to the measured distortion and noise output. Limits for intermediate levels are found by drawing straight lines between the successive coordinates in the table on a linear (dB signal level) – linear (dB ratio) scale. Table 4 – Handset Send Signal-to-Total Distortion and Noise Ratio Limits -30 Send Ratio 160 to 803 Hz (dB) 26 Send Ratio 1004 Hz (dB) 26 Send Ratio 2008 to 3150 Hz (dB) NA -25 26 26 26 -20 31 31 31 -10 33 33 33 0 33 33 33 Send Level at MRP (dB Pa) +5 33 33 33 +10 26 26 26 +15 NA 20 NA Note: 20 dB = 10%, 26 = 5%, 31 = 2.8%, 33 = 2.2%. NA = Not Applicable. Note; Fewer tests are warranted. If one column is emphasized, 400 Hz more appropriate because this is more prominent during speech. Move lower frequency boundary to 203 Hz (unless lower Send Mask frequency response revised to go down to 160 Hz). 6.5.2. Handset Receive Distortion and Noise 6.5.2.1. Method of Measurement Apply a digitally simulated sine wave, with the signal levels given in Table 5 and the following frequencies: 160, 315, 502, 1004, 2008 and 3150 Hz. The ratio of the signal-to-total distortion and noise power is measured with the artificial ear. The test frequency tolerance is 3%, and even submultiples of the sampling frequency must not be used. 6.5.2.2. Requirement The ratio of signal-to-total distortion and noise (SDN) measured with the artificial ear shall be above the limits given in Table 5, with A-weighting applied to the measured distortion and noise output, unless the signal in the artificial ear exceeds +10 dBPa or is less than -50 dBPa. 24 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Table 5 – Handset Receive Signal-to-Total Distortion and Noise Ratio Limits Receive level at the digital interface (dBm0) -40 Receive Ratio @ 160 Hz (dB) 20 Receive Ratio @ 315 Hz (dB) 24 Receive Ratio @ 502 to 3150 Hz (dB) 24 -34 24 24 24 -27 28 30 30 -20 28 32 32 -10 28 32 32 -6 28 32 32 -3 28 28 28 0 24 24 28 Note: 20 dB = 10%, 24= 6.3%, 28 = 4%, 30=3.2%, 32 = 2.5%. Note; Fewer tests are warranted. If one column is emphasized, 400 Hz more appropriate because this is more prominent during speech. Move lower frequency boundary to 203 Hz (unless lower Receive Mask frequency response revised to go down to 160 Hz). Tests that vary level only and frequency only might be listed separately. 6.6. Weighted Terminal Coupling Loss (TCLw) The weighted terminal coupling loss (TCLw) provides a measure of the echo performance under normal conversation, i.e., single far-end talker conditions. It is possible that echo control devices such as echo suppressors or echo cancellers with non-linear processing may be used on handset connections to provide sufficient echo return loss to mitigate increased echo associated with longer network delays. The use of echo control devices on the handset can affect the measurement of TCLw. The result would likely be different under cases of either single far-end talker or double-talk. The TCLw measurement is intended to represent a single far-end talker. The ‘proper’ measurement of TCLw is addressed in IEEE Std. 269, Annex O. 6.6.1. Measurement Method TCLw shall be measured with the handset receiver at the high leak Standard Test Position on the HATS. The TCLw measurement shall be made at an input signal level of -16 and -10 dBm0. The test should be performed in a quiet environment (the ambient noise level shall be less than 30 dBA.) The TCLw measurement shall not be performed using sinusoidal test signal for the receive path input. The test signal may be a composite source signal (CSS) as defined in ITU-T P.501 or bursted white noise. The test signal shall be band-limited to 100 through 8000 Hz. The calibration shall be determined during the ON portions of the signal. The measurement shall be performed after system stability is reached (including convergence of any echo algorithms); this shall be accomplished by invoking the test signal for at least 2 seconds before the actual measurement occurs. th The attenuation from digital input (Receive) to digital output (Send) is measured in 1/12 octave bands, using the measurement arrangement shown in Figure 7. See Error! Reference source not found.. 25 Deleted: This may provide idealized and unrealistic performance measurements when non-linear processing on the transmit side is used as part of the echo control algorithm. It may be more appropriate to measure TCLw either with non-linear processing disabled or with a near-end signal present that is a) capable of enabling echo control’s double-talk detector with the subsequent removal of non-linear processing and b) can be filtered out from the final return signal so as not to affect the accuracy of the TCLw measurement. The latter may be the only method that can be used consistently across products in a blackbox testing setup. A suitable signal may be a pulsed sine wave, but it will depend on the temporal characteristics of the double-talk detector.¶ ¶ Deleted: then becomes specific to the echo control implementation. These issues are still under study and are not addressed in these requirements. For further information see Deleted: SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) The weighted terminal coupling loss is calculated according to ITU-T G.122 Annex B.4 (Trapezoidal rule) using the frequency range of 300 to 6700 rather than 300 to 3400. Figure 7 – Terminal Coupling Loss Measurement Method vSEND (Echo Return) Ear Simulator Decoder v Coder GEN Digital Set Interface vRCV Anechoic Chamber Reference Codec 6.6.2. Requirements The normalized value of TCLw at the high leak position shall be greater than 52 dB for IP sets and 45 dB for PCM sets when measured under free field conditions and with SLR normalized to 8 dB and RLR normalized to 2 dB. It is desirable that the normalized value of TCLw for PCM sets be greater than 50 dB to meet ITU-T Recommendation G.131 talker echo objective requirements. For example, if the measured TCLw is 48 dB, the measured SLR is 11 dB and the measured RLR is 0 dB, then the normalized value of TCLw = TCLw measured + (8 - SLR) dB + (2 - RLR) dB = 48 dB + (8 - 11) dB + (2 - 0) dB = 47 dB. 26 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) NOTES 1. The requirement of 52 dB for IP sets is a function of the -16 dBm0 test signal level and the -68 dBm0A send noise requirement. Measuring TCLw > 52 dB can be difficult. 2. If equipped with adjustable receive level, the un-normalized TCLw will decrease in proportion with the increased gain relative to the nominal RLR in most cases. For example, if the measured TCLw is 45 dB at nominal RLR and the adjustable receive level adds 12 dB of gain, then un-normalized TCLw (maximum receive level) = 45 dB - 12 dB = 33 dB. 3. The echo impairment perceived by the person at the opposite end of the connection from a telephone set is a function of the magnitude of the talker echo signal as well as the talker echo path delay. The echo signal becomes more disturbing as the talker echo path delay increases. Thus, a telephone set with adequate TCLw performance on low delay connections may provide satisfactory performance while the same may not be true for connections that have a long delay. 4. Temporally weighted terminal coupling loss (TCLt) is an alternate method for echo measurement, which may be more subjectively relevant, especially in devices with echo suppression or cancellation features. (See IEEE Std 1329.) The performance requirements may need to be changed when using this method. Figure 8 – Reference Corner 25 cm 50 cm 6.7. Stability Loss The stability loss is a measure of the contribution of the telephone set to the overall network stability requirements. Stability loss is defined as the minimum loss from the digital input (receive) to the digital output (send), at any frequency. 27 Deleted: test SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 6.7.1. Measurement Method The stability measurement shall be made at input signal levels of -16 and -10 dBm0. The TCLw measurement shall not be performed using a sinusoidal test signal for the receive path input. The test signal may be CSS or bursted white noise, band-limited to 100 through 8000 Hz and represented by the L16-256 codec. The measurement and calibration shall be determined during the ON portions of the signal, not the average of on and off times. With the handset and transmission circuit fully active, measure the attenuation from the digital input to the digital output using Method 1 and Method 2. 6.7.1.1. Method 1 Place the handset in the reference corner, as shown in Figure 8, with the earcap and mouthpiece facing a hard, smooth surface. The handset shall be placed along the diagonal from the apex of the reference corner to the outside corner, with the earcap end of the handset 250 mm from the apex. The telephone set shall be fully active. The reference corner consists of three perpendicular plane, smooth, hard surfaces extending 0.5 m from the apex of the corner. 6.7.1.2. Method 2 Place the handset with the earcap and mouthpiece facing a hard, smooth surface free of any other object for 50 cm. The telephone set shall be fully active. 6.7.2. Requirement Stability loss using both method 1 and method 2, (i.e., minimum loss, at any frequency) shall be greater than 6 dB. It is desirable that this loss be greater than 10 dB. Telephone sets with adjustable receive level should maintain stability over the entire range of adjustable receive levels. 6.8. Long Duration Maximum Acoustic Pressure (Steady State Input) 6.8.1. General The long duration maximum acoustic pressure is the steady state (longer than 500 ms) sound pressure disturbance emitted from a telephone receiver, caused by the maximum excursions of the receive digital signal. Additional consideration should be given to the acoustic pressure caused by tones, other audio signals or long duration, high amplitude electrical signals applied to power, network, handset or auxiliary leads of the digital telephone. 6.8.2. Measurement Method The steady-state A-weighted sound pressure level shall be measured using the digital terminals test procedure in IEEE Standards 269, 269. 6.8.3. Requirements The measured maximum rms level shall be less than 125 dB(A) at ERP, required in UL/CSA 609502003. 6.9. Short Duration Maximum Acoustic Pressure (Peak) 6.9.1. General The short duration maximum acoustic pressure is the sound pressure impulse (less than 500 ms) emitted from a handset receiver. 28 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) This short duration test stresses nonlinear processes, like AGC, and doesn’t directly replace a short duration surge. Additional consideration should be given to the peak acoustic pressure caused by tones or short duration, high amplitude electrical pulses applied to power, network, handset or auxiliary leads of the digital telephone. 6.9.2. Measurement Method The peak acoustic pressure level shall be measured using the digital terminals test procedure in IEEE Standards 269, 269. 6.9.3. Requirements The maximum peak acoustic pressure shall be less than 136 dBSPL at ERP, as required in UL/CSA 60950-2003. 6.10. VoIP Telephone Delay Delay is a complex end-to-end issue. Certain aspects of delay can be optimized in VoIP telephones, such as the internal hardware/firmware delay and the optimization of the jitter buffer operation, which must trade-off the impairment of packet loss against the expected delay variation of the far-end telephone and/or the network. Other aspects, such as packetization and depacketization are also important sources of delay, but they are a function of the selected codec and the number of speech frames per packet, so they cannot be optimized in VoIP telephones. Therefore, this standard now specifies delay in terms of categories for network planning purposes, similar to ITU-T Rec. P.1010. When reporting compliance with this standard, only the category with the largest measured delay shall be reported, if the send and receive categories are different. If codecs or speech frame rates other than those specified in the measurement methods are used, then they must be clearly identified when reporting compliance. 6.10.1. Requirement The wired terminals shall be configurable so that requirements of at least Category B are met. Wireless terminals should be configurable so that the requirements of at least Category C are met. 6.10.2. Handset Send Delay 6.10.2.1. General The send delay is defined here as the time from when an acoustic signal leaves an artificial mouth playing into a VoIP telephone’s handset to the time its digitized, packetized representation arrives at that telephone’s packet network interface. 6.10.2.2. Measurement Method A digital audio measuring device capable of measuring the delay between an injected signal (to the mouth simulator) and a digitally transmitted signal should be connected to the artificial mouth and directly to the network output of the telephone. All delays inherent in the measurement system itself must be calibrated out. The telephone should be set to transmit L16-256 packets with a speech frame rate of 20 ms and with one speech frame per packet. An acoustic signal of -4.7 dBPa shall be generated at the artificial mouth. The delay between the time the pulse left the mouth to the time it was received at the telephone’s packet network interface shall be measured. The send delay shall be used to determine the corresponding category. • • • Category A: Ts ≤ 25 ms Category B: Ts ≤ 35 ms Category C: Ts ≤ 50 ms 29 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) • Category D: Ts > 50 ms 6.10.3. Handset Receive Delay 6.10.3.1. General The receive delay is defined here as the time from when a digitized, packetized representation of a signal arrives at that VoIP telephone’s packet network interface to the time its analog reproduction is received at an artificial ear sealed to that telephone’s handset. The handset receive delay requirements include the depacketization, hardware/firmware processing and de-jitter delays, plus any delay associated with the radio link for wireless products. 6.10.3.2. Measurement Method A digital audio measuring device capable of measuring the delay between an injected digital packet signal and the output of an artificial ear should be connected to the packet network input of the telephone and to the artificial ear. All delays inherent in the measurement system itself must be calibrated out. The telephone should be set to receive L16-256 packets with a speech frame rate of 20 ms and with one speech frame per packet. A pulsed digital signal of -16 dBm0 shall be injected as packets to the telephone’s network interface. The delay between the time the packet was injected at the telephone network interface to the time it was received at the artificial ear shall be measured. The receive delay shall be used to determine the corresponding category. • • • • Category A: Tr ≤ 30 ms Category B: Tr ≤ 65 ms Category C: Tr ≤ 100 ms Category D: Tr > 100 ms 30 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 7. Headset Technical Requirements The requirements of this section apply to the headset and terminal together. It is not intended to be a specification for a headset as a component separate from the terminal. All tests shall be performed with a Type 3.3 ear simulator. The Type 3.3 ear simulator shall comply with the specifications given in ITU-T Recommendation P.57. The Type 3.3 shall have a hardness of 35 ±6 degrees Shore-OO, as measured according to ASTM 2240. All tests involving the headset receiver shall be done with the same ear and mouth simulator. All test reports shall document the model of ear and mouth simulator used. All tests shall be preformed with the receive volume control set to the nominal volume control setting, unless otherwise specified. The headset test method is given in IEEE Std. 269 and 269a. 7.1. Headset Frequency Response 7.1.1. Headset Send Frequency Response The send frequency response is the overall response of the transducer, send amplifier, and the codec send filter. Send sensitivity is the ratio of the voltage output of the reference codec, or digital bit stream equivalent, to the sound pressure at the Mouth Reference Point (MRP) for each frequency or frequency band (Fi) as shown in the equation below: SMJ = 20 log (VSEND / PM) dB rel 1 V / Pa [3] Where SMJ PM VSEND Send Sensitivity, Mouth to Junction, at Fi. Sound pressure at the MRP at Fi. RMS output voltage of the reference codec, or digital bit stream equivalent at Fi. 7.1.1.1. Measurement Method The test setup is shown in Figure 3, clause 6.1.1.1, except the handset is replaced by the headset. Measurements should be done in ISO 1/12th octave bands or R40 intervals or smaller, over a minimum range of 100 Hz through 8000 Hz using the measurement set-up shown in Figure 3 except the handset is replaced by the headset. Direct digital processing may be employed as explained in clause 4.13. The test signal level shall be −4.7 dBPa at the MRP. . Measurements should be done at the Recommended Test Position (RTP) from the manufacturer. If the RTP is not available, then the headset should be positioned using the guidelines outlined in IEEE Std. 269 and 269a. 7.1.1.2. Requirement The headset send frequency response shall fall between the upper limit and the lower limit given in Error! Reference source not found. and shown in Error! Reference source not found.. The limit curves shall be determined by straight lines joining successive co-ordinates given in the table, when frequency response is plotted on a linear dB scale against frequency on a logarithmic scale. Note that the frequency response mask is a floating or “best fit” mask. 31 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Table 6 – Co-ordinates of Headset Send Response Limits Limit Curve Frequency (Hz) Send Response Limit (dB) [arbitrary level] upper limit 100 120 1000 2000 6000 8000 -1 +4 +4 +9 +9 +8 lower limit 200 200 250 1000 3000 6500 6500 - infinity -6 -3 -3 -1 -10 - infinity Figure 9 – Headset Send Frequency Response Mask 15 Arbitrary Level (dB) 10 5 0 -5 -10 -15 100 1000 Frequency (Hz) 32 10000 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) (Editor; Accepted for now. However, in future handset and headset Tx responses should be harmonized. Needs further investigation. Variability of headset boom mic position is a factor discussed, and reduced noise for boom mics, and some claim less signal in speech justifies high frequency roll off to be more aggressive than previous draft. Also SNR improves if no high frequency component of speech exists, some claim, due to aggressive high frequency rolloff.) 33 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 7.1.2. Headset Receive Frequency Response The receive frequency response is the overall response of the codec receive filter, receive amplifier and transducer. The receive frequency response is the ratio of the sound pressure measured in the ear simulator to the voltage input to the reference codec, or digital bit stream equivalent, for each frequency or frequency band (Fi) as shown in the equation below: SJE = 20 log (PE / VRCV) dB rel 1 Pa / V [4] Where Receive Sensitivity, Junction to Ear, at Fi. SJE ERP Sound pressure measured by ear simulator at Fi. Measurement data are PE converted from the Drum Reference Point, DRP to the ERP.. RMS Input voltage to the reference codec, or digital bit stream equivalent, at Fi. VRCV 7.1.2.1. Measurement Method The receive frequency response is measured according to IEEE Std. 269a using the HATS position and using the measurement set-up shown in Figure 5 of the handset section except the handset is replaced by the headset. Direct digital processing may be employed as explained in clause 4.13. Measurements should be done in 1/12th octave bands, or R40 intervals over a range of 100 Hz through 8000 Hz. Measurements should be done in 1/3rd octave bands for the R40 frequencies of 4000, 5000, 6300, 8000 Hz. The test signal level shall be -18.2 dBV (-16 dBm0), or digital bit stream equivalent. The frequency response measured with the ear simulator must be transformed to the ERP, refer to Annex D. The measurement is made at the reference volume control setting. Note: It is useful to look at 1/12th octave resolution all the way up to 8000 kHz, in order to better understand variability in the receive side, low pass characteristics, etc. 7.1.2.2. Requirement The receive frequency response shall be within the upper limit and lower limits given in 34 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Table 7 and shown in Figure 10. The limit curves shall be determined by straight lines joining successive co-ordinates given in the table, when frequency response is plotted on a linear dB scale against frequency on a logarithmic scale. Note: The frequency response mask is a floating or “best fit” mask. Deleted: ¶ 35 15 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 10 Limit Curve Upper Limit Lower Limit Frequency (Hz) Desirable Receive Response Limit (dB) [arbitrary level] Mandatory Receive Response Limit (dB) [arbitrary level] 100 130 700 1000 1400 2000 4000 8000 200 200 800 4000 6800 6800 +1 +4 +1 +4 +4 +4 0 -5 -10 +9 +9 +8 - infinity -10 -4 -4 -10 -infinity +9 +8 - infinity -10 -4 -4 -10 -infinity 10 5 0 -5 -10 1000 Frequency (Hz) 36 -15 100 Deleted: ¶ Deleted: Limit Curve Deleted: ¶ ¶ 15 Arbitrary Level (dB) 5 +9 Figure 10 – Headset Receive Frequency Response Mask -15 100 Arbitrary Level (dB) Table 7 – Co-ordinates of Headset Receive Response Limits 10000 ... [1] SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 7.2. Headset Wideband Loudness Ratings The loudness rating algorithm is defined in ITU-T Recommendation P.79 and summarized in Annex A. Loudness ratings are calculated from a telephone’s send, receive and sidetone frequency response measurement data. They provide single number metrics, which describe how loud the telephone will sound to a user. An important characteristic of P.79 loudness ratings is that the louder the telephone, the more negative the loudness rating. 7.2.1. Headset Wideband Send Loudness Rating (SLR) The SLR is the loudness loss in the send direction from the acoustic signal at the mouth reference point to the send signal at the digital reference point. Refer to Annex A and ITU-T Recommendation P.79. 7.2.1.1. Measurement Method The SLR shall be calculated using the 1/3rd octave sensitivity data collected from the send frequency response measurement referred to in 7.1.1. Use equation [A1] of Annex A and bands 4 to 17 of Table 14. Note; In Ottawa it was resolved to use NB LR computation for all 920A LR requirements, but corresponding new LR targets were not yet resolved, and the reference to equation [A1] of Annex A has not yet been modified consistently. Regarding targets, options were keeping current LR targets, making SLR 2 dB quieter, making SLR & RLR 2 dB quieter. This global change will obsolete the above reference to bands 1 through 20 here and elsewhere, since bands 1 to 3 (<200 Hz) and bands 18 to 20 (>4000 Hz) are not included in the NB LR computation. 7.2.1.2. Requirement The terminal shall be designed to have an SLR value of 10 dB, with a tolerance of ±5.0 dB. Note; This SLR target is 2 dB quieter relative to a handset, because people typically speak louder into a headset relative to a handset. The tolerance is wider because the mic boom position my vary relative to the MRP. 7.2.2. Headset Wideband Receive Loudness Rating (RLR) The RLR is the loudness loss in the receiving direction from the digital reference point to the ear reference point. Refer to Annex A and ITU-T Recommendation P.79. 7.2.2.1. Measurement Method The RLR shall be calculated using the 1/3rd octave sensitivity data collected from the receive frequency response measurement referred to in 7.1.2. Use equation [A2] of Annex A and bands 4 to 17 of Table 14. The reference volume control setting shall be used. 7.2.2.2. Requirement The monaural terminal shall have an RLR value of 0 dB, with a tolerance of -4.0/+8.0 dB. The binaural terminal should have an RLR value of 6 dB, with a tolerance of -4.0/+8.0 dB, for each of the left and right receivers measured separately. Editor’s note; The old 920 RLR requirement was 2 dB +/-4 dB, but 810-B is 0 dB +/-4 dB. The proposed “shift” to a 0 dB target to harmonize with 810-B seems a bit artificial, given that a 0 dB +4/8 dB requirement is essentially identical to a 2 dB +/-6 dB requirement. NOTE: Either the terminal or the headset should have a receive volume control that is capable of amplification and attenuation. 37 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 7.2.3. Headset Talker Sidetone The sidetone masking rating (STMR) of a digital telephone set is the loudness of the path from the mouth to the ear of the same headset. STMR is calculated from the ratio of the acoustic output signal from the receiver at the ear reference point (ERP) to the acoustic input signal at the mouth reference point (MRP) over the specified frequency band. It’s desirable for the STMR to be constant over the receive volume control range. 7.2.3.1. Measurement Method The test signal level at the MRP shall be -4.7 dBPa. For each frequency given in Table 14, bands 1 to 20, the sound pressure in the ear simulator shall be measured. The frequency response measured with the ear simulator must be transformed to the ear reference point (ERP). Refer to Annex D. The STMR shall be calculated using equation [A3] of Annex A. Telephone sets with adjustable receive levels shall be tested at the minimum, nominal and maximum volume control settings. 7.2.3.2. Requirement For any adjustable receive level, the value of STMR shall be within the range of 21 dB ±6 dB for supra-aural, 18 dB ±6 dB for insert, 20 dB ±6 for intra-conch, e.g. earbud.. The value of STMR for binaural terminals should be 6 dB quieter, for each of the receivers measured separately. NOTE - In practice, sidetone measurements in the high leak position are limited to a value of approximately 24 dB by the influence of the test setup (HATS). 7.2.4. Headset Sidetone Delay In a digital telephone, sidetone echo occurs when significant delay is introduced into the speech path between the headset microphone and the headset receiver by the sidetone feedback algorithm. Ideally, the sidetone signal should be a real-time signal. Sidetone delay less than 5 ms is generally perceived as normal sidetone. Sidetone delay between 5 and 10 ms is generally perceived as unnatural sidetone, with an uncomfortable hollow characteristic. Sidetone delay greater than 10 ms is generally perceived as a distinct talker echo signal. Since the sidetone level could be as loud, or louder than a talker echo signal, sidetone delay greater than 5 ms is undesirable. 7.2.4.1. Measurement Method See the method described in IEEE Standards 269, 269 and 269a. 7.2.4.2. Requirement Sidetone delay shall be less than 5 ms. Sidetone delay should be less than 1 ms. 7.3. Headset Noise 7.3.1. Headset Send Noise 7.3.1.1. General The send noise of a digital telephone is the 5 second average background noise level at the digital transmit output with the telephone headset transmitter isolated from sound input and mechanical disturbances. 38 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 7.3.1.2. Measurement Method In a quiet environment (ambient noise less than 30 dBA), free of mechanical disturbances, measure the A-weighted, 5 second average, noise level at the digital interface output or the reference codec decoder output over the frequency range of 100 to 8000 Hz. 7.3.1.3. Requirement The overall send noise shall be less than or equal to -64 dBm0, A-weighted. (This agrees with the 810B headset requirement, and is 4 dB louder than the 810B handset requirement, ignoring the discrepancy in spectral weighting. Typically the headset mic boom may not be as close to the MRP as the handset.) 7.3.2. Headset Send Single Frequency Interference 7.3.2.1. General Narrow-band noise, including single frequency interference, is an impairment that can be perceived as a tone depending on its level relative to the overall weighted noise level. 7.3.2.2. Measurement Method In a quiet environment (ambient noise less than 30 dBA), free of mechanical disturbances, measure the A-weighted noise level at VSEND with a selective voltmeter or spectrum analyzer having an effective bandwidth of not more than 31 Hz, over the frequency range of 100 to 8000 Hz. If FFT analysis is used, then “Flat Top” windowing shall be employed. 7.3.2.3. Requirement The A-weighted send single frequency interference shall be no greater than -74 dBm0. 7.3.3. Headset Receive Noise 7.3.3.1. General The receive noise of a digital telephone is the 5 second average noise level measured at the output of the telephone receiver with the digital telephone receiving the digital quiet code. Receive noise measurement results must be transformed from the DRP of the ear simulator to the ERP. If a single wideband measurement is made, the transfer function must be realized using a minimum phase, parametric filter (or equivalent). Refer to IEEE Std. 269a. 7.3.3.2. Measurement Method A signal corresponding to a decoder quiet code is applied at the digital interface. The A-weighted noise level is measured in the ear simulator over the frequency range of 100 to 8500 Hz. The ambient noise for this measurement shall not exceed 30 dBA. 7.3.3.3. Requirement The receive noise shall be less than 40 dBA for a monaural headset. The receive noise for binaural headsets should be less than 34 dBA, for each of the receivers measured separately. 39 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 7.3.4. Headset Receive Single Frequency Interference 7.3.4.1. General Narrow-band noise, including single frequency interference, is an impairment that can be perceived as a tone depending on its level relative to the overall weighted noise level. This test measures the weighted noise level characteristics in narrow bands of not more than 31 Hz, which can then be compared to the overall weighted background noise level. Narrow-band noise is measured at the output of the telephone receiver with the digital telephone receiving the digital quiet code. 7.3.4.2. Measurement Method A signal corresponding to a decoder quiet code is applied at the digital interface. The A-weighted noise level is measured in the ear simulator with a selective voltmeter or spectrum analyzer with an effective bandwidth of not more then 31 Hz, over the frequency range of 100 to 8500 Hz. If FFT analysis is used, then “Flat Top” windowing shall be employed. The ambient room noise for this measurement shall not exceed 30 dBA. 7.3.4.3. Requirement The receive A-weighted single frequency interference shall be 10 dB quieter than the overall Aweighted receive noise and shall be below 30 dBA. Editors Note; A very quite phone could in principle have , for example, 10 dBA overall noise but 1 dBA noise within 180 +/-15.5 Hz and thus fail this receive noise requirement. The committee felt that room and measurement mic noise are present at higher levels both overall and from 100 to 8500 Hz, so this type of theoretical unmerited failure does not occur in practice.. 7.4. Headset Distortion and Noise The distortion and noise requirements only apply to linear 16 bit PCM at 256 kbit/s. 7.4.1. Headset Send Distortion and Noise 7.4.1.1. Method of Measurement The highest signal levels employed for this test may exceed the published specifications for the mouth simulator. At high test levels, short duty cycles may be required to prevent overheating of the mouth simulator. Prior to testing the telephone, the output level and distortion of the mouth simulator should be verified at the maximum sound pressure level used for each frequency. This need not be verified before each test. The distortion of the mouth simulator should be at least 10 dB less than the maximum allowable telephone distortion for each frequency and level used for test. The mouth verification should be done at the MRP. Apply a sine wave signal at the MRP, at the levels given in Table 8 and at the following frequencies: 160, 315, 502,803, 1004, 2008 and 3150 Hz. The ratio of the signal-to-total-distortion and noise power of the digital encoded signal output is measured. The test frequency tolerance is 3%, but even submultiples of the sampling frequency must not be used. Note: In cases where the sound pressure exceeds +6 dBPa, the linearity of the artificial mouth should be checked, as it exceeds the limits of ITU-T Recommendation P.51. 7.4.1.2. Requirement The ratio of the signal-to-total distortion and noise (SDN) of the digitally encoded signal output shall be above the limits given in Table 8, when the A-weighting is applied to the measured distortion and noise output. Limits for intermediate levels are found by drawing straight lines between the successive coordinates in the table on a linear (dB signal level) – linear (dB ratio) scale. 40 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Table 8 – Headset Send Signal-to-Total Distortion and Noise Ratio Limits -30 Send Ratio 160 to 803 Hz (dB) 26 Send Ratio 1004 Hz (dB) 26 Send Ratio 2008 to 3150 Hz (dB) NA -25 26 26 26 -20 31 31 31 -10 33 33 33 Send Level at MRP (dB Pa) 0 33 33 33 +5 33 33 33 +10 26 26 26 +15 NA 20 NA Note: 20 dB = 10%, 26 = 5%, 31 = 2.8%, 33 = 2.2%. NA = Not Applicable. Note: Most sound source equipment will generate significant distortion for acoustic signals above 5 dBPa. Editor’s Note: Note 160 Hz is below the lower mask. Work is in progress on a major revision of the SDN measurement procedure, using gated 1/3 octave band limited white noise stimulus. 7.4.2. Headset Receive Distortion and Noise 7.4.2.1. Method of Measurement Apply a digitally simulated sine wave, with the signal levels given in Table 9 and the following frequencies: 160, 315, 502, 1004, 2008 and 3150 Hz. The ratio of the signal-to-total distortion and noise power is measured in the ear simulator. The test frequency tolerance is 3%, and even submultiples of the sampling frequency must not be used. 7.4.2.2. Requirement The ratio of the signal-to-total distortion and noise (SDN) measured in the ear simulator, shall be above the limits given in Table 9, with A-weighting applied to measured distortion and noise output, unless the signal in the ear simulator exceeds +10 dBPa or is less than -50 dBPa. Table 9 – Headset Receive Signal-to-Total Distortion and Noise Ratio Limits Receive level at the digital interface (dBm0) -40 Receive Ratio @ 160 Hz (dB) 20 Receive Ratio @ 315 Hz (dB) 24 Receive Ratio @ 502 to 3150 Hz (dB) 24 -34 24 24 24 -27 28 30 30 -20 28 32 32 -10 28 32 32 -6 28 32 32 -3 28 28 28 0 24 24 Note: 20 dB = 10%, 24 = 6.3%, 28 = 4%, 30 = 3.2%, 32 = 2.5%. 41 28 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Editor’s Note: We must remove 160 Hz during the future major rework of this requirement, because the Rx lower mask only goes to 200 Hz, and high pass filtering is encouraged below this frequency. 7.5. Weighted Terminal Coupling Loss (TCLw) The weighted terminal coupling loss (TCLw) provides a measure of the echo performance under normal conversation, i.e., single far-end talker conditions. It is possible that echo control devices such as echo suppressors or echo cancellers with non-linear processing may be used on headset connections to provide sufficient echo return loss to mitigate increased echo associated with longer network delays. The use of echo control devices on the headset can affect the measurement of TCLw. The result world likely be different under cases of either single far-end talker or double-talk. The TCLw measurement is intended to represent a single far-end talker. The ‘proper’ measurement of TCLw is addressed in IEEE std. 269, Annex O. 7.5.1. Measurement Method The TCLw measurement shall be made at an input signal level of -16 and -10 dBm0. The test shall be performed in a quiet environment (the ambient noise level shall be less than 30 dBA). The TCLw measurement shall not be performed using a sinusoidal test signal for the receive path input. The test signal may be a composite source signal (CSS) as defined in ITU-T P.501 or bursted white noise. The test signal shall be band-limited to 100 through 8000 Hz. The calibration shall be determined during the ON portions of the test signal, not the average of on and off times. The measurement shall be performed after system stability is reached (including convergence of any echo algorithms): this shall be accomplished by invoking the test signal for at least 2 seconds before the actual measurement occurs. The attenuation from digital input (receive) to digital output (send) is measured at 1/12th octave bands, using the measurement arrangement shown in Figure 11. See Annex C. The weighted terminal coupling loss is calculated according to ITU-T G.122 Annex B.4 (trapezoidal rule) using the frequency range of 300 to 6700 Hz rather than 300 to 3400 Hz. Figure 11 – Terminal Coupling Loss Measurement Method Headset Suspended vSEND (Echo Return) v Decoder Digital Set Interface Coder GEN vRCV Anechoic Chamber Reference Codec 42 Deleted: This may provide idealized and unrealistic performance measurements when non-linear processing on the transmit side is used as linear processing disabled or with a near-end signal present that is a) capable of enabling echo control’s double-talk detector with the subsequent removal of non-linear processing and b) can be filtered out from the final return signal so as not to affect the accuracy of the TCLw measurement. The latter may be the only method that can be used consistently across products in a black-box testing setup. A suitable signal may be a pulsed sine wave, but it will depend on the temporal characteristics of the double-talk detector.¶ ¶ Deleted: then becomes specific to the echo control implementation. These issues are still under study and are not addressed in these requirements. For further information see SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 7.5.2. Requirements The normalized value of TCLw loss shall be greater than 52 dB for IP sets and 45 dB for PCM sets when measured under free field conditions and with SLR normalized to 10 dB and RLR normalized to 0 dB. Note that the normalized value of TCLw for IP sets to be greater than 50 dB to meet ITU-T recommendation G.131 talker echo objective requirements. Editor’s Note: This needs further investigation. G.131 references P.310 (45 dB) and P.311 (35 dB) and ETSI (55 dB?). Perhaps this sentence should be removed for handset and headset entirely because the 50 dB reference cannot be found. For example, if the measured TCLw is 48 dB, the measured SLR is 11 dB and the measured RLR is 2 dB, then the normalized value of TCLw = 48 dB + (10 - 11) dB + (0 – (-2)) dB = 49 dB. See clause 7.6 for additional notes regarding TCLw. NOTES 1. The requirement of 52 dB for IP sets is a function of the -16 dBm0 test signal level and the -68 dBm0A send noise requirement. Measuring TCLw > 52 dB can be difficult. 2. If equipped with adjustable receive level, the un-normalized TCLw will decrease in proportion with the increased gain relative to the nominal RLR in most cases. For example, if the measured TCLw is 45 dB at nominal RLR and the adjustable receive level adds 12 dB of gain, then un-normalized TCLw (maximum receive level) = 45 dB - 12 dB = 33 dB. 3. The echo impairment perceived by the person at the far-end of the connection from a telephone set is a function of the magnitude of the talker echo signal as well as the talker echo path delay. The echo signal becomes more disturbing as the talker echo path delay increases. Thus, a telephone set with adequate TCLw performance on low delay connections may provide satisfactory performance while the same may not be true for connections that have a long delay. 4. Temporally weighted terminal coupling loss (TCLt) is an alternate method for echo measurement, which may be more subjectively relevant, especially in devices with echo suppression or cancellation features. (See IEEE Std 1329.) The performance requirements may need to be changed when using this method. 7.6. Stability Loss The stability loss is a measure of the contribution of the telephone set to the overall network stability requirements. Stability loss is defined as the minimum loss from the digital input (receive) to the digital output (send), at any frequency. 7.6.1. Measurement Method The stability measurement shall be made at input signal levels of -16 and -10 dBm0. The TCLw measurement shall not be performed using a sinusoidal test signal for the receive path input. The test signal may be CSS or bursted white noise, band-limited to 100 through 8000 Hz and represented by the L16-256 codec. The measurement and calibration shall be determined during the ON portions of the signal, not the average of on and off times. With the headset and transmission circuit fully active, measure the attenuation from the digital input to the digital output using Method 1 and Method 2. 43 Deleted: test SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 7.6.1.1. Method 1 Place the headset in the reference corner, as shown in Figure 12. Test stability while placing headset on both sides (face down and face up) with mic boom closest to corner. If there is a retractable mic boom, extend mic to normal use position for HATS. The headset shall be placed along the diagonal from the apex of the reference corner to the outside corner, with the receiver end of the headset 250 mm from the apex. The telephone set shall be fully active. The reference corner consists of three perpendicular planes, smooth, hard surfaces extending 0.5 m from the apex of the corner. 7.6.1.2. Method 2 Place the headset with the receiver and mouthpiece facing a hard, smooth surface free of any other object for 50 cm. The telephone set shall be fully active. Figure 12 – Reference Corner 25 cm 50 cm Editor’s note: needs to change Figure 12 to show headset. 7.6.2. Requirement Stability loss using both method 1 and method 2, (i.e., minimum loss, at any frequency) shall be greater than 6 dB. It is desirable that this loss be greater than 10 dB. Telephone sets with adjustable receive level should maintain stability over the entire range of adjustable receive levels. 7.7. Headset Long Duration Maximum Acoustic Pressure (Steady State) 7.7.1. General The long duration maximum acoustic pressure is the steady state (longer than 500 ms) sound pressure disturbance emitted from a telephone receiver, caused by the maximum excursions of the receive digital signal. 44 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Additional consideration should be given to the acoustic pressure caused by tones, other audio signals or long duration, high amplitude electrical signals applied to power, network, headset or auxiliary leads of the digital telephone. 7.7.2. Measurement Method The steady-state A-weighted sound pressure level shall be measured using the digital terminals test procedure in IEEE 269. 7.7.3. Requirements The measured maximum rms level shall be less than 118 dB(A) at ERP, required in UL/CSA 609501, 2007. 7.8. Short Duration Maximum Acoustic Pressure (Peak) 7.8.1. General The short duration maximum acoustic pressure is the sound pressure impulse (less than 500 ms) emitted from a headset receiver. This short duration test stresses nonlinear processes, like AGC, and doesn’t directly replace a short duration surge. Additional consideration should be given to the peak acoustic pressure caused by tones or short duration, high amplitude electrical pulses applied to power, network, headset or auxiliary leads of the digital telephone. 7.8.2. Measurement Method The peak acoustic pressure level shall be measured using the digital terminals test procedure in IEEE 269.. 7.8.3. Requirements The maximum peak acoustic pressure shall be less than 136 dBSPL, as required in UL/CSA 60950-1, 2007. Editor’s Note: Some committee members felt UL should be referenced and no safety requirements should be in this document. Al Baum will check out the TIA legal policy regarding liability on this issue. Editor’s note: The Clause 6.20 of VoIP about telephone delay does not belong to handset. It should be a standalone clause including both handset and headset. In general handset and headset do not affect the overall VoIP telephone delay very much, and handsfree may have only a few msecs more delay for noise filtering, etc.) 45 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 8. Speakerphone Technical Requirements (Advisory) All speakerphone requirements are advisory. Speakerphone test methods are given in IEEE Std. 1329. Speakerphone modes of operation almost always employ algorithms that are sensitive to the temporal and spectral characteristics of the signals. For these devices the use of sinusoidal test signals is not recommended. Speakerphone telephones designed for other than traditional tabletop or desktop positioning should be tested with the appropriate user positioning in mind. This position shall be defined as the “recommended test position” (RTP). The RTP should be obtained from the manufacturer, and should be based upon the product’s intended use. For testing purposes, this will dictate the distance and position geometry relationship between the speakerphone telephone and the mouth simulator and microphone. Measurements performed at other distances or positions shall be noted, and in the absence of an RTP, the 50 cm test position as defined in IEEE Std. 1329 is recommended. 8.1. Speakerphone Frequency Response 8.1.1. Speakerphone Send Frequency Response The send frequency response is the overall response of the transducer, send amplifier, and the codec send filter. The send sensitivity is expressed in terms of dBV/Pa. 8.1.1.1. Measurement Method The send frequency response is measured directly in 1/3rd octave bands or converted to 1/3rd octave bands according to IEEE Std. 1329, clause 9.3.8. The measurement is made over a minimum range of 100 Hz through 8000 Hz. For desktop speakerphones, use the measurement set-up shown in Figure 13. For other speakerphone devices, use the RTP. The test signal level shall be -4.7 dBPa at the MRP. Figure 13 – Speakerphone Send Frequency Response Measurement Method vSEND Mouth Simulator Send GEN Decoder pM Interface 50 cm 30 cm Coder 40 cm Anechoic Chamber Digital Set (On Table) Reference Codec 46 v Measuring Amplifier SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 8.1.1.2. Requirement The speakerphone send response should be within the limits given in Table 10 and shown in Figure 14. The limits break at the crossover frequencies of the ISO R10 series of preferred frequencies. The crossover frequencies may be adjusted up to 3% to accommodate non-R10 1/3rd octave measurement data. The frequency response mask is a floating or “best fit” mask. Table 10 – Co-ordinates of Speakerphone Send Response Limits Limit Curve Frequency Bands Send Response Limit (dB) [arbitrary level] Upper Limit 100 to 1120 1120 to 1780 1780 to 2820 2820 to 4470 4470 to 7080 7080 to 8910 0 +1 +2 +3 +4 -5 Lower Limit 224 to 282 282 to 355 355 to 4470 4470 to 5620 -12 -10 -8 -12 Figure 14 – Speakerphone Send Frequency Response Mask 10 Arbitrary Level (dB) 5 0 -5 -10 -15 -20 100 1000 Frequency (Hz) 47 10000 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 8.1.2. Speakerphone Receive Frequency Response The receive frequency response is the overall response of the codec receive filter, receive amplifier and transducer. The receive sensitivity is expressed in terms of dBPa/V. 8.1.2.1. Measurement Method The receive frequency response is measured directly in 1/3rd octave bands or converted to 1/3rd octave bands according to IEEE Std. 1329 clause 9.3.8. The measurement is made over a minimum range of 100 Hz through 8000 Hz. For desktop speakerphones, use the measurement set-up shown in Figure 15. For other speakerphone devices, use the RTP. The test signal level shall be -25 dBV. The speakerphone reference volume control setting shall be used. A free field microphone is used for the measurement. Figure 15 – Speakerphone Receive Frequency Response Measurement Method To Sound Pressure Measuring Amplifier Decoder Free Field Microphone Receive pE Interface 50 cm 30 cm 40 cm Anechoic Chamber GEN Coder Digital Set (On Table) vRCV Reference Codec 8.1.2.2. Requirement The speakerphone receive frequency response should be below the 1/3rd octave band upper limit and above the 1/3rd octave band lower limit given in Table 11 and shown in Figure 16. The limits break at the crossover frequencies of the ISO R10 series of preferred frequencies. The crossover frequencies may be adjusted up to 3% to accommodate non-R10 1/3rd octave measurement data. Table 11 – Co-ordinates of Speakerphone Receive Response Limit Curves Limit Curve Frequency Receive Response Limit (dB) [arbitrary level] upper limit 100 to 7080 7080 to 8910 0 -5 lower limit 224 to 282 282 to 4470 4470 to 5620 -14 -12 -14 48 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Figure 16 – Speakerphone Receive Frequency Response Mask 10 Arbitrary Level (dB) 5 0 -5 -10 -15 -20 100 1000 Frequency (Hz) 10000 8.2. Speakerphone Wideband Loudness Ratings and Receive Volume Control Correlation factors relating speakerphone loudness ratings to handset loudness ratings are under investigation and are not used in this standard. The currently accepted correlation factors for personal, wireline telephone applications are described in sub clause 8.2.2.2. These correlation factors may not be appropriate for other speakerphone applications such as conference, hand-held, in-car applications or any applications where the relationship between the talker and the speakerphone varies from the 50 cm position, or the reverberation characteristics or the background noise levels vary from typical office environments. The loudness rating algorithm is defined in ITU-T Recommendation P.79 and summarized in Annex A. Loudness ratings are calculated from a telephone’s send, receive and sidetone frequency response measurement data. They provide single number metrics, which describe how loud the telephone will sound to a user. An important characteristic of P.79 loudness ratings is that the louder the telephone, the more negative the loudness rating. 8.2.1. Speakerphone Wideband Send Loudness Rating (SLR) The SLR is the loudness loss in the send direction from the acoustic signal at the mouth reference point to the send signal at the digital reference point. Refer to Annex A and ITU-T Recommendation P.79. 8.2.1.1. Measurement Method The SLR shall be calculated using the 1/3rd octave sensitivity data collected from the send frequency response measurement. Use equation [A4] of Annex A and bands 1 to 20, Table 14. 49 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 8.2.1.2. Requirement The terminal should be designed to have a speakerphone SLR value of 13 dB, with a tolerance of ±4.0 dB. 8.2.2. Speakerphone Wideband Receive Loudness Rating (RLR) 8.2.2.1. Measurement Method The RLR shall be calculated from the 1/3rd octave sensitivity data collected from the receive frequency response measurement. Use equation [A5] of Annex A and bands 1 to 20, Table 14. 8.2.2.2. Requirement At the reference volume control setting the terminal should be designed to have a speakerphone RLR of 2 dB, with a tolerance of ±4.0 dB. Note that in most standards, such as IEEE Std. 1329 and ITU-T Recommendation P.340, there is a correction factor of 14 dB subtracted from calculated RLR (RLR = RLR -14). In order to be consistent with TIA-810-B, this standard does not use the correction factor. In order to compare RLR as measured by other standards the correction factor may be used. For example, an RLR of 16 dB measured per this standard would be: 16 dB -14 dB = +2 dB. If a correction factor is used, it must be stated with the data. 8.2.3. Speakerphone Receive Volume Control The speakerphone receive volume control shall provide greater than or equal to 8 dB of gain relative to the reference volume control setting. The volume control should provide at least 16 dB of attenuation relative to the reference volume control setting. The volume control step size shall be less than 6 dB. 8.3. Speakerphone Noise 8.3.1. Speakerphone Send Noise 8.3.1.1. General The send noise of a digital telephone is the 5 second average noise level at the digital transmit output with the microphone isolated from sound input and mechanical disturbances. 8.3.1.2. Measurement Method Speakerphone send noise is measured according to IEEE Std. 1329, clause 9.3.4. With the speakerphone telephone in a quiet environment (ambient noise less than 30 dBA), the A-weighted noise level at the digital output is measured. 8.3.1.3. Requirement The speakerphone send noise should be no greater than -63 dBm0, A-weighted. 8.3.2. Speakerphone Send Single Frequency Interference 8.3.2.1. General Narrow-band noise, including single frequency interference, is an impairment that can be perceived as a tone depending on its level relative to the overall weighted noise level. This test measures the weighted noise level characteristics in narrow bands of not more than 31 Hz, which can then be compared to the overall weighted background noise level. 50 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 8.3.2.2. Measurement Method Speakerphone send noise is measured according to IEEE Std. 1329, clause 9.3.4. In a quiet environment (ambient noise less than 30 dBA), measure the noise level at VSEND with a selective voltmeter or spectrum analyzer having an effective bandwidth of not more than 31 Hz, over the frequency range of 100 to 8000 Hz. If FFT analysis is used, then “Flat Top” windowing shall be employed. 8.3.2.3. Requirement The speakerphone send single frequency interference should be less than -70 dBm0. 8.3.3. Speakerphone Receive Noise 8.3.3.1. General The receive noise of a digital speakerphone telephone is the 5 second average noise level measured at the speaker output with the digital telephone receiving the digital quiet code. 8.3.3.2. Measurement Method The speakerphone receive A-weighted noise level is measured according to IEEE Std. 1329 clause 9.4.4. A signal corresponding to the quiet code is applied at the digital interface. The ambient noise for this measurement shall not exceed 30 dBA. 8.3.3.3. Requirement The speakerphone receive noise should be less than 40 dBA at the maximum volume control setting and less than 35 dBA at the reference volume control setting, with the comfort noise turned off. 8.3.4. Speakerphone Receive Single Frequency Interference 8.3.4.1. General Narrow-band noise, including single frequency interference, is an impairment that can be perceived as a tone depending on its level relative to the overall weighted noise level. This test measures the weighted noise level characteristics in narrow bands of not more than 31 Hz, which can then be compared to the overall weighted background noise level. Narrow-band noise is measured at the output of the telephone receiver with the digital telephone receiving the digital quiet code. 8.3.4.2. Measurement Method Speakerphone receive noise is measured according to IEEE Std. 1329, clause 9.4.4. A signal corresponding to a decoder quiet code is applied at the digital interface. The A-weighted noise level is measured with a selective voltmeter or spectrum analyzer having an effective bandwidth of not more then 31 Hz, over the frequency range of 100 to 8000 Hz. If FFT analysis is used, then “Flat Top” windowing shall be employed. The ambient noise for this measurement shall not exceed 30 dBA. The reference volume control setting shall be used. 8.3.4.3. Requirement The receive A-weighted single frequency interference shall be 10 dB quieter than the overall Aweighted receive noise. In no case shall the single frequency interference be greater than 28 dBA in each measured band. 8.4. Speakerphone Distortion and Noise The distortion and noise requirements only apply to linear PCM at 256 kbits/s. 51 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 8.4.1. Speakerphone Send Distortion and Noise 8.4.1.1. Method of Measurement Speakerphone send distortion and noise is measured according to IEEE Std. 1329, clause 9.3.6. Apply a test signal at the MRP, at the levels given in Table 12 and at the following frequencies: 315, 502, 803, 1004, 2008 and 3150 Hz. The ratio of the signal power –to the total A-weighted distortion and noise power of the digitally encoded signal output is measured. Table 12 – Speakerphone Send Signal-to-Total Distortion and Noise Ratio Limits Send level at the MRP (dBPa) Send Ratio (dB) @ 315 Hz Send Ratio (dB) @ 502 to 3150 Hz -10 -5 0 +5 +10 26 30 30 30 30 26 30 30 30 30 Note: 26 dB = 5%, 30 dB = 3.2% 8.4.1.2. Requirement The ratio of the signal power to the total distortion and noise power of the digitally encoded signal output should be above the limits given in Table 12. Limits for intermediate levels are found by drawing straight lines between the successive coordinates in the table on a linear (dB signal level) – linear (dB ratio) scale. 8.4.2. Speakerphone Receive Distortion and Noise 8.4.2.1. Method of Measurement Speakerphone receive distortion and noise is measured according to IEEE Std. 1329, 9.4.6. Apply a test signal, at the levels given in Table 13 and at the following frequencies: 315, 502, 1004, 2008 and 3150 Hz. The ratio of the signal power to the total A-weighted distortion and noise power is measured. The reference volume control setting shall be used. 8.4.2.2. Requirement The ratio of the signal power to the total A-weighted distortion and noise power should be above the limits given in Table 13, unless the measured sound pressure is less than -50 dBPa. The measurement microphone may be placed at 25 cm for this measurement if the measured signal levels are too low. 52 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Table 13 – Speakerphone Receive Signal-to-Total Distortion and Noise Ratio Limits Receive level at the digital interface (dBm0) Receive Ratio (dB) @ 315 & 502 Hz Receive Ratio (dB) @ 803 to 2008 Hz -30 -20 -10 -6 -3 28 28 28 28 24 28 30 30 30 24 Note: 24 dB = 6.3%, 28 dB = 4%, 30 dB =3.2% 53 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) 8.5. Weighted Terminal Coupling Loss (TCLw) The weighted terminal coupling loss (TCLw) provides a measure of the echo performance under normal conversation, i.e., single far-end talker conditions. It is possible that echo control devices such as echo suppressors or echo cancellers with non-linear processing may be used on speakerphone connections to provide sufficient echo return loss to mitigate increased echo associated with longer network delays. The use of echo control devices on the speakerphone can affect the measurement of TCLw. The result would likely be different under cases of either single far-end talker or double-talk. The TCLw measurement is intended to represent a single far-end talker. The ‘proper’ measurement of TCLw is addressed in IEEE Std. 269, Annex O. Deleted: . 8.5.1. Measurement Method Refer to IEEE Std. 1329, clause 11, for the test method. The test range is 300 to 6700 Hz. The TCLw measurement shall not be performed using a sinusoidal test signal for the receive path input. 8.5.2. Requirements The normalized value of speakerphone TCLw loss should be greater than 45 dB when measured under free field conditions and with SLR normalized to 13 dB and RLR normalized to 2 dB. It is desirable that the normalized value of TCLw be greater than 50 dB to meet ITU-T Recommendation G.131 talker echo objective requirements. For example, if the measured speakerphone TCLw is 27 dB, the measured SLR is 15 dB and the measured RLR is -1 dB, then the normalized value of TCLw = 27 dB + (2-(-1)) dB + (13 - 15) dB = 28 dB. 8.6. Stability Loss The stability loss is a measure of the telephone set's contribution to the overall network stability requirements. Stability loss is defined as the minimum loss from the digital input (receive) to the digital output (send), at any frequency. 8.6.1. Measurement Method Place the speakerphone telephone in the middle of a hard, smooth surface free of any other object for 0.5 m. The telephone set shall be fully active. The surface must be at least 1 square meter, with no horizontal dimension of the table less than 0.8 m (see IEEE Std. 1329, clause 7.3.4). The stability measurement is made at an input signal level greater than or equal to -10 dBm0 and less than or equal to 0 dBm0, at 1/12th octave bands, or R40 intervals centered at 300 Hz to 6700 Hz. The TCLw measurement shall not be performed using a sinusoidal test signal for the receive path input. The test signal may be CSS or bursted white noise, band-limited to 100 through 8000 Hz and represented by the L16-256 codec. The measurement and calibration shall be determined during the ON portions of the signal, not the average of on and off times. With the speakerphone and transmission circuit fully active, measure the attenuation from the digital input to the digital output. 8.6.2. Requirement The speakerphone stability loss, i.e., minimum loss, at any frequency should be greater than 6 dB. It is desirable that this loss be greater than 10 dB. Telephone sets with adjustable receive level should maintain a minimum 6 dB stability loss over the entire range of adjustable receive levels. 54 Deleted: test SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Annex A (normative) – Calculation of Loudness Ratings This Annex details the loudness rating calculations and weighting factors relevant to this document. Loudness ratings are a measure of loudness loss and are used in network planning to insure that the loudness of a connection from the Mouth Reference Point (MRP) of the talker to the Ear Reference Point (ERP) of the far end listener is at a satisfactory level. The loudness of the complete path is designated as the wideband Overall Loudness Rating (OLR). The wideband Send Loudness Rating (SLR) is the loudness loss from the MRP to the electrical output. The wideband Receive Loudness Rating (RLR) is the loudness loss from the electrical input to the ERP. The Sidetone Masking Rating (STMR) is the loudness loss from the MRP to the ERP via the electric sidetone path. Loudness ratings are used rather than simple level measurements because of better subjective correlation. Loudness ratings more closely account for the ear’s different sensitivity at different frequencies and its nonlinear response to varying sound levels. The following calculations are based on the ITU-T Recommendation P.79. ITU-T Recommendation P.79 provides information on the derivation of the loudness rating algorithm. Loudness ratings determined in accordance with P.79 are analogous to loss, resulting in the characteristic that the louder the telephone, the more negative the loudness rating. ITU-T Recommendation P.79 Annex G has the weighting factors used for wideband SLR and RLR calculation. The STMR weighting factors are unchanged. For convenience, the P.79 weighting information is included in this document in Table 14. Wideband Send Loudness Rating (Handset and Headset): Band 20 SLR = - 57.1 log10 ∑ 10 (0.1 * 0.175 * (SMJ – Wsi )) [A1] i = Band 1 Where: i SMJ Wsi Frequency bands from Table 14, bands 1-20. Send frequency response data (Sensitivity, Mouth-to-Junction) in dBV/Pa measured per this standard. Send weighting factor from Table 14. Wideband Receive Loudness Rating (Handset and Headset): Band 20 RLR = - 57.1 log10 ∑ 10 (0.1 * 0.175 * (SJE – Wri )) [A2] i = Band 1 Where: i SJE Wri Frequency bands from Table 14, bands 1-20. Receive frequency response data (Sensitivity, Junction-to-Ear Reference Point) in dBPa/V measured per this standard. See Annex D for DRP to ERP information. Receive weighting factor from Table 14. Note: There is no Leakage Correction, LE, when using the Type 2, 3.2, 3.3 or Type 3.4 ear simulator. 55 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Sidetone Masking Rating (Handset and Headset): Band 20 STMR = - 44.4 log10 ∑ 10 (0.1 * 0.225 * (SmeST – WMSi)) [A3] i = Band 1 Where: i SmeST WMSi Frequency bands from Table 14, bands 1-20. Sidetone frequency response data (Sensitivity, Mouth-to-Ear) in dB Pa/Pa measured per this standard. Sidetone weighting factor from Table 14. Wideband Send Loudness Rating (Speakerphone): Band 20 SLR = - 57.1 log10 ∑ 10 (0.1 * 0.175 * (SMJ – Wsi )) [A4] i = Band 1 Where: i SMJ Wsi Frequency bands from Table 14, bands 1-20. Send frequency response data (Sensitivity, Mouth-to-Junction) in dBV/Pa measured per this standard. Send weighting factor from Table 14. Wideband Receive Loudness Rating (Speakerphone): Band 20 RLR = - 57.1 log10 ∑ 10 (0.1 * 0.175 * (SJE – Wri )) - CorrRFF [A5] i = Band 1 Where: Frequency bands from Table 14, bands 1-20. Receive frequency response data (Sensitivity, Junction-to-Ear) in dBPa/V measured per this standard. Receive weighting factor from Table 14. Wri CorrRFF Correction of 14 dB for receive measured in the free field as recommended in ITU-T P.340 (05/2000) i SJE Note that in most standards, such as IEEE Std. 1329 and ITU-T Recommendation P.340, there is a correction factor of 14 dB subtracted from calculated RLR (RLR = RLR-14). In order to be consistent with TIA-810-A, this standard does not require the correction factor. In order to compare RLR as measured by other standards the correction factor may be used. For example, an RLR of 16 calculated using equation [A5] would be 8 dB-14 dB = 2 dB. If a correction factor is used, it must be stated with the data. 56 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Table 14 – ITU-T P.79 Annex G Weighting factors for calculating wideband loudness ratings Band No. Midfrequency (Hz) Send Wsi Receive Wri Sidetone WMSi 1 100 154.5 152.8 110.4 Deleted: 110.4 2 125 115.4 116.2 107.7 Deleted: 107.7 3 160 89.0 91.3 104.6 Deleted: 104.6 4 200 77.2 85.3 98.4 Deleted: 98.4 5 250 62.9 75.0 94.0 Deleted: 94.0 6 315 62.3 79.3 89.8 Deleted: 89.8 7 400 45.0 64.0 84.8 Deleted: 84.8 8 500 53.4 73.8 75.5 Deleted: 75.5 9 630 48.8 69.4 66.0 Deleted: 66.0 10 800 47.9 68.3 57.1 Deleted: 57.1 11 1000 50.4 69.0 49.1 Deleted: 49.1 12 1250 59.4 75.4 50.6 Deleted: 50.6 13 1600 57.0 70.7 51.0 Deleted: 51.0 14 2000 72.5 81.7 51.9 Deleted: 51.9 15 2500 72.9 76.8 51.3 Deleted: 51.3 16 3150 89.5 93.6 50.6 Deleted: 50.6 17 4000 117.3 114.1 51.0 Deleted: 51.0 18 5000 157.3 144.6 49.7 Deleted: 49.7 19 6300 172.2 165.8 50.0 Deleted: 50.0 20 8000 181.7 166.7 52.8 Deleted: 52.8 Note: The send and receive weighting values for bands 4-17 are different from the values used for narrowband (200 to 4000 Hz) calculation and therefore should not be used for narrowband loudness calculation. 57 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Annex B (informative) – Measurement and Level Conversions General The following describes how to convert between various units of measurement used in telephone testing. Useful Conversions and Procedures 0 dBm (0 VU) is accepted as 1 mW, typically using a circuit impedance of 600 Ω or 900 Ω. 0 dBm = 10 log 1(mW) dBV = 10 log V2 = 20 log V or, V = 10 dBV/20 P = V2/R, where for dBm reference, R = 600 Ω dBm = 10 log (V2/R * 1000) = 10 log (V2/600 * 1000) = 10 log (V2/0.600) Therefore, for 0 dBm, V = 774.6 mV or 0 dBm = -2.218 dBV @ 600 Ω (use -2.2 dB) P = V2/R, where for dBm reference, R = 900 Ω dBm = 10 log (V2/R * 1000) = 10 log (V2/900 * 1000) = 10 log (V2/0.900) Therefore, for 0 dBm, V = 948.7 mV or 0 dBm = -0.458 dBV @ 900 Ω (use -0.5 dB) This means that if we substitute 600 Ω for 900 Ω or vice versa, and the voltage remains constant, then we have: Correction (dB) = -10 log 0.600/0.900 = 10 log 0.900/0.600 = 1.761 dB To simplify, Correction (dB) = 10 log( |Z1| / |Z2| ), that is, the log of the ratio of the magnitude of the impedances, when converting from impedance Z1 to Z2. If converting from "Z1 = 600 Ω" to "Z2 = 900 Ω", the correction factor is -1.76 dB (use -1.8 dB), therefore subtract 1.8 dB from the measurement. At this point, depending on the impedance, conversion factors can be applied dB for dB to the measured or calculated result. For example, to convert a 600 Ω, -20 dBm signal to dBV, subtract 2.2 to get -22.2 dBV. Another example is if -20 dBm is measured across 600 Ω, then when measuring across 900 Ω, add a correction of -1.8 dB to get -21.8 dBm (since less power is dissipated by the higher resistance). 58 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Acoustic Sound Pressure Conventions dB Pa (dB Pascal) dBSPL (dB Sound Pressure Level) Where, 0 dB Pa = 94 dBSPL, and 0 dBSPL = 20 microPascals, 1 Pa = 1 N/m2 An A-weighted sound pressure level in dB (dBSPL, A-weighted) is often abbreviated to “dBA”. 59 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Annex C (informative) – R40 Preferred Frequencies The ISO 3, R40 basic series preferred frequencies are listed in Table 15. The frequencies highlighted in Italics are the R10 series of preferred frequencies. The R40 series of preferred frequencies is based on 1/12th octave frequencies, but the numbers are rounded in a convenient pattern. The R10 series is based on the 1/3rd octave frequencies. Table 15 – R40 Preferred Frequencies # 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 Preferred Frequencies (Hz) # 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 100 106 112 118 125 132 140 150 160 170 180 190 200 212 224 236 250 265 280 300 315 335 355 375 400 425 450 475 500 530 560 600 630 670 710 750 800 850 900 950 1000 60 Preferred Frequencies (Hz) 1000 1060 1120 1180 1250 1320 1400 1500 1600 1700 1800 1900 2000 2120 2240 2360 2500 2650 2800 3000 3150 3350 3550 3750 4000 4250 4500 4750 5000 5300 5600 6000 6300 6700 7100 7500 8000 8500 9000 9500 10000 SP-3-4705-RV1 (to be published as ANSI/TIA-920-A) Annex D (normative) – DRP to ERP Transfer Function Frequency response measurements using the Type 3.3 or 3.4 ear are made at the ear Drum Reference Point (DRP). Data collected shall be converted to the Ear Reference Point (ERP) before comparing against the tolerance limits or calculating loudness ratings. The conversion is accomplished by adding the correction factor, SDE, given in Table 16 to the measured data. For complete information refer to ITU-T Recommendation P.58 and IEEE Standard 269 and 269a. Table 16 – DRP to ERP Correction Factors Frequency 92 97 103 109 115 122 130 137 145 154 163 173 183 193 205 218 230 244 259 274 SDE (dB) 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 −0.1 −0.1 0.0 0.1 0.0 −0.1 −0.2 −0.3 −0.3 Frequency (Hz) 290 307 325 345 365 387 410 434 460 487 516 546 579 613 649 688 729 772 818 866 SDE (dB) −0.3 −0.2 −0.2 −0.2 −0.4 −0.5 −0.4 −0.6 −0.3 −0.7 −0.6 −0.6 −0.6 −0.6 −0.8 −0.8 −1.0 −1.1 −1.1 −1.2 Frequency (Hz) 917 972 1029 1090 1155 1223 1296 1372 1454 1540 1631 1728 1830 1939 2053 2175 2304 2441 2585 2738 61 SDE (dB) −1.3 −1.4 −1.8 −2.0 −2.3 −2.4 −2.6 −3.1 −3.3 −3.9 −4.4 −4.8 −5.3 −6.0 −6.9 −7.5 −8.1 −9.1 −9.5 −10.4 Frequency (Hz) 2901 3073 3255 3447 3652 3868 4097 4340 4597 4870 5158 5464 5788 6131 6494 6879 7286 7718 8175 8659 SDE (dB) −11.0 −10.5 −10.2 −9.1 −8.0 −6.9 −5.8 −5.0 −4.2 −3.3 −2.7 −2.4 −2.4 −2.5 −3.3 −4.5 −5.9 −9.0 −14.2 −20.7 Page 36: [1] Deleted a0216538 4/25/2008 3:00 PM Limit Curve Frequency (Hz) Receive Response Limit (dB) [arbitrary level] Upper Limit 100 130 1000 2000 4000 8000 1 4 4 9 9 8 Lower Limit 200 200 800 4000 6600 6600 -infinity -10 -4 -4 -10 -infinity