Introducing G.729

G.729 is an ITU-T standard compression algorithm for audio data that compresses digital voice in packets of 10 ms duration. Its official name is Coding of speech at 8 kbps using conjugate-structure algebraic-code-excited linear prediction (CS-ACELP).

It is the lowest bit rate ITU-T standard with toll quality, offering opportunities for significant increases in bandwidth utilization in existing telephony and wireless applications. G.729 offers high quality, robust speech performance at the price of complexity at a low bit rate (8 kbps).

It requires 10 ms input frames and generates frames of 80 bits in length. Each of these generated 80 bit frames contains linear prediction factors, excitation code book indices, and gain parameters used by the decoder to reproduce speech. With the G.729 coder processing signals in 10 ms frames and a 5 ms look-ahead, the total algorithmic delay is 15 ms.

Use of this codec in a product requires licensing.

Due to its low bandwidth requirements, G.729 is commonly used by various VoIP applications. To provide more benefits the codec has been extended with various features, designated at G.729A and G.729B.

However, DTMF tones, Fax transmissions and high-quality audio cannot be transported with this algorithm.

Applications

• Digital telephony VoIP, Voice over ATM, Voice over Frame Relay, wireless, satellite, cable
• Videoconferencing, teleconferencing
• Multimedia, e.g., Digital Simultaneous Voice and Data
• Voicemail
• Media gateways
• Test and measurement equipment
• PDA/handset-hosted applications

Features

• MIPS/memory requirements for various platforms are available
• PSQM/PSQM+ values under different network conditions are also available
• Full and half duplex modes of operation
• Passes ITU test vectors
• Optimized for high performance on leading edge DSP architectures
• Multichannel implementation
• Multi-tasking environment compatible
• DAA interface using linear codec at 8.0 kHz sample rate
• Direct interface to 8.0 kHz PCM data stream (A-law or μ-law)
• North American/International Telephony (including caller ID) support available
• Simultaneous DTMF detector operation available - (less than 150 hits on Bellcore test tape typical)
• MF tone detectors, general purpose programmable tone detectors/generators available
• Data/Facsimile/Voice Distinction available
• Common compressed speech frame stream interface to support systems with multiple speech coders
• Dynamic speech coders selection if multiple speech codecs available
• Can be integrated with G.168 Echo Canceller and Tone Detection/Regeneration modules
• Multiple ports can be executed on a single DSP

Technology

• Encoded bandwidth: ~ 200-3400 Hz
• Standardized: ITU-T 1995
• Coding type: CS-ACELP (Conjugate-Structure Algebraic Code Excited Linear Prediction)
• Bit rate: 8 kbps
• Delay (ms):
  • Frame size: 10
  • Look-ahead: 5
• Quality: Toll
• Complexity:
  • MIPS: 20-25
  • RAM (words): < 4 K

Summary for G729 codec

G.729A

G.729A is a compatible extension of G.729 and it is a Conjugate-Structure Algebraic-Code-Excited Linear Prediction (CS-ACELP) speech compression algorithm defined in ITU-T G.729 Annex A. It requires less computational power than G.729. However, lower complexity results in reduced speech quality. It can be employed for VoIP or similar applications using multimedia, voice, and/or data.

Technology

• Encoded bandwidth: ~ 200-3400 Hz
• Standardized: ETSI 1999
• Coding type: CS-ACELP (Conjugate-Structure Algebraic Code Excited Linear Prediction)
• Bit rate: 8 kbps
• Delay (ms):
  • Frame size: 10
  • Look-ahead: 5
• Quality: Toll
• Complexity:
  • MIPS: 10
  • RAM (words): 2 K

Features

• Compliant with G.729, Annex A, Annex B specifications
• Full and half duplex modes to operate
• Passes ITU test vectors
• Optimized for high performance on leading edge DSP architectures
• Multichannel implementation
• Multi-tasking environment compatible
• Common compressed speech frame stream to support systems with multiple codecs
• Voice Activity Detection and Comfort Noise Generation

Applications

• Voice over Internet (VoIP)
• Digital satellite systems
• Voice/Fax/Data relay systems
• Multimedia products
• DCME, PSTN, ISDN
• Voice store and forward
• Video phones

Back to G729

G.729B

G.729B is an extension of G.729 speech codec. G.729 is an ITU-T recommendation and it has been designed to achieve a reduction in the transmitted bit rate in a way that silent periods of human speech has been exploited.

This speech codec codes speech and audio signals that are used in multimedia applications at 8 kbps. It employs Conjugate-Structure Algebraic-Code Excited Linear Prediction (CS-ACELP).

G.729B enables VAD (Voice Activity Detection). It is useful to detect voice activity in the signal. DTX (Discontinuous Transmission) module is also included to increase the overall efficiency.

(DTX is a method that allows momentarily to power-down or mute the telephone if there is no voice input.)

CNG (Comfort Noise Generation) is also included in this process by using 2 byte SID (Silence Insertion Descriptor). It is necessary because speakers may think that the link has been cut if it goes quite in case of no speech. In these cases CNG ensures that an analog hiss is simulated to make the receiver sure that the link is active.

Technology

• Standardized: ITU-T 1996
• Coding type: CS-ACELP (Conjugate-Structure Algebraic Code Excited Linear Prediction)
• Bit rate: <= 8 kbps

Features

• Full duplex and half duplex mode of operation
• Multiple channel compatibility
• All ITU-T test vectors are bit exact compatible
• Common compressed speech frame stream to support systems with multiple codecs
• VAD and CNG compatible
• Seamless integration with RTP/RTCP, DTMF, VAD and jitter management

Applications

• Voice over Internet (VoIP)
• Digital satellite systems
• Voice/Fax/Data relay systems
• Multimedia products
• DCME, PSTN, ISDN
• Voice store and forward
• Video phones

Back to G729

Related Pages

• Supported codecs
• iLBC
• Speex