Speex is a free software speech codec that may be used on VoIP applications and podcasts. Speex claims to be free of any patent restrictions and is licensed under the revised (3-clause) BSD license. It may be used with the Ogg container format or directly transmitted over UDP/RTP.
The Speex designers see their project as complementary to the Vorbis general-purpose audio compression project.
Speex is a lossy format, meaning quality is permanently degraded to reduce file size.
Unlike many other speech codecs, Speex is not targeted at cellular telephony but rather at Voice over IP (VoIP) and file-based compression. The design goals have been to make a codec that would be optimized for high quality speech and low bit rate. To achieve this the codec uses multiple bit rates, and supports ultra-wideband (32 kHz sampling rate), wideband (16 kHz sampling rate) and narrowband (telephone quality, 8 kHz sampling rate). Designing for Voice over IP (VoIP) instead of cell phone use means that Speex must be robust to lost packets, but not to corrupted ones since the User Datagram Protocol (UDP) ensures that packets either arrive unaltered or do not arrive at all. All this led to the choice of Code Excited Linear Prediction (CELP) as the encoding technique to use for Speex. One of the main reasons is that CELP has long proven that it could do the job and scale well to both low bit rates (as evidenced by DoD CELP @ 4.8 kbit/s) and high bit rates (as with G.728 @ 16 kbit/s). The main characteristics can be summarized as follows:
* Free software/open-source, patent and royalty-free
* Integration of narrowband and wideband in the same bit-stream
* Wide range of bit rates available (from 2 kbit/s to 44 kbit/s)
* Dynamic bit rate switching and Variable bit-rate (VBR)
* Voice Activity Detection (VAD, integrated with VBR)
* Variable complexity
* Ultra-wideband mode at 32 kHz (up to 48 kHz)
* Intensity stereo encoding option
Speex is mainly designed for three different sampling rates: 8 kHz (the same sampling rate to transmit telephone calls), 16 kHz, and 32 kHz. These are respectively referred to as narrowband, wideband and ultra-wideband.
Speex encoding is controlled most of the time by a quality parameter that ranges from 0 to 10. In constant bit-rate (CBR) operation, the quality parameter is an integer, while for variable bit-rate (VBR), the parameter is a real (floating point) number.
With Speex, it is possible to vary the complexity allowed for the encoder. This is done by controlling how the search is performed with an integer ranging from 1 to 10 in a way similar to the -1 to -9 options to gzip compression utilities. For normal use, the noise level at complexity 1 is between 1 and 2 dB higher than at complexity 10, but the CPU requirements for complexity 10 is about five times higher than for complexity 1. In practice, the best trade-off is between complexity 2 and 4, though higher settings are often useful when encoding non-speech sounds like DTMF tones, or if encoding is not in real-time.
Variable Bit-Rate (VBR)
Variable bit-rate (VBR) allows a codec to change its bit rate dynamically to adapt to the "difficulty" of the audio being encoded. In the example of Speex, sounds like vowels and high-energy transients require a higher bit rate to achieve good quality, while fricatives (e.g. s and f sounds) can be coded adequately with fewer bits. For this reason, VBR can achieve lower bit rate for the same quality, or a better quality for a certain bit rate. Despite its advantages, VBR has two main drawbacks: first, by only specifying quality, there is no guarantee about the final average bit-rate. Second, for some real-time applications like voice over IP (VoIP), what counts is the maximum bit-rate, which must be low enough for the communication channel.
Average Bit-Rate (ABR)
Average bit-rate solves one of the problems of VBR, as it dynamically adjusts VBR quality in order to meet a specific target bit-rate. Because the quality/bit-rate is adjusted in real-time (open-loop), the global quality will be slightly lower than that obtained by encoding in VBR with exactly the right quality setting to meet the target average bitrate.
Voice Activity Detection (VAD)
When enabled, voice activity detection detects whether the audio being encoded is speech or silence/background noise. VAD is always implicitly activated when encoding in VBR, so the option is only useful in non-VBR operation. In this case, Speex detects non-speech periods and encodes them with just enough bits to reproduce the background noise. This is called "comfort noise generation" (CNG).
Discontinuous Transmission (DTX)
Discontinuous transmission is an addition to VAD/VBR operation, that allows to stop transmitting completely when the background noise is stationary. In a file, 5 bits are used for each missing frame (corresponding to 250 bit/s).
Perceptual enhancement is a part of the decoder which, when turned on, tries to reduce (the perception of) the noise produced by the coding/decoding process. In most cases, perceptual enhancement makes the sound further from the original objectively (signal-to-noise ratio), but in the end it still sounds better (subjective improvement).
Every codec introduces a delay in the transmission. For Speex, this delay is equal to the frame size, plus some amount of "look-ahead" required to process each frame. In narrowband operation (8 kHz), the delay is 30 ms, while for wideband (16 kHz), the delay is 34 ms. These values don't account for the CPU time it takes to encode or decode the frames.