User:Gmaxwell/celt: Difference between revisions

From XiphWiki
Jump to navigation Jump to search
No edit summary
 
Line 1: Line 1:
==Short term CELT TODO==
{|width="70%"
|
===Overview===


===Updated version with these features===
The CELT codec is a compression algorithm for audio. Like MP3, Vorbis, and
*Allocation out of bounds memory access patch (done:  3b34e188ede286d7a4ecc9fdf66831397a352cb3)
AAC it is suitable for transmitting music with high quality. Unlike these  
*Denorm fixes  (done: de1fd28740a33930b661a64a0ca52bd7c32fa20e)
formats CELT imposes very little delay on the signal, even less than is
*96kHz support (done: 179206a5ac6079470f53d2c231264aa40783e63f)
typical for speech centric formats like Speex, GSM, or G.729.
*Merging of the last band if it's 20% of the prior band size or smaller (not done yet)
*Celtclient fixes (done: ae2fb591b89a8a5cc29fcb89fc53639ebfe84314)
*Celtclient proper UI (easy, but not done)
*Celtclient selective building (screw you autotools; not done)
*Activity interface (started: 06768a7fe7fb3368a00b03540cb9ec83f321005f)
*Lowpass intelligence (not decided how this should work)
**Lowpass as a function of sample rate?
**Lowpass as a function of bytes/frame? Bitrate? Bitrate is problematic because the mode doesn't know the bitrate, although I feel this would be best.
*Merge of raw bits (done: 241b5f0c4506f4d01611c98f296a8c337c2f1f4c)
**bit to uint changes for mode bits (done: 363be091b32324d4b552109b5c174d9b304e18d1)
*Fixes for torbenh's LF tone test case (done: 8256ed4cf745436f4c9b9cd5857917614193f22c)
**Ability to allocate more bits to narrow bands? (more pulses? memory usage? some other approach?)
*CELT_SET_USE_LTP function to disable the use of pitch regardless of the complexity (done: 98046cadcfc1160df434d97b0bdd40f3116f3161)
*Fully tested


==CELT regular testplan==
Using CELT application developers can build software that allows musicians to
*In both --fixed-point and floating point mode
perform together across the Internet, or simply build great sounding
**Using a single very short test file:
telephony systems. Why shouldn't your telephone sound as good as your
***testcelt in valgrind at frame sizes of 64,128,256, 512, with all supported sample rates
stereo?
**Using a collection of a 8 samples for PEAQ testing plus 8 torture samples:
***testcelt in valgrind at 64,128,256, 512 at 32,44.1,48, and 96kHz, at 8,20,40,60,80,100,200 bytes/frame
***Collect PEAQ results on PEAQ test samples (?)
*PEAQ scan - Collection of 8 PEAQ test files (takes ~16 CPU days)
**In both LC and normal mode at 48kHz PEAQ results for every framesize and byte/frame combination


==Paranoia==
CELT is open and unencumbered technology from the Xiph.Org Foundation.
Candidates for assert instrumentation:
|}
* enc_bits/enc_uint should range check their inputs

Latest revision as of 15:18, 14 July 2009

Overview

The CELT codec is a compression algorithm for audio. Like MP3, Vorbis, and AAC it is suitable for transmitting music with high quality. Unlike these formats CELT imposes very little delay on the signal, even less than is typical for speech centric formats like Speex, GSM, or G.729.

Using CELT application developers can build software that allows musicians to perform together across the Internet, or simply build great sounding telephony systems. Why shouldn't your telephone sound as good as your stereo?

CELT is open and unencumbered technology from the Xiph.Org Foundation.