IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
CELT 0.5.0 is out, New low-complexity mode
jmvalin
post Oct 14 2008, 05:12
Post #1


Xiph.org Speex developer


Group: Developer
Posts: 475
Joined: 21-August 02
Member No.: 3134



I'd like to announce version 0.5.0 of CELT. CELT is a very low delay (<10 ms), high quality audio codec suitable for applications such as network music performance and high-quality teleconferencing.

This release includes simplifications to the bit allocation code and includes a new low-complexity mode that can be enabled (without breaking compatibility) on the encoder side and which results in a 50% saving on the CPU. There are several other optimisations and bug fixes included. The API has also been slightly modified, so code using CELT will have to be updated (minor changes).

Note that the bit-stream still isn't frozen yet. So the good thing is that you can still give us feedback so we can improve it. The bad thing is that you can't rely on backward compatibility yet.
Go to the top of the page
+Quote Post
NullC
post Dec 19 2008, 17:55
Post #2





Group: Developer
Posts: 200
Joined: 8-July 03
Member No.: 7653



QUOTE (jmvalin @ Oct 13 2008, 20:12) *
I'd like to announce version 0.5.0 of CELT. CELT is a very low delay (<10 ms), high quality audio codec suitable for applications such as network music performance and high-quality teleconferencing.


Perhaps not worth a new full announcement:

QUOTE (celt-codec.org @ Dec 17 2008)
Version 0.5.1 is out. This release fixes a number of issues with 0.5.0. Several cases where the encoder could fail have been fixed, including one that could cause memory corruption (and possibly a segfault). In addition, the VQ search has been improved, which results in a small CPU reduction, while fixing some quality problems with low frequency tones. Other than that, there should be no significant quality changes.


CELT is still undergoing active development but we're pushing towards finalizing the bitstream. We still have much to work on: Dynamic bit allocation, improved stereo coding, and overall psychoacoustic tuning are still TODO items. Now is a good a time as any to get involved if you'd like a chance to help improve the codec *before* the bitstream is set in stone. It would be especially useful to discover if anyone is aware of any torture samples with which CELT currently gives poor results.

If you'd like to help out with CELT but don't know where to start, feel free to drop me an email or join the CELT discussion in #vorbis on freenode irc.

Cheers
Go to the top of the page
+Quote Post
rbrito
post Mar 1 2009, 08:28
Post #3





Group: Members
Posts: 55
Joined: 16-August 05
Member No.: 23948



QUOTE (jmvalin @ Oct 14 2008, 01:12) *
(...)
Note that the bit-stream still isn't frozen yet. So the good thing is that you can still give us feedback so we can improve it. The bad thing is that you can't rely on backward compatibility yet.


This part is one that interests me somewhat. Please, don't rush to freeze the bit-stream. In this stage, please announce it so that more people know that the bit-stream isn't fixed yet, so that they shouldn't rely on it for storage yet, but this is for the good of the format.

Perhaps announcing it at a high volume site like Slashdot would be a good thing (if I had not checked Hydrogenaudio, I would not even know about CELT). Of course, there will be whiners (they are always there), but the thing is that it will raise more exposure to CELT and some moderately informed people may become aware of it.


Thanks, Rogério Brito.
Go to the top of the page
+Quote Post
NullC
post Mar 2 2009, 17:38
Post #4





Group: Developer
Posts: 200
Joined: 8-July 03
Member No.: 7653



QUOTE (rbrito @ Mar 1 2009, 00:28) *
This part is one that interests me somewhat. Please, don't rush to freeze the bit-stream. In this stage, please announce it so that more people know that the bit-stream isn't fixed yet, so that they shouldn't rely on it for storage yet, but this is for the good of the format.

Perhaps announcing it at a high volume site like Slashdot would be a good thing (if I had not checked Hydrogenaudio, I would not even know about CELT). Of course, there will be whiners (they are always there), but the thing is that it will raise more exposure to CELT and some moderately informed people may become aware of it.


Perhaps you can suggest some good things to have completed (obviously not the bitstream) before we make a push for wider attention. There are a number of obvious things, like a more accessible explanation of the purpose of CELT on the web-page. I recently added win32 builds of the simple example encoder/decoder tools, which I considered a pre-requisite (so that non-developers on Windows could try out the quality of CELT), but I'm curious to hear if you have other ideas.

I think that the general public doesn't really get that codecs add delay, nor really appreciate the importance of delay until they experience a first hand comparison, nor do they appreciate the quality improvement of full-bandwidth speech vs wideband speech. So I'd expect a slashdot announcement to complains by folks who don't appreciate the purpose of CELT being unhappy that CELT isn't something else entirely. I also don't expect there to be a lot of hidden codec developers lurking on slashdot. Of course, I could be totally wrong, but this is why I wouldn't rush to draw that kind of attention.





Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 23rd July 2014 - 18:02