Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: AUPECg2 audio compression (Read 3788 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

AUPECg2 audio compression

I found a very interseting codec here:

http://students.washington.edu/mvinton/



I've done some listening tests and the quaity at 128 is in most cases better than ogg or aac. At 96 (cut-off @ 18,5kHz!!!) it provides very good quality (most files). I've done the tests on various Vangelis tracks, which are very hard to encode.

I have a little problem with english, and I don't understand if this codec is being developed since 1998.

AUPECg2 audio compression

Reply #1
Quote
Originally posted by wolfman
I've done some listening tests and the quaity at 128 is in most cases better than ogg or aac. At 96 (cut-off @ 18,5kHz!!!) it provides very good quality (most files).

I downloaded and tested it on Fatboy.wav. 128kbps is really bad, worse than mp3. I haven't tried higher bitrates yet.

Edit: Now I tested it with higher 256kbps, still bad. Highest setting 392kbps was very good, but it had an annoying click in the beginning.

AUPECg2 audio compression

Reply #2
I tried it a while ago and it had clear artifacts at any bitrate. Should sound quite bad at lower bitrates since it lacks joint stereo modes.

AUPECg2 audio compression

Reply #3
Quote
Originally posted by wolfman
At 96 (cut-off @ 18,5kHz!!!)


The cutoff is a completely meaningless indicator of quality in itself.

But if you feel this is important, you'll be happy to know that the
next version of Ogg won't have a cutoff at all.

--
GCP

AUPECg2 audio compression

Reply #4
Quote
Originally posted by wolfman
I found a very interseting codec here:

http://students.washington.edu/mvinton/ 



I've done some listening tests and the quaity at 128 is in most cases better than ogg or aac. At 96 (cut-off @ 18,5kHz!!!) it provides very good quality (most files). I've done the tests on various Vangelis tracks, which are very hard to encode.

I have a little problem with english, and I don't understand if this codec is being developed since 1998.


I tried this codec a few months ago and I have to say that the quality is horrible.  It's nowhere even close to being as good as aac or ogg.  If anything, the quality is similar to the Bink audio codec (which someone made similar claims about ).  Anyway, this codec artifacted quite heavily on any remotely difficult test clip I tried... it's not even as good as LAME.  I'd stear clear of it if I were you

AUPECg2 audio compression

Reply #5
I've also tried this long ago (it's not updated). Quality was very bad.. It's in essence just a demonstration codec. It's not seriously tweaked.
Juha Laaksonheimo

AUPECg2 audio compression

Reply #6
Quote
Originally posted by JohnV
I've also tried this long ago (it's not updated). Quality was very bad.. It's in essence just a demonstration codec. It's not seriously tweaked.


What I like of that codec is, it does handle Spahm.wav much better than AAC at 128kbps.

 

AUPECg2 audio compression

Reply #7
I wouldn't judge codec quality with fatboy.wav and spahm.wav,

You should try many other samples, and I am sure that you will find out that AAC is probably superior.

I am very busy these days so I can't check it by myself, but - one hint before making various claims:

- Are there any scientific papers regarding particular technologies used in the codec?

- Who are the scientists/developers that were researching these algorithms?

- Their background, previous work... all in all - scientific proof that they made an improvement.

Take a look at AAC for example (not because I am the "supporter" of the particular format, but because it is a very nice example)

- Inventors are biggest names in the audio coding industry (M. Bosi, K. Brandenburg, J. Herre, J. Johnston, H. Fuschs and many others) - working for biggest companies (AT&T, Dolby, FhG IIS, Nokia and Sony) and universities.

- It was developed during four years, with legacy of about 10 years of intensive R&D

- Many scientific papers with scientific proofs were written, as well as Ph.D's

- Many subjective tests were carried out with proof that it is good.

- It is a well-established worldwide ISO/MPEG standard


It is not easy to "invent" a good coding system - and it is very hard to tune one implementation to be state-of-the-art. And, verification of the quality is a very tough process.