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: [USAC] Unified Speech and Audio Coding Ref Code (Read 10531 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

[USAC] Unified Speech and Audio Coding Ref Code

Hi,
I have got USAC reference code. I have used it, listened and check it quality in PEAQ ODG measure. Now I am disappointed of its quality because it's very poor (e.g. <-3 ODG for 128kbps). In publicly available tests of USAC I have read that it has very good quality.

Am I testing it right? Maybe should I search for some more options in usac encoder to get better quality? Or maybe reference encoder and encoder used in published tests are completely different encoders?

Has anybody tested USAC encoder?
TIA
Guffi

[USAC] Unified Speech and Audio Coding Ref Code

Reply #1
Hi,
I have got USAC reference code. I have used it, listened and check it quality in PEAQ ODG measure. Now I am disappointed of its quality because it's very poor (e.g. <-3 ODG for 128kbps). In publicly available tests of USAC I have read that it has very good quality.

Am I testing it right? Maybe should I search for some more options in usac encoder to get better quality? Or maybe reference encoder and encoder used in published tests are completely different encoders?

Has anybody tested USAC encoder?
TIA
Guffi


Hi Guffi, I am guessing the reference encoder is just a FYI version. The true good quality versions are all not publicly available implementations. Could you point me to the link where you found this reference implementation? I want to take a look and see how good it is as well..

Thanks

[USAC] Unified Speech and Audio Coding Ref Code

Reply #2
Don't expect quality from USAC and AAC reference implementations. There aren't real things.

There is no publicly available high quality USAC encoder despite it was published as standard 3 years ago. In fact there are already a few audio standards which have no available good implmentations (sinusoidal coding, MPEG Surround and now USAC).

Opus has the same quality as USAC at 64 kbps and higher bitrates.

Oh, and it's available

[USAC] Unified Speech and Audio Coding Ref Code

Reply #3
Anyone tried the fork of the reference encoder, called JAME?
It's a dead project so I'm not really expecting anything, just curious.

[USAC] Unified Speech and Audio Coding Ref Code

Reply #4
Anyone tried the fork of the reference encoder, called JAME?
It's a dead project so I'm not really expecting anything, just curious.

Thanks very much for that link! I was looking for JAME for quite some time.
I remember during the MPEG-D USAC standardization that the JAME encoder gave much better quality than the default reference encoder, which as the others said is just a "valid-bit-stream generator".
I don't know its sweet-spot, though (i.e. which bit-rate(s) it was optimized for) and which DEcoder it includes (a custom one?). Are Jeongook Song or Hong-Goo Kang from Yonsei reading this forum?

Edit: now that the Zip file from that link finally finished downloading, it seems the sweet-spots are 24-16 kbit/s stereo, so very low bit-rates which are probably hardly of interest on HA. Still, a comparison with Opus would be interesting. I might give it a try with my HA test-set on the weekend.

Quote
Or maybe reference encoder and encoder used in published tests are completely different encoders?

Yes, the encoder used for the verification tests is Fraunhofer's proprietary USAC encoder which, like for previous MPEG audio codecs, must be purchased.

Chris
If I don't reply to your reply, it means I agree with you.