IPB

Welcome Guest ( Log In | Register )

New opus test build
NullC
post Apr 16 2013, 23:31
Post #1





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



I've posted up a new win32 build of libopus git master: https://people.xiph.org/~greg/opus-tools_exp_a8783a1.zip

There aren't many major changes from the prior alpha but a number of small bugs fix, some of which impact quality. I think we're particular interested in knowing about any quality regressions vs prior versions and particular cases where the speech/music detection get confused. (For the actual release we'll likely back that detector off at high rates but for now it's set more aggressively in order to find problems)
Go to the top of the page
+Quote Post
 
Start new topic
Replies
ChristianK
post Jun 16 2013, 22:02
Post #2





Group: Members
Posts: 8
Joined: 14-June 13
Member No.: 108661



Well, I'll just go ahead and post my question and findings here.

I have compiled opus and opus-tools from the git a few days back and compared it to the latest stable version (libopus 1.0.2).
Overall it's a really nice codec which performs very good (especially when you compare it with AAC hev1/2. Where AAC makes music sound a little different Opus still sounds natural).
However I've noticed that at around 42 Kbps it tends to lose a lot of stereo information, is this where Silk kicks in?

When comparing git with the latest stable (libopus 1.1-alpha-161-g28733d1 and 1.0.2) you can really hear a difference at 42 Kbps (git sounds better, no weird distortion but stable still has more stereo).
When going lower than that it now really just sounds like mono to me whereas AAC hev2 still sounds acceptable stereo-wise (and overall too for bitrates like 24 Kbps).

At Wikipedia I read that silk goes from 6 to 40 kbit/s and Celt supports 24 kbit/s to 128 kbit/s.
Although these are probably old figures from before the Opus merge, shouldn't there be more stereo information left at about 32 Kbps?

I've mainly tested this with My God Is the Sun from Queens of the Stone Age and some other songs of that album which seems pretty decent to test for compression artefacts.
Go to the top of the page
+Quote Post
jmvalin
post Jun 16 2013, 22:16
Post #3


Xiph.org Speex developer


Group: Developer
Posts: 485
Joined: 21-August 02
Member No.: 3134



QUOTE (ChristianK @ Jun 16 2013, 17:02) *
I have compiled opus and opus-tools from the git a few days back and compared it to the latest stable version (libopus 1.0.2).
Overall it's a really nice codec which performs very good (especially when you compare it with AAC hev1/2. Where AAC makes music sound a little different Opus still sounds natural).
However I've noticed that at around 42 Kbps it tends to lose a lot of stereo information, is this where Silk kicks in?

When comparing git with the latest stable (libopus 1.1-alpha-161-g28733d1 and 1.0.2) you can really hear a difference at 42 Kbps (git sounds better, no weird distortion but stable still has more stereo).
When going lower than that it now really just sounds like mono to me whereas AAC hev2 still sounds acceptable stereo-wise (and overall too for bitrates like 24 Kbps).


Actually, this has nothing to do with SILK and it's actually done on purpose. The newer version of the encoder gradually reduces the stereo width as the bitrate goes below 48 kbps. The exact amount hasn't been subject to much tuning yet and right now it's just linear between 48 kbps and 32 kb/s (at which point it's all mono). If you're interested in doing testing on this, I can show you how to control the width reduction. Also note that he-aac v2 works very differently from Opus when it comes to stereo because it only encodes a mono stream and then codes some "stereo cues" to fake the stereo effect on the decoder side. This is called parametric stereo and is not something Opus supports.
Go to the top of the page
+Quote Post

Posts in this topic
- NullC   New opus test build   Apr 16 2013, 23:31
- - Dynamic   I was trying to test the speech/music detection, b...   Apr 17 2013, 19:32
|- - jmvalin   QUOTE (Dynamic @ Apr 17 2013, 14:32) I wa...   Apr 17 2013, 21:13
- - Gainless   Good to see the Muse Breaks fix finally bundled. B...   Apr 17 2013, 20:32
|- - jmvalin   QUOTE (Gainless @ Apr 17 2013, 15:32) Goo...   Apr 17 2013, 21:13
|- - Gainless   QUOTE (jmvalin @ Apr 17 2013, 22:13) I th...   Apr 19 2013, 14:57
- - CoRoNe   Using this test build (libopus 1.1a-67) to encode ...   Apr 18 2013, 20:27
- - Gainless   Forgot to add this one: Probably the worst of the...   Apr 20 2013, 10:52
- - IgorC   Gainless, Can You provide some ABC/HR logs compar...   Apr 20 2013, 16:14
|- - Gainless   QUOTE (IgorC @ Apr 20 2013, 17:14) Gainle...   Apr 21 2013, 10:33
- - IgorC   Here are some results for 96 kbps 1.0.2 https://f...   Apr 21 2013, 02:32
- - Kamedo2   I visualized IgorC's precious results on the p...   Apr 21 2013, 03:17
- - IgorC   Thank You, Kamedo2. I was thinking to add it but h...   Apr 21 2013, 04:00
- - Gainless   Ok, I've done ABC-HR tests with the samples at...   Apr 21 2013, 15:18
- - IgorC   There are two first samples (01 and 05) those have...   Apr 25 2013, 05:39
- - Gainless   What I don't really get, is why Opus's VBR...   Apr 25 2013, 10:41
|- - jmvalin   QUOTE (Gainless @ Apr 25 2013, 05:41) Wha...   Apr 25 2013, 13:06
|- - Gainless   QUOTE (jmvalin @ Apr 25 2013, 14:06) That...   Apr 25 2013, 14:00
|- - jmvalin   QUOTE (Gainless @ Apr 25 2013, 09:00) Can...   Apr 25 2013, 16:45
|- - Gainless   QUOTE (jmvalin @ Apr 25 2013, 17:45) Depe...   Apr 25 2013, 21:27
- - IgorC   Gainless, Some time ago You have submited one sam...   Apr 25 2013, 21:56
|- - Gainless   QUOTE (IgorC @ Apr 25 2013, 22:56) Gainle...   Apr 25 2013, 22:11
- - IgorC   Yes, it was Project 100 Sample. http://www.hydrog...   Apr 27 2013, 19:35
- - Gainless   Ok, then this is at least clear. What's probab...   Apr 27 2013, 21:11
- - IgorC   Have tried "Iron Man" sample with the ex...   May 11 2013, 19:26
- - Gainless   Igor, you may also take this sample into considera...   May 12 2013, 12:26
- - IgorC   Gainless, Yes, 1.0.2 was better than the experi...   May 12 2013, 21:42
- - Gainless   Maybe some framesize otpimation/detector can be do...   May 13 2013, 20:33
|- - jmvalin   QUOTE (Gainless @ May 13 2013, 15:33) Btw...   May 14 2013, 04:52
|- - Gainless   Sorry^1000 to Jmvalin and Igor, was my fault with ...   May 14 2013, 08:10
- - Gainless   Here's another sample that could benefit from ...   May 23 2013, 17:15
|- - jmvalin   QUOTE (Gainless @ May 23 2013, 12:15) Her...   May 24 2013, 08:43
|- - Gainless   QUOTE (jmvalin @ May 24 2013, 09:43) QUOT...   Jun 2 2013, 21:41
- - ChristianK   Hi, is this the place to also discus the latest ve...   Jun 14 2013, 19:02
- - ChristianK   Well, I'll just go ahead and post my question ...   Jun 16 2013, 22:02
|- - jmvalin   QUOTE (ChristianK @ Jun 16 2013, 17:02) I...   Jun 16 2013, 22:16
- - ChristianK   Yes, I'm curious on how to control the stereo....   Jun 17 2013, 19:10
|- - jmvalin   QUOTE (ChristianK @ Jun 17 2013, 14:10) Y...   Jun 17 2013, 19:25
|- - kabal4e   QUOTE You should replace the second line simply wi...   Jun 18 2013, 23:06
- - ChristianK   QUOTE (jmvalin @ Jun 17 2013, 20:25) Wher...   Jun 17 2013, 21:22
|- - jmvalin   QUOTE (ChristianK @ Jun 17 2013, 16:22) O...   Jun 18 2013, 00:23
|- - ChristianK   QUOTE (jmvalin @ Jun 18 2013, 01:23) You ...   Jun 30 2013, 20:50
- - Bostedclog   Could you tell me which opus encoder we should be ...   Jun 25 2013, 18:00
- - IgorC   Previously I gave a try to new temporal VBR at 48...   Jul 1 2013, 04:40
- - Gainless   A question: Does the tonality detector inside Opus...   Jul 4 2013, 15:03


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 November 2014 - 00:55