Hydrogenaudio Forums > Orion_II.wav with Vorbis 1.1.0 @q6

Full Version: Orion_II.wav with Vorbis 1.1.0 @q6

Hydrogenaudio Forums > Hydrogenaudio Forum > Listening Tests

lotr

Feb 26 2005, 09:31

Hi all!

This is my first post and my first ABX test. I hope you find the results useful.

I ABXed Orion_II.wav against the same sample encoded with Vorbis 1.1.0 @q6

CODE

23:08:00 : Test started.
23:09:12 : 01/01 50.0%
23:09:44 : 02/02 25.0%
23:09:52 : 03/03 12.5%
23:10:11 : 04/04 6.3%
23:11:01 : 05/05 3.1%
23:11:23 : 06/06 1.6%
23:11:37 : 07/07 0.8%
23:11:58 : 08/08 0.4%
23:12:11 : 09/09 0.2%
23:12:43 : 10/10 0.1%
23:12:55 : 11/11 0.0%
23:13:23 : 12/12 0.0%
23:14:00 : 13/13 0.0%
23:14:52 : 14/14 0.0%
23:15:19 : 15/15 0.0%
23:17:20 : 15/16 0.0%
23:17:29 : Test finished.

----------
Total: 15/16 (0.0%)

There was a "smudged" kind of distortion (don't know how to describe it exactly) during the first 6 seconds. It was a bit difficult to spot.

Feel free to suggest any other samples/settings I should test.

QuantumKnot

Feb 27 2005, 00:02

Thank you for your valuable tests. I was wondering whether you can perform the same test using my recent experimental version of Vorbis that sports a new assistive block switching algorithm that may reduce the smearing:

http://rarewares.org/quantumknot/oggenc-qktune20050227.exe

Many thanks

lotr

Feb 27 2005, 08:52

Ok, here are the results of Orion_II.wav vs your version of the encoder at q6:

CODE

foo_abx v1.2 report
foobar2000 v0.8.3
2005/02/27 16:09:50

File A: file://C:\Audio Installation\samples\Orion_II.wav
File B: file://C:\Documents and Settings\X\Επιφάνεια εργασίας\Vorbis tuned\001 Orion_II_gk.ogg

16:09:50 : Test started.
16:10:48 : 01/01 50.0%
16:11:36 : 02/02 25.0%
16:11:44 : 03/03 12.5%
16:11:57 : 04/04 6.3%
16:12:09 : 05/05 3.1%
16:12:15 : 06/06 1.6%
16:12:21 : 07/07 0.8%
16:12:48 : 08/08 0.4%
16:12:54 : 09/09 0.2%
16:13:00 : 10/10 0.1%
16:13:11 : 11/11 0.0%
16:13:18 : 12/12 0.0%
16:13:23 : 13/13 0.0%
16:13:33 : 14/14 0.0%
16:13:39 : 15/15 0.0%
16:13:46 : 16/16 0.0%
16:13:52 : 17/17 0.0%
16:14:09 : 18/18 0.0%
16:14:15 : 19/19 0.0%
16:14:20 : 20/20 0.0%
16:14:27 : 21/21 0.0%
16:14:33 : 22/22 0.0%
16:14:45 : Test finished.

----------
Total: 22/22 (0.0%)

The same smearing was there.

Then I tested Vorbis 1.1.0 at q6 against your version at q6:

CODE

foo_abx v1.2 report
foobar2000 v0.8.3
2005/02/27 16:16:22

File A: file://C:\Documents and Settings\X\Επιφάνεια εργασίας\Vorbis 110\001 Orion_II_110.ogg
File B: file://C:\Documents and Settings\X\Επιφάνεια εργασίας\Vorbis tuned\001 Orion_II_gk.ogg

16:16:23 : Test started.
16:18:06 : 01/01 50.0%
16:18:35 : 02/02 25.0%
16:19:12 : 03/03 12.5%
16:19:59 : 04/04 6.3%
16:20:05 : 04/05 18.8%
16:20:15 : 05/06 10.9%
16:21:03 : 05/07 22.7%
16:21:57 : 05/08 36.3%
16:22:33 : 06/09 25.4%
16:23:14 : 06/10 37.7%
16:23:23 : 06/11 50.0%
16:23:50 : 06/12 61.3%
16:23:58 : 07/13 50.0%
16:28:14 : 08/14 39.5%
16:28:21 : 09/15 30.4%
16:28:28 : 10/16 22.7%
16:28:39 : 11/17 16.6%
16:28:53 : 12/18 11.9%
16:30:01 : 13/19 8.4%
16:30:45 : 13/20 13.2%
16:30:55 : 14/21 9.5%
16:31:00 : 14/22 14.3%
16:31:08 : 15/23 10.5%
16:31:28 : 16/24 7.6%
16:31:41 : 16/25 11.5%
16:31:52 : 16/26 16.3%
16:32:01 : 16/27 22.1%
16:32:05 : Test finished.

----------
Total: 16/27 (22.1%)

I noticed no difference in smearing at all.

At about trial 11, I concentrated between seconds 4-6. I was given the impression that 1.1.0 produced a fuller sound in that part of the sample, but after 2-3 trials that impression dissipated and I couldn't tell any difference between the two samples.

Hope this helps!

QuantumKnot

Feb 27 2005, 17:39

Ah, so it may not be a block switching problem then. Many thanks.

This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.