The Beta5 has finally arrived!
# The action of noise normalization has been improved. This has an effect in the sound roughness and tremor problem etc. in the low bitrate.
# The threshold of a stereo mode change was calculated dynamically. Some of problems which originate in channel coupling by this will improve.
# Noise control of an impulse block was changed (quality 0-10 / 32-48kHz). And pre-echo decreased slightly.
# Tuning of each part was redone according to above-mentioned changed part and additional part.
http://www.geocities.jp/aoyoume/aotuv/
Full Version: AoTuV Beta 5 released!
Excellent new! The changelog looks great.
I'll try to start a new listening test very soon.
Just tried my standard bad samples harp40_1, herding_calls and trumpet on b5.
At -q2 no good quality can be expected from these bad samples, but herding_calls turns out to be pretty acceptable. Not quite so with trumpet and harp.
At -q3 I'd call trumpet not annoying anymore.
At -q4 even harp is quite acceptable to me.
These are very good results!
Something else: With the simple commandline 'venc -qx y.wav' which seems to be the only way to use venc I have seen no way to use venc from within foobar.
Any solutions?
At -q2 no good quality can be expected from these bad samples, but herding_calls turns out to be pretty acceptable. Not quite so with trumpet and harp.
At -q3 I'd call trumpet not annoying anymore.
At -q4 even harp is quite acceptable to me.
These are very good results!
Something else: With the simple commandline 'venc -qx y.wav' which seems to be the only way to use venc I have seen no way to use venc from within foobar.
Any solutions?
QUOTE(halb27 @ Oct 24 2006, 21:40) 
Something else: With the simple commandline 'venc -qx y.wav' which seems to be the only way to use venc I have seen no way to use venc from within foobar.Any solutions?
I guess you missed this:
CODE
http://www.geocities.jp/aoyoume/aotuv/binary/oggenc_aoTuV_b5.zip
Aoyumi released also an oggenc.exe version.
Edit: Perhaps external linking is prohibited. I changed the link to code.
I tried with a few samples, but I'm unable to hear a difference between the latest Lancer and this new beta. I tried my favourite Vorbis problem sample (the_product) at both low and high quality settings, as well as a few normal samples at -q0, but I was unable to distinguish between the two encoded versions. I didn't do many compressed/compressed comparsions before, so I might just be too "unexperienced" to spot the improvements.
Edit: I finally ABXed a few differences, at -q-2. b5 sounded better to me in all cases so far.
Edit: I finally ABXed a few differences, at -q-2. b5 sounded better to me in all cases so far.
QUOTE(Alex B @ Oct 24 2006, 20:59)
I guess you missed this: .... Aoyumi released also an oggenc.exe version.
You're right. Thanks a lot.
Alex B: that oggenc link isn't working.
Edit: strange, I can't get it to download with firefox, but wget had no problems...
Edit: strange, I can't get it to download with firefox, but wget had no problems...
QUOTE(Firon @ Oct 24 2006, 21:41)
Alex B: that oggenc link isn't working. ...
I couldn't use it either but I found the link on Aoyumi's page (just a few lines below the venc link).
It didn't work when I tried from Aoyumi's page either, but I just did a wget on it from my linux machine and all was fine.
Firefox was caching the broken response, opening it on a new tab and doing Ctrl F5 fixed it.
Firefox was caching the broken response, opening it on a new tab and doing Ctrl F5 fixed it.
It's a strange link. For me, it didn't work directly from my post, but when I pasted it into the browser's address bar it worked. (Opera 9.02)
I am going to retest the same samples I tried with the earlier pre-beta 5 version @ -q 1 here:
http://www.hydrogenaudio.org/forums/index....st&p=401514
I suppose it would be helpful for Aoyumi if others who are interested about -q 1 could try the same samples.
I would like to test -q 6 or 7 also, but I am not sure if have enough guts for that. I may have found a couple of new samples that may be problematic at -q 6-7, but ABXing them is still very difficult (at least for me). I would first need to be sure that I can ABX the r1 encoded samples and then try b5.
I am going to retest the same samples I tried with the earlier pre-beta 5 version @ -q 1 here:
http://www.hydrogenaudio.org/forums/index....st&p=401514
I suppose it would be helpful for Aoyumi if others who are interested about -q 1 could try the same samples.
I would like to test -q 6 or 7 also, but I am not sure if have enough guts for that. I may have found a couple of new samples that may be problematic at -q 6-7, but ABXing them is still very difficult (at least for me). I would first need to be sure that I can ABX the r1 encoded samples and then try b5.
Good news.
Speed test (PentiumM 715, 1500Mhz)
Speed test (PentiumM 715, 1500Mhz)
CODE
C:\temp\test>oggenc_lancer.exe naraya_nonamed.wav
Opening with wav module: WAV file reader
Encoding "naraya_nonamed.wav" to
"naraya_nonamed.ogg"
at quality 3,00
[100,0%] [ 0m00s remaining] |
Done encoding file "naraya_nonamed.ogg"
File length: 3m 25,0s
Elapsed time: 0m 7,735s
Rate: 26,539668
Average bitrate: 104,7 kb/s
C:\temp\test>oggenc_aotuv_b451.exe naraya_nonamed.wav
Opening with wav module: WAV file reader
Encoding "naraya_nonamed.wav" to
"naraya_nonamed.ogg"
at quality 3,00
[100,0%] [ 0m00s remaining] \
Done encoding file "naraya_nonamed.ogg"
File length: 3m 25,0s
Elapsed time: 0m 34,0s
Rate: 6,0378
Average bitrate: 104,7 kb/s
C:\temp\test>oggenc_aotuv_b5.exe naraya_nonamed.wav
Opening with wav module: WAV file reader
Encoding "naraya_nonamed.wav" to
"naraya_nonamed.ogg"
at quality 3,00
[100,0%] [ 0m00s remaining] \
Done encoding file "naraya_nonamed.ogg"
File length: 3m 25,0s
Elapsed time: 0m 38,0s
Rate: 5,4022
Average bitrate: 104,8 kb/s
Opening with wav module: WAV file reader
Encoding "naraya_nonamed.wav" to
"naraya_nonamed.ogg"
at quality 3,00
[100,0%] [ 0m00s remaining] |
Done encoding file "naraya_nonamed.ogg"
File length: 3m 25,0s
Elapsed time: 0m 7,735s
Rate: 26,539668
Average bitrate: 104,7 kb/s
C:\temp\test>oggenc_aotuv_b451.exe naraya_nonamed.wav
Opening with wav module: WAV file reader
Encoding "naraya_nonamed.wav" to
"naraya_nonamed.ogg"
at quality 3,00
[100,0%] [ 0m00s remaining] \
Done encoding file "naraya_nonamed.ogg"
File length: 3m 25,0s
Elapsed time: 0m 34,0s
Rate: 6,0378
Average bitrate: 104,7 kb/s
C:\temp\test>oggenc_aotuv_b5.exe naraya_nonamed.wav
Opening with wav module: WAV file reader
Encoding "naraya_nonamed.wav" to
"naraya_nonamed.ogg"
at quality 3,00
[100,0%] [ 0m00s remaining] \
Done encoding file "naraya_nonamed.ogg"
File length: 3m 25,0s
Elapsed time: 0m 38,0s
Rate: 5,4022
Average bitrate: 104,8 kb/s
Yay, finally the long wait is over! Gotta test this release on my favourite samples, and start encoding right away. Thanks, Aoyumi!
I'll wait for an OggDropXPd compile from John33 (hurry up, mate ! 
) and then start testing the new sound from Aoyumi 
) and then start testing the new sound from Aoyumi
I've already started encoding my stuff to -q2. The biggest change I've noticed is that the bitrates are slightly lower over 4.51.
QUOTE(Firon @ Oct 24 2006, 19:19)
The biggest change I've noticed is that the bitrates are slightly lower over 4.51.
I noticed that, too. I wonder if it's intentional, as if to combat slight bitrate inflation present during past tests? Although, to tell the truth, I didn't notice an issue there.
But then again, maybe it's simply due to a new bitrate spread as a result of new algorithms...
Cool!! Thanks Aoyumi. 
Now me waits for fresh lancer builds.
Now me waits for fresh lancer builds.
OMG Very Surprised 
QUOTE(beto @ Oct 24 2006, 18:28)
Cool!! Thanks Aoyumi.
Now me waits for fresh lancer builds.
You read my mind.
Congrats on the milestone, Aoyumi!
I am happy that the somehow new version was releasable.
QUOTE(VEG @ Oct 25 2006, 06:23)
Speed test (PentiumM 715, 1500Mhz)
...
This is normal. Processing required for encoding is increasing (especially channel coupling). ...
The encoders and dlls have been added at Rarewares.
Thanks Aoyumi.
The encoders and dlls have been added at Rarewares.
oggdropXPd V.1.8.9 using aoTuVb5 2006-10-25 doesn't work on my computer (lpac_codec_api.dll)
Where is the Generic - P3/AMD - P4 Only compile ?
QUOTE(john33 @ Oct 25 2006, 17:38)
The encoders and dlls have been added at Rarewares.
oggdropXPd V.1.8.9 using aoTuVb5 2006-10-25 doesn't work on my computer (lpac_codec_api.dll)
Where is the Generic - P3/AMD - P4 Only compile ?
Thanks Aoyumi!
Btw, would be nice to see Lancer's SSE optimized versions.
Btw, would be nice to see Lancer's SSE optimized versions.
QUOTE(hugo25 @ Oct 25 2006, 17:01)
...
oggdropXPd V.1.8.9 using aoTuVb5 2006-10-25 doesn't work on my computer (lpac_codec_api.dll)
Can you provide more info please? Was there an error message? It works fine here.
QUOTE(hugo25 @ Oct 25 2006, 17:01)
Where is the Generic - P3/AMD - P4 Only compile ?
The single compile should work reasonably well on all systems. I'm trying to get away from producing dozens of compiles each time there's a change; it takes me far too long to get through them.
These compiles use Blacksword's (of Lancer fame) routine for runtime optimisation with the Intel compiler.
QUOTE(john33 @ Oct 26 2006, 02:04)
QUOTE(hugo25 @ Oct 25 2006, 17:01)
...
oggdropXPd V.1.8.9 using aoTuVb5 2006-10-25 doesn't work on my computer (lpac_codec_api.dll)
Can you provide more info please? Was there an error message? It works fine here.
Current compile never run without following dlls also on my WinXP.
LPAC dlls (54Kb) - Monkey's Audio dll (71Kb) - OptimFROG dll (81Kb) on Rarewares
QUOTE(hugo25 @ Oct 25 2006, 08:01)
...
oggdropXPd V.1.8.9 using aoTuVb5 2006-10-25 doesn't work on my computer (lpac_codec_api.dll)
...
I get an error saying I miss that dll too.
Sorry!! 
I forgot to set the "delayload" option on the decoding dlls. I'll fix it soon. Just as well you let me know!!
I forgot to set the "delayload" option on the decoding dlls. I'll fix it soon. Just as well you let me know!!
Fixed and uploaded again. This zip is 425KB.
Thanks .
.
Thanks John.
Thanks John!
Thanks Aoyumi!! 
<> Beta 5 of aoTuV
<> New (non-beta!) WMA10 Pro codecs coming out in 5 days...
<> November is going to be a blast
Let's get ready for a massive low-bitrate shootout!!
<> Beta 5 of aoTuV
<> New (non-beta!) WMA10 Pro codecs coming out in 5 days...
<> November is going to be a blast
Let's get ready for a massive low-bitrate shootout!!
Is this competitive with AAC+ at 64 kbps now or am I imagining things?
It might very well be, since roughness was improved at low bitrates and quality improved overall.
Test it and find out.
Test it and find out.
QUOTE(The Sheep of DEATH @ Oct 26 2006, 03:32)
<> New (non-beta!) WMA10 Pro codecs coming out in 5 days...
Is it certain? Could you name the source?
It really interests me
On a short ABX test, I managed to discern aoTuV b5 from original at -q5 on three different samples. It seems the r1's -q5 sounds the same, cause it has the same problems and I failed to abx r1 from b5 at this quality setting. Additional details, the samples used and ABX logs are here.
Still, great job at -q3 and below, Vorbis is once again unmatched at mid-low range.
Still, great job at -q3 and below, Vorbis is once again unmatched at mid-low range.
QUOTE(guruboolez @ Oct 26 2006, 16:04)
QUOTE(The Sheep of DEATH @ Oct 26 2006, 03:32)
<> New (non-beta!) WMA10 Pro codecs coming out in 5 days...
Is it certain? Could you name the source?
It really interests me
Zambelli said that WMP11 ought to be out any day now.
But I get this from someone in Doom9
http://neowin.net/index.php?act=view&id=35798
sample name: ae_gg
sample origin: Mo0zOoH
encoder versions: aoTuV beta 4.51 & aoTuV beta 5
encoder setting: -q4,00
test: ABC/HR + ABX (hidden results) including both encodings in the same session
Interesting sample; aoTuV encodings sound unusually dirty (a bit "hollow" and clearly noisy) at this setting.
While REF vs OGG comparisons was rather obvious OGG vs OGG was impossible for me. In other words I didn't noticed any improvement with beta 5 over beta 4.51 (aka RC1).
sample name: eig
sample origin: Mo0zOoH
encoder versions: aoTuV beta 4.51 & aoTuV beta 5
encoder setting: -q4,00
test: ABC/HR + ABX (hidden results) including both encodings in the same session
Excellent sample for testing pre-echo/smearing.
aoTuV beta 4.51 sounds bad on this sample (2.0/5 ; 8/8 pval = 0.0030): smearing is really strong and annoys me a lot.
aoTuV beta 5 is much less irritating but smearing is still perceptible (4.0/5 ; 11/12 pval = 0.0030). This obvious improvement has no significant impact on bitrate (+1 kbps).
There are no results for nin_quake.wv: I can't hear anything wrong with both aoTuV versions at -q4,00.
EDIT: ABCHR logs:
ABC/HR for Java, Version 0.5b, 26 octobre 2006
Testname:
Tester:
1L = C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav
2L = C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
---------------------------------------
General Comments:
---------------------------------------
1L File: C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav
1L Rating: 3.0
1L Comment: Disapointing performance of Vorbis/aoTuV on this sample!
It really sounds different (empty, or metallic - and obviously dirty).
---------------------------------------
2L File: C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
2L Rating: 3.0
2L Comment: sounds identical.
---------------------------------------
ABX Results:
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav
11 out of 12, pval = 0.0030
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
12 out of 12, pval < 0.001
C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
7 out of 12, pval = 0.387
---- Detailed ABX results ----
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav
Playback Range: 06.203 to 09.217
2:27:23 PM p 1/1 pval = 0.5
2:27:27 PM p 2/2 pval = 0.25
2:27:30 PM p 3/3 pval = 0.125
2:27:33 PM p 4/4 pval = 0.062
2:27:37 PM p 5/5 pval = 0.031
2:27:40 PM f 5/6 pval = 0.109
2:28:01 PM p 6/7 pval = 0.062
2:28:04 PM p 7/8 pval = 0.035
2:28:10 PM p 8/9 pval = 0.019
2:28:12 PM p 9/10 pval = 0.01
2:28:15 PM p 10/11 pval = 0.0050
2:28:17 PM p 11/12 pval = 0.0030
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
Playback Range: 06.203 to 09.217
2:28:34 PM p 1/1 pval = 0.5
2:28:37 PM p 2/2 pval = 0.25
2:28:40 PM p 3/3 pval = 0.125
2:28:46 PM p 4/4 pval = 0.062
2:28:48 PM p 5/5 pval = 0.031
2:28:50 PM p 6/6 pval = 0.015
2:28:53 PM p 7/7 pval = 0.0070
2:28:55 PM p 8/8 pval = 0.0030
2:28:58 PM p 9/9 pval = 0.0010
2:29:00 PM p 10/10 pval < 0.001
2:29:03 PM p 11/11 pval < 0.001
2:29:06 PM p 12/12 pval < 0.001
C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
Playback Range: 06.203 to 09.217
2:31:13 PM p 1/1 pval = 0.5
2:31:27 PM f 1/2 pval = 0.75
2:31:29 PM p 2/3 pval = 0.5
2:31:56 PM p 3/4 pval = 0.312
2:32:06 PM f 3/5 pval = 0.5
2:32:09 PM p 4/6 pval = 0.343
2:32:14 PM f 4/7 pval = 0.5
2:32:19 PM p 5/8 pval = 0.363
2:32:22 PM f 5/9 pval = 0.5
2:32:26 PM f 5/10 pval = 0.623
2:32:30 PM p 6/11 pval = 0.5
2:32:33 PM p 7/12 pval = 0.387
ABC/HR for Java, Version 0.5b, 26 octobre 2006
Testname:
Tester:
1R = C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav
2L = C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
---------------------------------------
General Comments:
---------------------------------------
1R File: C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav
1R Rating: 2.0
1R Comment: strong smearing
---------------------------------------
2L File: C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
2L Rating: 4.0
2L Comment: much better but still perceptible
---------------------------------------
ABX Results:
C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
16 out of 16, pval < 0.001
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
11 out of 12, pval = 0.0030
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav
8 out of 8, pval = 0.0030
---- Detailed ABX results ----
C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
Playback Range: 02.304 to 03.966
2:37:50 PM p 1/1 pval = 0.5
2:37:53 PM p 2/2 pval = 0.25
2:38:01 PM p 3/3 pval = 0.125
2:38:03 PM p 4/4 pval = 0.062
2:38:05 PM p 5/5 pval = 0.031
2:38:07 PM p 6/6 pval = 0.015
2:38:10 PM p 7/7 pval = 0.0070
2:38:13 PM p 8/8 pval = 0.0030
2:38:17 PM p 9/9 pval = 0.0010
2:38:20 PM p 10/10 pval < 0.001
2:38:23 PM p 11/11 pval < 0.001
2:38:25 PM p 12/12 pval < 0.001
2:38:28 PM p 13/13 pval < 0.001
2:38:30 PM p 14/14 pval < 0.001
2:38:32 PM p 15/15 pval < 0.001
2:38:35 PM p 16/16 pval < 0.001
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
Playback Range: 02.304 to 03.966
2:36:36 PM p 1/1 pval = 0.5
2:36:39 PM p 2/2 pval = 0.25
2:36:42 PM p 3/3 pval = 0.125
2:36:45 PM p 4/4 pval = 0.062
2:36:51 PM p 5/5 pval = 0.031
2:36:53 PM p 6/6 pval = 0.015
2:36:58 PM p 7/7 pval = 0.0070
2:37:13 PM f 7/8 pval = 0.035
2:37:16 PM p 8/9 pval = 0.019
2:37:19 PM p 9/10 pval = 0.01
2:37:26 PM p 10/11 pval = 0.0050
2:37:29 PM p 11/12 pval = 0.0030
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav
Playback Range: 02.304 to 03.966
2:36:02 PM p 1/1 pval = 0.5
2:36:05 PM p 2/2 pval = 0.25
2:36:07 PM p 3/3 pval = 0.125
2:36:10 PM p 4/4 pval = 0.062
2:36:12 PM p 5/5 pval = 0.031
2:36:14 PM p 6/6 pval = 0.015
2:36:16 PM p 7/7 pval = 0.0070
2:36:18 PM p 8/8 pval = 0.0030
sample origin: Mo0zOoH
encoder versions: aoTuV beta 4.51 & aoTuV beta 5
encoder setting: -q4,00
test: ABC/HR + ABX (hidden results) including both encodings in the same session
Interesting sample; aoTuV encodings sound unusually dirty (a bit "hollow" and clearly noisy) at this setting.
While REF vs OGG comparisons was rather obvious OGG vs OGG was impossible for me. In other words I didn't noticed any improvement with beta 5 over beta 4.51 (aka RC1).
sample name: eig
sample origin: Mo0zOoH
encoder versions: aoTuV beta 4.51 & aoTuV beta 5
encoder setting: -q4,00
test: ABC/HR + ABX (hidden results) including both encodings in the same session
Excellent sample for testing pre-echo/smearing.
aoTuV beta 4.51 sounds bad on this sample (2.0/5 ; 8/8 pval = 0.0030): smearing is really strong and annoys me a lot.
aoTuV beta 5 is much less irritating but smearing is still perceptible (4.0/5 ; 11/12 pval = 0.0030). This obvious improvement has no significant impact on bitrate (+1 kbps).
There are no results for nin_quake.wv: I can't hear anything wrong with both aoTuV versions at -q4,00.
EDIT: ABCHR logs:
CODE
ABC/HR for Java, Version 0.5b, 26 octobre 2006
Testname:
Tester:
1L = C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav
2L = C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
---------------------------------------
General Comments:
---------------------------------------
1L File: C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav
1L Rating: 3.0
1L Comment: Disapointing performance of Vorbis/aoTuV on this sample!
It really sounds different (empty, or metallic - and obviously dirty).
---------------------------------------
2L File: C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
2L Rating: 3.0
2L Comment: sounds identical.
---------------------------------------
ABX Results:
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav
11 out of 12, pval = 0.0030
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
12 out of 12, pval < 0.001
C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
7 out of 12, pval = 0.387
---- Detailed ABX results ----
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav
Playback Range: 06.203 to 09.217
2:27:23 PM p 1/1 pval = 0.5
2:27:27 PM p 2/2 pval = 0.25
2:27:30 PM p 3/3 pval = 0.125
2:27:33 PM p 4/4 pval = 0.062
2:27:37 PM p 5/5 pval = 0.031
2:27:40 PM f 5/6 pval = 0.109
2:28:01 PM p 6/7 pval = 0.062
2:28:04 PM p 7/8 pval = 0.035
2:28:10 PM p 8/9 pval = 0.019
2:28:12 PM p 9/10 pval = 0.01
2:28:15 PM p 10/11 pval = 0.0050
2:28:17 PM p 11/12 pval = 0.0030
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
Playback Range: 06.203 to 09.217
2:28:34 PM p 1/1 pval = 0.5
2:28:37 PM p 2/2 pval = 0.25
2:28:40 PM p 3/3 pval = 0.125
2:28:46 PM p 4/4 pval = 0.062
2:28:48 PM p 5/5 pval = 0.031
2:28:50 PM p 6/6 pval = 0.015
2:28:53 PM p 7/7 pval = 0.0070
2:28:55 PM p 8/8 pval = 0.0030
2:28:58 PM p 9/9 pval = 0.0010
2:29:00 PM p 10/10 pval < 0.001
2:29:03 PM p 11/11 pval < 0.001
2:29:06 PM p 12/12 pval < 0.001
C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b45.wav vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\ae_gg.b5.wav
Playback Range: 06.203 to 09.217
2:31:13 PM p 1/1 pval = 0.5
2:31:27 PM f 1/2 pval = 0.75
2:31:29 PM p 2/3 pval = 0.5
2:31:56 PM p 3/4 pval = 0.312
2:32:06 PM f 3/5 pval = 0.5
2:32:09 PM p 4/6 pval = 0.343
2:32:14 PM f 4/7 pval = 0.5
2:32:19 PM p 5/8 pval = 0.363
2:32:22 PM f 5/9 pval = 0.5
2:32:26 PM f 5/10 pval = 0.623
2:32:30 PM p 6/11 pval = 0.5
2:32:33 PM p 7/12 pval = 0.387
CODE
ABC/HR for Java, Version 0.5b, 26 octobre 2006
Testname:
Tester:
1R = C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav
2L = C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
---------------------------------------
General Comments:
---------------------------------------
1R File: C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav
1R Rating: 2.0
1R Comment: strong smearing
---------------------------------------
2L File: C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
2L Rating: 4.0
2L Comment: much better but still perceptible
---------------------------------------
ABX Results:
C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
16 out of 16, pval < 0.001
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
11 out of 12, pval = 0.0030
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav
8 out of 8, pval = 0.0030
---- Detailed ABX results ----
C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
Playback Range: 02.304 to 03.966
2:37:50 PM p 1/1 pval = 0.5
2:37:53 PM p 2/2 pval = 0.25
2:38:01 PM p 3/3 pval = 0.125
2:38:03 PM p 4/4 pval = 0.062
2:38:05 PM p 5/5 pval = 0.031
2:38:07 PM p 6/6 pval = 0.015
2:38:10 PM p 7/7 pval = 0.0070
2:38:13 PM p 8/8 pval = 0.0030
2:38:17 PM p 9/9 pval = 0.0010
2:38:20 PM p 10/10 pval < 0.001
2:38:23 PM p 11/11 pval < 0.001
2:38:25 PM p 12/12 pval < 0.001
2:38:28 PM p 13/13 pval < 0.001
2:38:30 PM p 14/14 pval < 0.001
2:38:32 PM p 15/15 pval < 0.001
2:38:35 PM p 16/16 pval < 0.001
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b5.wav
Playback Range: 02.304 to 03.966
2:36:36 PM p 1/1 pval = 0.5
2:36:39 PM p 2/2 pval = 0.25
2:36:42 PM p 3/3 pval = 0.125
2:36:45 PM p 4/4 pval = 0.062
2:36:51 PM p 5/5 pval = 0.031
2:36:53 PM p 6/6 pval = 0.015
2:36:58 PM p 7/7 pval = 0.0070
2:37:13 PM f 7/8 pval = 0.035
2:37:16 PM p 8/9 pval = 0.019
2:37:19 PM p 9/10 pval = 0.01
2:37:26 PM p 10/11 pval = 0.0050
2:37:29 PM p 11/12 pval = 0.0030
Original vs C:\Documents and Settings\autre\Mes documents\Mes tests\petits tests\aotuv beta 5\eig.b45.wav
Playback Range: 02.304 to 03.966
2:36:02 PM p 1/1 pval = 0.5
2:36:05 PM p 2/2 pval = 0.25
2:36:07 PM p 3/3 pval = 0.125
2:36:10 PM p 4/4 pval = 0.062
2:36:12 PM p 5/5 pval = 0.031
2:36:14 PM p 6/6 pval = 0.015
2:36:16 PM p 7/7 pval = 0.0070
2:36:18 PM p 8/8 pval = 0.0030
QUOTE(guruboolez @ Oct 26 2006, 04:04)
QUOTE(The Sheep of DEATH @ Oct 26 2006, 03:32)
<> New (non-beta!) WMA10 Pro codecs coming out in 5 days...
Is it certain? Could you name the source?
It really interests me
Yeah, I heard if from someone who claims to have read it in an MS newsgroup. And the link Jillian posted above helps
This is really making it hard for me not to swtich back to Rockbox
Awesome, thank you so much, Aoyumi! The hard work you put into this for a few months really paid off, a number of piano recordings that hiss aren't as objectionable at q2 now. (Not to say the artifacts are gone, but I've taken to running a white noise removal filter over any songs that show artifacts anyway - cleans the sound right up.)
Just recompiled vorbis-tools and xmms-oggre with aoTuV b5.
That, plus Ubuntu 6.10 being released, made my week.
That, plus Ubuntu 6.10 being released, made my week.
i downloaded Oggenc2.83 using aoTuVb5 2006-10-25 but the command line still says OggEnc v2.83 (libvorbis aoTuVr1) 
-andy-
-andy-
QUOTE(AndyMutz @ Oct 30 2006, 00:26)
i downloaded Oggenc2.83 using aoTuVb5 2006-10-25 but the command line still says OggEnc v2.83 (libvorbis aoTuVr1)
-andy-
Ooops, sorry!!
That's only embedded text, the vendor string displayed within foobar will tell you which library version was used and that should be showing the correct version. It's one of those things that are easily forgotten in the rush to recompile. Thanks for drawing my attention to this, it is somewhat misleading so I'll recompile and upload again later, but it is using the correct libs.
Whoops - the .dlls at rarewares for use with CDEx don't work, they do not show up in the codec selection dropdown menu. What's wrong? The aoTuV r1 version works normally.
EDIT: If I put the dll package from aoyumi's webpage into the folder, it is correctly recognized (aoTuV b5).
EDIT: If I put the dll package from aoyumi's webpage into the folder, it is correctly recognized (aoTuV b5).
QUOTE(ImAlive @ Oct 30 2006, 17:52)
Whoops - the .dlls at rarewares for use with CDEx don't work, they do not show up in the codec selection dropdown menu. What's wrong? The aoTuV r1 version works normally.
EDIT: If I put the dll package from aoyumi's webpage into the folder, it is correctly recognized (aoTuV b5).
Hmmm, they work fine here!?!? I specifically tested to make sure as there have been problems in the past. What system are you using? Anyone else have this problem?
XP Home, Service Pack 1 (d'oh), german, with CDEx 1.71b2 unicode. I tried a fresh reinstall of CDEx, however, after copying the dlls into the directory, they did not show up in the list. As mentioned, r1 package and b5 package from aoyumi's page work.
QUOTE(ImAlive @ Oct 30 2006, 21:58)
XP Home, Service Pack 1 (d'oh), german, with CDEx 1.71b2 unicode. I tried a fresh reinstall of CDEx, however, after copying the dlls into the directory, they did not show up in the list. As mentioned, r1 package and b5 package from aoyumi's page work.
Hmmm, very puzzling.
These are plain MSVC8 compiles compiled on an Athlon64 3700+, Xp Pro SP2 fully up to date, and they run on this system fine. My other test system is a Pentium D 940, same OS, and they're not seen!! I'll try to get to the bottom of this tomorrow, but if anyone has any bright ideas?
Well, I've uploaded a set of b5 dlls compiled using VC6 that run on both my test systems, but I'll continue to pursue the VC8 compile mystery.
QUOTE(vinnie97 @ Oct 25 2006, 22:30)
Is this competitive with AAC+ at 64 kbps now or am I imagining things?
Yes indeed--especially with raised lowpass (I raised it to 16.5khz and I have trouble discerning it from the original in some cases).
But then again, WMA10Pro is certainly competitive with both of them...
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.