IPB

Welcome Guest ( Log In | Register )

Multiformat Listening Test @ 64 kbps - FINISHED
Sebastian Mares
post Aug 16 2007, 00:00
Post #1





Group: Members
Posts: 3633
Joined: 14-May 03
From: Bad Herrenalb
Member No.: 6613



The much awaited results of the Public, Multiformat Listening Test @ 64 kbps are ready - partially. So far, I only uploaded an overall plot along with a zoomed version. The details will be available tomorrow. You can also download the encryption key on the results page that is located here:

http://www.listening-tests.info/mf-64-1/results.htm
http://www.listening-tests.info/mf-64-1/resultsz.png

Nero and WMA Professional 10 are tied and WMA Professional 10 is tied to Vorbis. Vorbis however performed worse than Nero. Of course, High Anchor is best and Low Anchor loses.

This one goes to the experts: How would you rank codecs in such a situation, where A=B and B=C, but C<A?


--------------------
http://listening-tests.hydrogenaudio.org/sebastian/
Go to the top of the page
+Quote Post
 
Start new topic
Replies
lexor
post Aug 19 2007, 15:44
Post #2





Group: Members
Posts: 216
Joined: 20-July 03
Member No.: 7896



QUOTE (Sebastian Mares @ Aug 15 2007, 20:00) *
This one goes to the experts:

How would you rank codecs in such a situation, where A=B and B=C, but C<A?

That is actually not a contradiction as such (though further expert opinion on the actual statistical metric used is needed).

You think that is a contradiction, because such situation doesn't happen in "normal" number systems, like integers, reals, etc. What you noticing is the property of total order breaking. However not all valuations have that property.

Take integers, if you take 2 integers at random there is the way to count from one to the other, precisely because there is a total order and you know what is less/greater than what, what equals what and what follows what.
On the other hand take Complex numbers, this is the first number system students usually exposed to in school that doesn't have total order on its elements (though school teachers don't usually mention that). Given 2 random Complex numbers there isn't "the" way to count from one to the other, in fact there are infinitely many ways, all correct in some sense.

So while I don't know if the underlying statistical measure produces set of values that has total order, your example (if not subject to some freaky error) shows that it doesn't, and should be read as raking:

1) HE = WMA
2) Ogg

My immediate intuition would be to use equivalence classes to solve this problem.
1) Make individual comparison between every possible pair
2) Look at the ones with strict inequalities
3) Pick the largest of them all (strictly greater, not >=)
4) Rank that first
5) Add all who are directly equal to it to its equivalence class (not ones that are equal by some chain of equalities)
6) Removing them from further consideration
7) From remaining, rank the next largest as number 2
8) Go to 5 and repeat for the rest of the ranking.

This post has been edited by lexor: Aug 19 2007, 16:00


--------------------
The Plan Within Plans
Go to the top of the page
+Quote Post
halb27
post Aug 19 2007, 22:37
Post #3





Group: Members
Posts: 2439
Joined: 9-October 05
From: Dormagen, Germany
Member No.: 25015



QUOTE (lexor @ Aug 19 2007, 16:44) *
QUOTE (Sebastian Mares @ Aug 15 2007, 20:00) *

This one goes to the experts:

How would you rank codecs in such a situation, where A=B and B=C, but C<A?

That is actually not a contradiction as such...

Same opinion for me.
For clearly defined objects A, B, C, a clearly defined identy and a clearly defined <-relation it would be a contradiction.
Here A and B correspond to the quality of Nero HE AAC and that of WMA pro, and C corresponds to that of Vorbis. Quality as measured with this test.
The problem is in the meaning of '=' and '<' as these are rough quality comparison operators which can easily make up for such a pseudo-contradiction.

The zoomed view is a major evil to me as it overestimates such a rough '<' comparison.
From absolute view ranging from 1.0 to 5.0 it's easy to say 'all these three encoders yield roughly the same quality with vorbis being a tiny bit behind.'
This is what is most important in practice, cause with these codecs you usually don't have the choice which one to use on a mobile device. No matter which one you use: you get state of the art 64 kbps technology regarding quality.

If it's up to elaborating differences between the encoders the very personal preferences are much more of concern than the overall small quality differences according to the test.

This post has been edited by halb27: Aug 19 2007, 22:39


--------------------
lame3100m -V1 --insane-factor 0.75
Go to the top of the page
+Quote Post

Posts in this topic
- Sebastian Mares   Multiformat Listening Test @ 64 kbps - FINISHED   Aug 16 2007, 00:00
- - guruboolez   Wow, thanks a lot for posting so fast these result...   Aug 16 2007, 00:06
|- - -Nepomuk-   Compare to the last 48kbit/s listening test, 64kbi...   Aug 16 2007, 00:27
||- - ff123   QUOTE (-Nepomuk- @ Aug 15 2007, 16...   Aug 16 2007, 01:43
||- - guruboolez   QUOTE (ff123 @ Aug 16 2007, 01:43) It...   Aug 16 2007, 01:52
|- - Sebastian Mares   QUOTE (guruboolez @ Aug 16 2007, 01:06) E...   Aug 16 2007, 00:32
- - rjamorim   Very interesting, Sebastian. Congratulations, and ...   Aug 16 2007, 00:28
- - kdo   Nice! I'm a little surprised that Vorbis ...   Aug 16 2007, 00:34
- - Sebastian Mares   All you need is Chunky! http://www.phong.org/c...   Aug 16 2007, 00:45
|- - kdo   QUOTE (Sebastian Mares @ Aug 16 2007, 01...   Aug 16 2007, 00:48
- - guruboolez   My personal results: CODE WMAPro high Vorbis low H...   Aug 16 2007, 01:21
- - kwanbis   QUOTE (Sebastian Mares @ Aug 15 2007, 23...   Aug 16 2007, 01:58
|- - echo   QUOTE (kwanbis @ Aug 16 2007, 03:58) QUOT...   Aug 17 2007, 00:36
|- - Woodinville   QUOTE (kwanbis @ Aug 15 2007, 17:58) QUOT...   Aug 21 2007, 20:48
- - kdo   Very interesting. After all my results are not so ...   Aug 16 2007, 02:00
- - kennedyb4   It seems that Itunes at 96 VBR has outscored Itune...   Aug 16 2007, 02:58
- - ff123   QUOTE (Sebastian Mares @ Aug 15 2007, 16...   Aug 16 2007, 05:42
- - vinnie97   Guru, my taste mirrors yours on Vorbis...anything ...   Aug 16 2007, 07:39
- - Slacker   Little Question: How do I use the key to see my re...   Aug 16 2007, 09:17
|- - kdo   QUOTE (Slacker @ Aug 16 2007, 10:17) Litt...   Aug 16 2007, 09:34
- - Alexxander   I used the key and decrypted results through java ...   Aug 16 2007, 09:46
|- - Sebastian Mares   QUOTE (Alexxander @ Aug 16 2007, 10:46) I...   Aug 16 2007, 12:49
- - muaddib   Wow, much more results than what I expected! T...   Aug 16 2007, 10:07
|- - muaddib   QUOTE (muaddib @ Aug 16 2007, 11:07) QUOT...   Sep 26 2007, 13:40
- - Alex B   Here are my personal results: CODE% Sample Averag...   Aug 16 2007, 11:53
- - thana   i downloaded the rar file and tried to process the...   Aug 16 2007, 14:12
|- - kdo   QUOTE (thana @ Aug 16 2007, 15:12) i down...   Aug 16 2007, 14:20
|- - Alex B   QUOTE (thana @ Aug 16 2007, 16:12) i down...   Aug 16 2007, 15:08
- - Rio   QUOTE (Sebastian Mares @ Aug 16 2007, 07...   Aug 16 2007, 15:11
|- - pdq   QUOTE (Rio @ Aug 16 2007, 10:11) QUOTE (S...   Aug 16 2007, 16:07
|- - benski   QUOTE (Rio @ Aug 16 2007, 10:11) QUOTE (S...   Aug 16 2007, 16:09
|- - ff123   QUOTE (benski @ Aug 16 2007, 08:09) One t...   Aug 16 2007, 16:20
|- - Whelkman   QUOTE (ff123 @ Aug 16 2007, 11:20) In the...   Aug 16 2007, 17:26
- - naylor83   Stupid question alert: If I ranked the reference,...   Aug 16 2007, 15:58
- - Sebastian Mares   The decrypted result files will then contain the r...   Aug 16 2007, 16:00
- - naylor83   Stupid question alert (again): I'm trying to ...   Aug 16 2007, 16:13
- - guruboolez   You can use MrQuestionMan, foobar2000 or several o...   Aug 16 2007, 16:17
|- - naylor83   QUOTE (guruboolez @ Aug 16 2007, 17:17) 1...   Aug 16 2007, 16:19
- - Sebastian Mares   Does anyone know how to make Excel to refer to the...   Aug 16 2007, 21:39
|- - mezenga   QUOTE (Sebastian Mares @ Aug 16 2007, 12...   Aug 17 2007, 00:17
- - Sebastian Mares   Uploaded the plots for each sample. The correspond...   Aug 16 2007, 23:31
- - ff123   Interesting. he-aac had some clear winners over w...   Aug 17 2007, 00:18
- - rockcake   I'd also like to give a big thankyou to Sebast...   Aug 17 2007, 05:32
|- - TechVsLife   QUOTE (rockcake @ Aug 17 2007, 00:32) I...   Aug 17 2007, 07:43
- - IgorC   Thanks for test. Nero has done a good work.   Aug 17 2007, 08:05
- - Alexxander   Congrat Nero! I can't believe WMA Pro 10 ...   Aug 17 2007, 09:30
|- - halb27   QUOTE (Alexxander @ Aug 17 2007, 10:30) ....   Aug 17 2007, 10:21
||- - Alexxander   QUOTE (halb27 @ Aug 17 2007, 11:21) ...Mo...   Aug 17 2007, 11:29
|- - muaddib   QUOTE (Alexxander @ Aug 17 2007, 10:30) C...   Aug 17 2007, 10:22
- - Ivan Dimkovic   CBR, in this context, means: "Fixed bit rate...   Aug 17 2007, 11:40
- - Alex B   Sebastian, Could you possibly post the average re...   Aug 17 2007, 12:45
|- - muaddib   QUOTE (Alex B @ Aug 17 2007, 13:45) Could...   Aug 17 2007, 13:05
|- - Alex B   QUOTE (muaddib @ Aug 17 2007, 15:05) It i...   Aug 17 2007, 13:37
- - muaddib   This looks much better now. Nero is at its place ...   Aug 17 2007, 14:02
- - lexor   QUOTE (Sebastian Mares @ Aug 15 2007, 20...   Aug 19 2007, 15:44
|- - halb27   QUOTE (lexor @ Aug 19 2007, 16:44) QUOTE ...   Aug 19 2007, 22:37
- - Jillian   How about use approximation instead of equality.   Aug 19 2007, 16:06
- - Nikaki   Interesting results. I only use Vorbis for high bi...   Aug 22 2007, 15:20
|- - TechVsLife   Wouldn't it make more sense just to go with on...   Aug 22 2007, 16:40
|- - Junon   QUOTE (TechVsLife @ Aug 22 2007, 17:40) W...   Aug 22 2007, 17:10
|- - TechVsLife   Thanks, I forgot about that whole dimension to the...   Aug 22 2007, 18:16
- - Woodinville   Out of curiousity, why is there no castinettes in ...   Aug 27 2007, 22:31
|- - [JAZ]   QUOTE (Woodinville @ Aug 27 2007, 23:31) ...   Aug 27 2007, 22:46
|- - Woodinville   QUOTE ' date='Aug 27 2007, 14:46' post...   Aug 28 2007, 18:03
- - Sebastian Mares   Well, I asked for sample suggestions long before t...   Aug 28 2007, 18:08
- - muaddib   How many of you that took test did use Headphones?...   Aug 31 2007, 15:28
|- - elmar3rd   QUOTE (muaddib @ Aug 31 2007, 14:28) How ...   Aug 31 2007, 15:54
||- - IgorC   Headphones for 17 samples. Loudspeakers for White...   Aug 31 2007, 17:19
|- - Alex B   I too used the Koss PortaPro headphones, which are...   Aug 31 2007, 20:08
- - ff123   headphones for me   Aug 31 2007, 15:35
- - kdo   good headphones (mid-price Beyerdynamic) with som...   Aug 31 2007, 17:44
- - naylor83   Headphones: Koss PortaPro   Aug 31 2007, 19:43
- - muaddib   It seems to me that in most cases headphones revea...   Sep 3 2007, 10:58
|- - naylor83   QUOTE (muaddib @ Sep 3 2007, 11:58) It se...   Sep 3 2007, 11:17
|- - muaddib   QUOTE (naylor83 @ Sep 3 2007, 12:17) Coul...   Sep 3 2007, 12:23
- - Whelkman   I used a pair of Sennheiser PX 100s for the test. ...   Sep 3 2007, 19:28
- - kdo   All of a sudden, I have got a small question -- ab...   Sep 6 2007, 13:54
- - Sebastian Mares   Within a sample plot, all bars should have the sam...   Sep 6 2007, 15:29
|- - kdo   QUOTE (Sebastian Mares @ Sep 6 2007, 16:2...   Sep 6 2007, 16:06
- - Sebastian Mares   Maybe someone with more knowledge in statistics ca...   Sep 6 2007, 17:20
- - robert   Who said all bars should be equal? What do you wan...   Sep 6 2007, 18:00
- - Sebastian Mares   In my results (and Roberto's, Guru's and f...   Sep 6 2007, 23:06
|- - naylor83   QUOTE (Sebastian Mares @ Sep 7 2007, 00:0...   Sep 6 2007, 23:11
- - ff123   For this type of analysis, the error bars are all ...   Sep 7 2007, 03:48
- - robert   So the bars do not represent the distribution of d...   Sep 7 2007, 10:58
|- - ff123   QUOTE (robert @ Sep 7 2007, 02:58) So the...   Sep 8 2007, 03:38
- - Moguta   I would've loved to see MP3 involved in this t...   Sep 8 2007, 02:56
- - Sebastian Mares   Yes, this is normal and depends on the mood, the l...   Sep 26 2007, 14:28
- - benwaggoner   Say, are there any plans for doing a new test here...   Apr 10 2008, 23:25
|- - benski   QUOTE (benwaggoner @ Apr 10 2008, 18:25) ...   Apr 10 2008, 23:29
|- - benwaggoner   QUOTE (benski @ Apr 10 2008, 14:29) QUOTE...   Apr 11 2008, 02:57
- - Sebastian Mares   The reason why WMA was tested in CBR mode is that ...   Apr 11 2008, 18:43
|- - benwaggoner   QUOTE (Sebastian Mares @ Apr 11 2008, 09...   Apr 11 2008, 21:20
- - Sebastian Mares   The test performed by NSTL featured WMA in CBR mod...   Apr 11 2008, 21:45
|- - benwaggoner   QUOTE (Sebastian Mares @ Apr 11 2008, 12...   Apr 11 2008, 23:28
- - hellokeith   QUOTE (benwaggoner @ Apr 11 2008, 15:20) ...   Apr 12 2008, 08:27
|- - benwaggoner   QUOTE (hellokeith @ Apr 11 2008, 23:27) N...   Apr 13 2008, 06:19
- - IgorC   I tried 1 and 2 pass CBR wma10 at 64 kbit/s in pas...   Apr 12 2008, 12:55
- - vinnie97   I'm still anxiously awaiting the forthcoming ~...   Apr 14 2008, 08:30


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 October 2014 - 22:45