IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
ABX+ a layman (or noob)'s proposal
LoFiYo
post Mar 25 2004, 16:27
Post #1





Group: Members
Posts: 133
Joined: 2-January 04
Member No.: 10896



When I see someone post ABX results like:

orig vs sample1 -- 12/12
orig vs sample2 -- 12/12
sample1 vs sample2 -- 12/12

unless there is a comment on the results, you couldn't tell which sample was better or harder to ABX. If there is a program that records how much time it took to reach the sufficient -p value, you can tell how difficult it was to ABX each sample.

For example, if you try to ABX fatboy.wav vs fatboy-cbr-96.mp3 (an mp3 file encoded in CBR96), it is very easy, therefore very quick to ABX the two. It might only take 1 minute to reach p = 0.001, whereas if you are trying to compare a very easy-to-encode sample vs alt-preset insane, it will definitely take much more time if possible at all.

So the correlation I see is, the more time it takes to reach the target p value, the more difficult it is to ABX the difference.

So if orig vs sample1 took 2 minutes to reach p=0.005, and if orig vs sample2 took 15 minutes to reach p=0.005, sample2 would be the clear winner.

If someone could make a quick and dirty prog that does this, people could test it and see if my theory is any good...

edit: for clarify

This post has been edited by LoFiYo: Mar 25 2004, 16:43
Go to the top of the page
+Quote Post
tigre
post Mar 25 2004, 16:44
Post #2


Moderator


Group: Members
Posts: 1434
Joined: 26-November 02
Member No.: 3890



Aren't there some ABX tools that log each trial with time? (latest fb2k ABX component does IIRC). Besides this, my 2 cent:
- I think ppl who are ready to spend the time necessary for doing valid ABX tests and posting the results, should be able to write one extra line describing the differences they heard and what sounded better/worse. Asking for such information if missing shouldn't be a problem.
- The time spent doesn't tell much generally, because of
-- listening fatigue / training effect can cause additional differences
-- sometimes you take a break or (as I do) ABX two different tracks alternately at the same time to avoid fatigue and save time.


--------------------
Let's suppose that rain washes out a picnic. Who is feeling negative? The rain? Or YOU? What's causing the negative feeling? The rain or your reaction? - Anthony De Mello
Go to the top of the page
+Quote Post
harashin
post Mar 25 2004, 16:47
Post #3





Group: Members
Posts: 339
Joined: 20-February 02
From: Kyoto, Japan
Member No.: 1362



Actually, I spent more time on Rosemary than on Castanets.


--------------------
Folding@Home Hydrogenaudio.org Team ID# 32639
http://folding.stanford.edu/
Go to the top of the page
+Quote Post
ff123
post Mar 25 2004, 17:01
Post #4


ABC/HR developer, ff123.net admin


Group: Developer (Donating)
Posts: 1396
Joined: 24-September 01
Member No.: 12



WinABX also logs the times

ff123
Go to the top of the page
+Quote Post
[proxima]
post Mar 25 2004, 17:07
Post #5





Group: Members
Posts: 197
Joined: 12-October 02
From: Italy
Member No.: 3537



I have to agree with tigre here. Nor time nor wrong trials are necessarily index of difficulty. Others motivations are:
-Sometimes i try to abx artifacts more subtle than the most annoying one in the sample, just to be sure they exist.
- I can obtain better results flushing my memory and waiting a little time before listening the other sample, fast switching doesn't help me to concentrate.

@LoFiYo: if you're referring to my abx results, try listening with more attention. The artifacts with rebel.wav are really easy to hear for me. I don't think it's a matter of good/bad ears. My equipment is nothing special (SB128 soundcard and HD 497 headphones). Maybe could be a matter of training.


--------------------
WavPack 4.3 -mfx5
LAME 3.97 -V5 --vbr-new --athaa-sensitivity 1
Go to the top of the page
+Quote Post
LoFiYo
post Mar 25 2004, 17:27
Post #6





Group: Members
Posts: 133
Joined: 2-January 04
Member No.: 10896



I see. I almost exclusively use ABC/HR for ABX testing, so I was ignorant of other options... I also ignored the fatigue factor etc. Obviously I have much to learn.

[proxima]: your results and others on this forum only initially made me start thinking about this idea, and I am in no way trying to discredit your results. I respect your spending that much time for the test. If you are any way offended, I am sorry. I will try that sample again later. Maybe I can do better this time.

Thanks smile.gif
Go to the top of the page
+Quote Post
[proxima]
post Mar 25 2004, 17:42
Post #7





Group: Members
Posts: 197
Joined: 12-October 02
From: Italy
Member No.: 3537



QUOTE (LoFiYo @ Mar 25 2004, 05:27 PM)
I am in no way trying to discredit your results. I respect your spending that much time for the test. If you are any way offended, I am sorry.

Offended.. ?? for what ?? blink.gif
Are you joking ?? laugh.gif
Really no reasons to excuse.


--------------------
WavPack 4.3 -mfx5
LAME 3.97 -V5 --vbr-new --athaa-sensitivity 1
Go to the top of the page
+Quote Post

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: 20th December 2014 - 17:13