Blind Test 2
Oct 22, 2010 at 11:31 PM Thread Starter Post #1 of 57

Crazy*Carl

500+ Head-Fier
Joined
Aug 8, 2009
Posts
852
Likes
28
I am going to treat this one a bit different.  Instead of giving away what the two different file types are, I want you guys to say which one is the higher quality.  Like my previous test, different flac compression level is used for each to cover up the file type.  This results in slightly different file sizes.
 
Track is Pink Floyd - Mother
 
http://www.fileden.com/files/2008/2/16/1763003/Pink_A.flac
 
http://www.fileden.com/files/2008/2/16/1763003/Pink_B.flac
 
Oct 23, 2010 at 6:03 PM Post #6 of 57
 

I got 7/9 abx score, the cymbals are darker one one of them.I guess it's b, higher quality, does the foobar abx comparator pick first file as a and second as b, or random. Still think it's b, but i am tired and can hardly tell.
 
Oct 23, 2010 at 9:24 PM Post #7 of 57


Quote:
I got 7/9 abx score, the cymbals are darker one one of them.I guess it's b, higher quality, does the foobar abx comparator pick first file as a and second as b, or random. Still think it's b, but i am tired and can hardly tell.


Should be random for each trial.
 
se
 
 
 
Oct 24, 2010 at 10:14 AM Post #8 of 57


Quote:
 
I got 7/9 abx score, the cymbals are darker one one of them.I guess it's b, higher quality, does the foobar abx comparator pick first file as a and second as b, or random. Still think it's b, but i am tired and can hardly tell.





 
Quote:
Should be random for each trial.
 
se
 
It only randomises X - A and B are static once allocated - you can view the text file it creates, so far in my tests first in list is always A - I just got 7/9 always guessing A = X and not even listening to the tracks - I have noticed this before with FooBar - once I ran it up to 20 trails it evened out a lot but even then it was only showing me as 74% probability of guessing - I  would suggest 20 trials as a minimum to overcome this.
 
 



 
Oct 24, 2010 at 7:07 PM Post #9 of 57


Quote:
 
It only randomises X - A and B are static once allocated - you can view the text file it creates, so far in my tests first in list is always A - I just got 7/9 always guessing A = X and not even listening to the tracks - I have noticed this before with FooBar - once I ran it up to 20 trails it evened out a lot but even then it was only showing me as 74% probability of guessing - I  would suggest 20 trials as a minimum to overcome this.


Thanks for the answer, i know the probabilities of seven out of nine, but i got tired and started to second guess everything. I did hear the difference between them. If i had to guess in an abx test without consciously hearing a difference i would have to consider the tracks identical. Also forgot about the text file, this is basically the second time i used the plugin. I too thought the guessing probability is wrong. At 7/9 the guessing probability was 12%. That sounds lower than it should be.
 
Oct 24, 2010 at 7:12 PM Post #10 of 57


Quote:
Thanks for the answer, i know the probabilities of seven out of nine, but i got tired and started to second guess everything. I did hear the difference between them. If i had to guess in an abx test without consciously hearing a difference i would have to consider the tracks identical. Also forgot about the text file, this is basically the second time i used the plugin. I too thought the guessing probability is wrong. At 7/9 the guessing probability was 12%. That sounds lower than it should be.


Wavoman would know the answer to the probability calculation.
 
 
Oct 24, 2010 at 7:30 PM Post #11 of 57
I couldn't hear a difference really. I'll try again in a bit though. I know I can tell the difference between 160 and FLAC on most files, and I can mostly tell 192 from FLAC. I suspect this is higher bitrate. It's really tough.
 
Oct 24, 2010 at 7:49 PM Post #12 of 57
I have no clue which one is lossless. I scored 0/2 and read PJ's comments about the vocals, which was then the focus of my foobar ABX session. I then scored 6/6 on the following attempts. Despite these results, I doubt I could repeat them and I could care less which copy I owned for casual listening.
 
Oct 25, 2010 at 7:29 AM Post #13 of 57
Cannot hear a difference.
 
Oct 26, 2010 at 4:31 PM Post #14 of 57
This thread hasn't had nearly the activity of the first - probably because most of us can't tell a difference. When are you gonna spill the beans about the difference between the files?
 
Oct 26, 2010 at 8:05 PM Post #15 of 57
Can't tell much of a difference either, but B sounds a little better.
 
BTW- 100th Post
beyersmile.png
!!!!!!!!!!!!!!!!!
 

Users who are viewing this thread

Back
Top