BLIND TEST: Lossless vs. MP3 320 - Page 7

The point is, if you didn't do the test properly, you didn't prove anything to yourself.  What it sounds like you mean to say is 'I stopped when I was happy with the results' which is very different than 'I completed the test and was able to hear a difference.'

Or at least, you think you did. This is why I have a hard time taking audiophiles with golden ears seriously: because they rarely go through rigorous testing. Their testimony remains purely subjective (and dare I say, sometimes purely fictional). What we end up with is a grossly inflated statistic of "people who can tell the difference".

I'm not really interested in what people think they hear. I'm more interested in what they actually hear. It's cool if you don't want to play.

I got lost during the comparison of 320 vs "lost-less".

I believe its perfectly possible to tell the difference between a FLAC and a mp3 320kps. I also believe its extremely difficult to tell which one is which, and I doubt the enjoyment factor will changes depending on lossless or lossy.


Ive downloaded the tracks, but I dont have the ABX plug-in. Ill get it, but I can already tell I will fail bigtime unless I spend considerable time with those songs (and I wont, because I hate them...). Actually, even if I spent "enough" time with them, im not sure I could differ between these two.

Also (for the tracks ive already spent 100s of hours with) I find far harder to spot the difference/tell which one is which with well master tracks, compared to really hot masters. Simple, acoustic songs are harder than stuff that covers most of the audible hearing range as well.


I think these tests are complete BS because no sane (plenty idiots though) person would claim that one of the files is more enjoyable than the other. To spot the difference, you need training and to be familiar with the track.



While Ive been writing this post I DLed the ABX plugin and gave it a go and the song had some exploits I could abuse:

foo_abx 1.3.4 report
foobar2000 v1.1.11
2012/04/17 02:31:38

File A: C:\Users\The Copy Cat\Desktop\A.wav
File B: C:\Users\The Copy Cat\Desktop\B.wav

02:31:38 : Test started.
02:32:23 : 01/01  50.0%
02:32:54 : 02/02  25.0%
02:33:20 : 03/03  12.5%
02:33:50 : 04/04  6.3%
02:35:27 : 05/05  3.1%
02:35:43 : 06/06  1.6%
02:35:56 : 07/07  0.8%
02:36:04 : 07/08  3.5%
02:36:14 : 08/09  2.0%
02:36:31 : 09/10  1.1%
02:36:47 : 10/11  0.6%
02:37:22 : 11/12  0.3%
02:37:42 : 12/13  0.2%
02:38:05 : 13/14  0.1%
02:38:13 : 14/15  0.0%
02:38:21 : 15/16  0.0%
02:38:31 : 15/17  0.1%
02:38:37 : 16/18  0.1%
02:38:43 : 17/19  0.0%
02:38:53 : 18/20  0.0%
02:38:59 : Test finished.

 ----------
Total: 18/20 (0.0%)


Before people start yelling; I cheated, I admit that at once. Liberate use of compressing and some EQing. If you are young or have good hearing range you can spot the differences.

In normal listening I would not manage those results.

If someone use these results to prove there is an audible difference that impacts the listening experience in a negative you should go fu(k yourself.

PS. I still dont know which one is lossless, but based on what I know about audio codecs (not enough) my guess is B is lossless.




 

I just want to point out that although lame mp3 is frequently chosen to represent lossy audio, there are better options available.

 

The 2 best encoders are actually Apple's AAC-LC encoder and Autov Vorbis imo.

So, is there anybody who was able to successfully pass the test without cheating?!! 

I don't hear any artifact (maybe I'm not trained to do so?) but IMO B has a richer music, exactly around 9 and 32 seconds. I voted for B lossless.

Edited by jHoNDoE - 9/10/12 at 12:22am

MP3 compression doesn't affect the overall sound (ie: "richer" "warmer"), it created distortion artifacts in sounds it has trouble encoding. If you hear an overall difference, odds are the volumes are a little different.

BTW: no eq used.

Edited by ghardashyan - 10/24/12 at 4:04am

And another pass:

 

Pretty easy to beat the test apparently. Or maybe, you just use ripped super low quality Mp3s so it was easy to hear. This seems especially possible since some many golden ears need what, 6 seconds to tell the difference? That's barely enough time to hear a damn thing.

 

Or maybe you just use a voice recording, where one says "I am MP3". The other says, "I am CD".

There is no way to prove the fact that I don't cheat, so you gonna just believe me.

No. .

 

 

Actually .... yes.  Look at post #25 in this thread.  Deadlylover did it - and I believe him without question - he's previously been active in many of these threads.  He's also been happy to admit it when he gets 320 vs FLAC which are completely transparent.  He's one of the very few that definitely knows what to look for - but who also is realistic :

[a] that it is extremely hard to isolate the difference

[b] admitted he couldn't tell which one was lossless and which one wasn't - just could tell the difference in the files when listened to critically

[c] would be happy with either one (if it was the type of music he was into)

 

Basically the differences are so small to be un-noticeable during normal listening conditions.

 

DL is definitely the exception (being able to differentiate) rather than the rule.

 

 

And sorry @ghardashyan - the mere fact that you asked whether anyone had done it without cheating, the time taken in your log, and the gear used (I'm doubting if it would be transparent/detailed enough - DL used a Stax system) - I'm guessing you cheated as well.  If you didn't - well then you are one of the very few with the skills to do this.  If you did cheat, I'd encourage you to come clean and just admit that you are normal - like the vast majority of us 

I agree with mentioned above. Same do state I.

Have you noticed the difference between dates of original post of mine and of the next 2 posts ( where I show the logs ).

What's wrong with the time in the logs?
I used my Brainwavz B2 IEM's ( which are very analytical and have decent extension on treble ) paired with FiiO E17 connected to my laptop @ 24/ 96. Also I state that my ears do hear up to 21.5 kHz( which is not a rare thing ). And I spent a lot of time to find a part in the track ( you can see it on screenshots ) where I was able to distinguish the difference in the sounding of high freq parts of these two tracks. It was difficult, but once you find what and where to listen, it starts to be easier to distinguish the difference. The difference was nor in detail, nor in separation, nor in soundstage etc. The difference was in the *aggressiveness* of the treble in certain, small amount of time in certain part of one of the tracks...

As mentioned above it is impossible to prove the fact that I am not cheating. So maybe you have to just believe me ( or not ).

And another thing: why do I need to fake the results here in the forum where nobody knows me?
And sorry for my bad English.