The MP3 challenge

Which file is which?

  • X is the MP3, Y is the original

    Votes: 3 25.0%

  • X is the original, Y is the MP3

    Votes: 8 66.7%

  • I can't perceive a difference

    Votes: 1 8.3%
  • Total voters
    12
Status
Not open for further replies.
mattr

mattr

Resident moody teenager
This is all just a bit of fun, but I hope it can be useful to some at the same time :)

There are much better methods of testing (I might do a quick Foobar ABX tutorial later on), however this is a nice simple, fun and interactive way that everyone can join in at once with. With only two files in the test there is a high probability of correctly guessing just by chance, but as I said its all just a bit of fun.

The challenge

Very simply, here are 2 files:
x.wav
y.wav


One is the uncompressed original WAV.
One is a WAV copy of a 192kbps MP3.


So, which one is which? I'll reveal the answer after a few days.

Most importantly, try not to be influenced by others, and no cheating!
Ears only, no frequency analysis or technical foolery like that!​



I'm sure some with well trained ears and good listening setups (probably the mastering guys) will be able to do this, but I'm sat here listening to the clips wondering where I would even begin if I didn't have a little post-it note next to me saying which was which :D

Bonus rep for anyone who can give correct or plausible reasons why/how they can tell the difference.



Clip taken from Take Five by Dave Brubeck, ripped from the CD "The Very Best of Dave Brubeck" (Copyright 2000 Dave Brubeck + Sony). The sound clips are used under fair use for educational purposes, and remain property of their respective owners.


The purpose of this isn't to say that MP3 is as good as an uncompressed original, because it isn't. Where possible its always best to keep uncompressed archives of the original files in case you ever need a different format, as you can never gain back the lost data and lossy-to-lossy transcoding can introduce some nasty artifacts. I think the main lesson is that you can't always blame a bad mix on the 'crappy MP3 quality' :p
 
Last edited:
hiding my response...

Y is the MP3. The ride cymbal falls off a cliff and splats on every stroke.



Drag your mouse over the white above to read.

mattr said:
So, which one is which? I'll reveal the answer after a few days.


I'm sure some with well trained ears and good listening setups (probably the mastering guys) will be able to do this
Click to expand...
You just need the ears. I listened on my computer speakers in work (the pack-in freebies) and it was plenty obvious.

Mastering guys? You don't need to be anywhere near that level to hear it.
 
Last edited:
I'd say, right away, that there's more difference between the wav and mp3 versions of my own stuff than there is between these two clips. Why's that?
 
dobro said:
I'd say, right away, that there's more difference between the wav and mp3 versions of my own stuff than there is between these two clips. Why's that?
Click to expand...

MP3 comes in 10,000 different flavors. Encoders work in entirely different ways. A 192 kbps mp3 encoded with codec "A" will not have much in common with a 192 kbps mp3 encoded with codec "B".
 
So if I can hear more difference between the wavs and the mp3s that come out of my studio than between the two clips posted here, then I should shift from the codec I'm using to the codec mattr's using, yes?
 
I say x is the original WAV file. I can hear a lot more detail in the lead sax part especially in its reverb tail. It just comes through more clearly to me. Other than that nothing really jumped out at me. I'm also listening on consumer headphones..

My answer is above in white text. Highlight to see it but only after you try the test for yourself!
 
I've listend to them a few times...and while I do hear some differences, I would only make a final decision after hearing them on my studio monitors...not PC speakers.

I just haven't had the chance to do that yet....
 
My response:

Y is the mp3, although Im only marginally confident. I noticed a lack of air around the cymbals probably due to the upper frequencies removed by the compression.

I listened to the samples on my Phonak Audeo IEMs, which have very good high end detail. (Better than my Grado SR60i's which are know for their detailed high end). Since I usually notice the effects of compression more on the high end, I was expecting to be readily able to tell, but it proved to be much more difficult than I thought.
 
The hihat came clearer and more defined in y than in x therefore i believe y is the wav I whited my answer too. listen for yourself!:D
 
Chibi Nappa said:
Yup. :D









(10 character minimum filler text)
Click to expand...


I use my mixing software to encode mp3s. The highest rate in that software is 128 Kbps. I'm thinking the difference between that and the 192 Kbps in the two samples above might account for some of the difference. But I'm going to download the LAME encoder and do an A/B myself.
 
Yeah, the mp3 encoded by my mixing software sound a bit brittle by comparison to the same tune encoded using Winlame. Interesting. Thanks for the idea, youse guys.
 
Nice. I wanted to do a test like this but never got around to it. My guess was that x was the wav, but I wouldn't be surprised if I was told otherwise.
 
Chibi Nappa said:
Moar comparisons naow!!!!!!
Click to expand...

I 2nd that. These are the types of listening exercises we should be doing much more often. Any ideas as to other critcal listening challenges other than mp3 verse wav? Maybe compression verse no compression? Maybe dither verse no dither?
 
Correct answer was... X = original, Y = MP3

7 correct votes, 4 incorrect votes.


Well done to all those that got it right :) The usual things that MP3 trips up on such as cymbal detail and reverb tails gave it away to some.

I'll try and come up with another one tonight.
 
Last edited:
Status
Not open for further replies.
Back
Top