[skchow]

http://mp3ornot.com/

Thought I would share this little test w/ all of you.  My ears aren't golden enough; I can hardly tell the difference.

Sunny.

[nathanm]

The funny thing about bitrates is that they can have an inverse psychological effect.  Usually we think we hear differences that are measurably not there, but with lossy compression you KNOW that the two files are measurably different and yet you may not necessarily  hear a difference!  It's weird.  If offered a quality setting in any given application I will usually crank it up on principle rather than real world results.  Even though I've never compared 128 to 320 in an ABX test, I still will set it on 320 because it "must" be better, right? Heh!

[laserboi]

I failed

[DeanSheen]

That was fun.

I passed via. my sweet KOSS TD\80 work supplied headphones.

Theres a pretty easy tell in there that I think they used to bias the results for joe blow.


*********spoiler***********spoiler************



One of the tracks was louder than the other, or at least sounded that way which may have been an intentional way to skew the results but I have'nt attempted to measure it.

[duff138]

  I listened to it a few times on the computer speakers, then put on some headphones.  Once the phones were on I could tell and picked the correct one.

 

[fiveoclockfriday]

I got it, with some cheap Sony headphones. The telling thing to me was the cymbal type sounds that came in after ~8 seconds or so. Cymbals and high end sounds sound very "swooshy" and fake with lots of compression. The vocals were very similar I thought, through my aforementioned cheap phones.

[sbrtoy]

This test is BAAAAD news for anyone into audio....listening through the typical $15 computer speakers, it is very hard to hear a difference.  This is how we end up with "good enough" sound such as MP3 instead of great sound such as newer high-rez formats....

My advice...don't forward to anyone unless you know they have their PC hooked up to an audio system capable of quality playback (or phones).  On my iPod with a pair of $60 Sennheisers the difference between 128 and 320 is immediately clear, and I have gone to all lossless files despite the sacrifice in storage.

[Tweaker]

I got it, with some cheap Sony headphones. The telling thing to me was the cymbal type sounds that came in after ~8 seconds or so. Cymbals and high end sounds sound very "swooshy" and fake with lots of compression. The vocals were very similar I thought, through my aforementioned cheap phones.

That's what I've noticed, too. I tried some 320kbps Mp3 downloads from amazon.com and every thing sounds really good except cymbals. It's definitely what to concentrate on when comparing.

[yo2tup]

on by cheapo computer speakers, they both sound like 96kbps MP3's.  fail!

[EchiDna]

geez I could pick it pretty easy on my monitors built in speakers...

but then again, I've done this kind of thing before and have compared many many digital file types.

As per the previous *spoiler* the difference is easy to pick in the highs
 

[ToddSTS]

Well, to me, I couldn't tell anything in crap earphones, but when I used my Alessandros, even just plugged in to the laptop I could hear a difference and chose correctly.  But, to be fair, the chosen music isn't as revealing as a drum solo (with cymbals) would be, or even some more complicated orchestral arrangements.  So, if you're going to compare, shouldn't you pick something a bit more "challenging" (to mp3 encoded at 128).  Otherwise, someone listens to this, doesn't hear a difference, encodes their music at 128 then one day realizes half their music sounds like crap after they upgrade their speakers/phones.  That would suck.

Todd

[nathanm]

An interesting test would be to find which bitrate ratios can be correctly identified at near 100% accuracy by all listeners; and then at what point the differences become negligible.  Or which bitrate can be identified as "bad" without any other reference.  Like, could people always identify 64K without comparison to something better?  Would it have to go even lower?  I wouldn't be surprised if you had to go to 8-bit audio in order to achieve 100% accuracy.  Hmmm…

[PaulHilgeman]

I think one of the easiest tracks for comparing digital compression formats is the first 30 seconds of "Hotel California" from the Live Hell Freezes Over album by The Eagles.

If the clapping isnt a dead givaway, the shaker sure is.  I have had non-audiophiles tell me that 128 sounds perfect and 192....320 are all way overkill.  I was able to show them the differences betwen all of the  MP3 formats and then even between 320MP3 and lossless.  Without any leading or baiting, it was even easy for them to hear the difference, and I always use that track, because after hearing it pure and uncolored it sounds awsome, big, like they are in the room with you.... you go back to 128 and it sounds like absolute trash, and almost anyone can hear it.

-Paul

[Tweaker]

I just took the test wearing some modestly priced Yamaha headphones and could easily tell the difference the first time. The first clip I was concentrating on the sound of the tambourine or shaker, whatever that was. Listening to the second clip though, it was obvious that it wasn't just the high frequencies that sounded different. Not subtle at all. The 128 kbps track really sounds lifeless and two dimensional.