digg

Join DiggAbout

Discover the best of the web!
Learn more about Digg by taking the tour.

Can you tell 128kbps AAC from the original? Take the test!

duxlist.com:81 — Apple just announced songs at 256kbps, but I wondered how many people could even tell 128kbps encodes from the originals. I whipped up this little app in the hopes of finding out just how good the average Digger's ears are, or how good Apple's encoder is, depending on how you look at it. So get out those Sennheisers and listen carefully!

Bury
show profanity settings
expand all
  • dadof2, on 10/12/2007, -14/+88This is awesome I got all 4 songs right! If you have good speakers/headphones it should be easily apparent. Great info!
    • slightlyoffbeat, on 10/12/2007, -18/+147Unfortunately, most people will probably be trying to compare the two song files by listening via their generic ipod ear buds, or cheap computer speakers. I used my studio headphones and the difference was easily noticeable.
    • aussieNickuss, on 10/12/2007, -10/+23I guessed the wrong one but (i may have forgotten which one was which) but one of them had the faint "hiss" in the background (like off old cassette tapes) that gives me the ***** about the songs I buy from iTunes.
    • Sanitarium, on 10/12/2007, -30/+11Did they download the 256kbps AAC file, and then compress it to 128kbps? Or did they have the lossless file and encode both from the same file?

      Lossy -> Lossy encoding should be a crime.
    • nullcodes, on 10/12/2007, -8/+56This test is not very helpful in determining if AAC sounds different than the original (which was the stated goal).
      Reason being, what if AAC makes the original sound better? That is, the artifacts of compression actually improve the sound (for example by making an instrument sound smoother).

      Looking at the results .. it seems this was the case for Crossfade - Starless ... since the "incorrect" recording was considered better.

      I'll change a quote from Mark Twain so it seems appropriate: "Crossfade's music is a lot worse than it sounds"
    • donelson, on 10/12/2007, -6/+22When I decided to buy a good set of speakers (36 years ago, still have them, Braun 810s, then $400 each), I took three LPs with me to Tech Hi-Fi in Cambridge, Mass, to have a listen: Two piano (Beethoven, Chopin) and one Pink Floyd.

      By far, the most demanding type of music for speakers is Classical Piano.

      ( Yes, I got all tests right using my Shure E3 in-ear phones )

      I also did tests when I first got iTunes, and decided AAC at 192 kbps was the best compormise for iPod battery life and storage space. (Yes, higher data rates burn more battery power due to more frequent trips to the hard disk to play them)

    • .Steven, on 10/12/2007, -14/+2I'm using crappy logitech headphones, because I my Bang&Olufsen are for my ipod. Still I got 3 correct!
    • zolf, on 10/12/2007, -1/+12nullcode: The objective of test is not to tell "which one is better" but the ability to distinguish them. If the probability of selecting the real one is equal or close to 0.5 then they are "the same" (random) for the user.

      Search for "ABX testing" or even for "winabx" software if you want to prove your ears/headphones are superior to others.
    • T0PS3O, on 10/12/2007, -27/+2@ Donelson - With iPods being Flash memory, is your point about reduced battery life due to increased disk reads even valid?
    • Smills, on 10/12/2007, -12/+3I listened to the songs on my Z-5500s and the difference was obvious.
    • Dygear, on 10/12/2007, -8/+1I would like to point out that for some stupid reason iTunes did not seem to always play the files A B, A B, A B, A B. It would stand to reason that these results might just be a little skewed.
    • ciram, on 10/12/2007, -13/+7The one with the bigger filesize is the original.
    • 0o0Moylan0o0, on 10/12/2007, -1/+9I dont suppose someone has a mirror of the files?
    • joeydoo, on 10/12/2007, -0/+7I got one correct. I did 2 with the "24bit crystalizer" from my sound card on.. and two with it off. The one I got correct was with it off (the third one)...... what this means, I have no idea. My speakers aren't too bad.

      ... I'm going to go to guess that I got one correct from chance and I'm actually tone deaf.
    • omatsei, on 10/12/2007, -22/+130I hate to ask (because I have a feeling I'll be the victim of a severe digg-beating), but if the point of this exercise is to prove that you can tell a difference, what does that accomplish? If you can, good for you. Do you intend to wear the giant-sized $300 headphones every time you're walking around listening to your ipod? Are you planning to spend thousands upon thousands of dollars to upgrade your car stereo system so you can actually hear the difference in there too? If you can't, then are you an inferior person, lacking the ability to divine even the simplest of things from 2 vastly different encodes? If by chance, you're talking to someone who can't tell the difference, would you ridicule them for not being able, or would you somehow try to explain how it's different, and why that matters?

      Sorry if these questions seem a little too ... practical for your tastes.
    • AJH16, on 10/12/2007, -10/+3I got them all, but then again, I have tested myself to see the difference between 320 and lossless and I can tell the difference there too. I guess that is 9 years of professional audio experience at work. (I can also tell the difference between CD "lossless" and live through a sound board, but that is a really big difference.)
    • tuzziel, on 10/12/2007, -9/+18The point the guy was making is that about 50% people get it wrong, that means that 50% people are unable to recognize the original, thats significant find and congrats to Apple for such a good codec.
    • samadam, on 10/12/2007, -14/+5@ omatsei: yep. I leave my sennheisers at home and take my also three hundred dollar Shure IEMs out with my iPod.
    • SVPirate, on 10/12/2007, -5/+8I'm at work so I can't take the test, but I know regarding my own collection where I have bought a 128kbps AAC from iTMS then bought the album on CD and ripped it at higher quality I have definitely noticed a difference in sharpness and depth of the sound. Of course I do listen via a good quality 1990s Japanese made Amp and 2 huge Wharfedale speakers that are easily older than I am. In the digital age, there's still no substitute for good analogue equipment!
    • Verdanic, on 10/12/2007, -6/+6@omatsei

      Good point, but I actually can't stand headphones like Apple's earbuds unless they're all I can wear in the situation. Usually I have my massive Sony MDR-Z700's on. Before that, my MDR-V300's. I don't see an issue in walking around with large headphones on. They fold right up and go in a bag when not in use, and when they are, well they just send the message that you care about sound quality which I really do not mind sending at all.
    • gr3yn3t, on 10/12/2007, -6/+4For anyone looking for good (cheap) earbuds, check out the JVC Marshmallows. They really do sound fantastic

    • pkulak, on 10/12/2007, -0/+9"This test is not very helpful in determining if AAC sounds different than the original (which was the stated goal).
      Reason being, what if AAC makes the original sound better? That is, the artifacts of compression actually improve the sound (for example by making an instrument sound smoother)."

      You're totally right, and at this point I'm really ***** that I didn't make it a true ABX test. I was a bit pressed for time though, and I cut corners. If I had known it was going to the front page, I would have probably spent a week on it instead of a couple hours.
    • darkamster07, on 10/12/2007, -5/+2got my Sennheisers (EH-150s), the problem is that I am at school and it was a little hard to listen to the tracks, but I got most of them right. One of the things that really ticks me off about the ITMS is the low quality, I hate having to pirate songs that I have already purchased just to get 320 kbps; although I guess it is not a problem for most people, who just use the ***** earbuds packaged with the iPod.
    • joeydoo, on 10/12/2007, -2/+30"The point the guy was making is that about 50% people get it wrong, that means that 50% people are unable to recognize the original,"

      No it doesn't. If they can't tell which is which then how are the people who don't know... going to pick the incorrect one??

      The flaw here is that it's a 50:50 question. If you don't know then are just as likely to get it right.... which is why the results have all come out 50:50 across the board. It's a completely meaningless poll.
      Every single person could have been unable to distinguish the difference ... picked at random, and the results we are seeing would be exactly the same.
      It's quite bizarre actually.
    • gotamd, on 10/12/2007, -0/+6I got it right, but it wasn't obvious to me. Of course, I was using a $50 Sennheiser headset and had a lot of background noise from traffic too. As other people have said, if nobody here could tell the difference you would expect results of about 50:50...which is about what they are for all the different samples. I think that this is pretty conclusive evidence that most people cannot tell the difference under normal listening conditions.
    • Buchardt, on 10/12/2007, -0/+10I've listened to the songs now, and I couldn't hear a damn difference, but again, my stereo setup now sounds like ***** (lame Linux drivers?). Anyways, when you guys say the the difference is obvious, I don't understand it. What do I need to listen to, to hear the difference?
      Distortion?
      Noise?
      Flat/dull tones?

      Although not working 24/7 with music, I'm rather interested in music, and I do play a lot of music myself aswell. But partially my music setup right now (I just installed Linux, and there are LOADS of background noise in my speakers, dunno why) sux, and partially I just don't know what to listen to. Maybe if I brought the music to some other DAC, I could hear a difference, but right now I don't know what to listen to.

      So all you "it's easy to hear" people, please enlighten us and tell us WHAT exactly is easy to hear.
    • wiihuck, on 10/12/2007, -1/+3i got them all right, but i'm not going to lie: i was kinda guessing. plus i was listening with my macbook's speakers.

      this "test" was interesting to me, because (like so many others i'm sure) i'm a live music collector, and i'm pretty anal about maintaining lossless in the community. so....it's making me rethink that whole deal...
    • izomiac, on 10/12/2007, -4/+0@omatsei

      Bulky $300 headphones? I happen to have a pair of $160 in-the-ear headphones. They aren't any bigger than standard ear buds, have a 20 dB passive noise reduction, and have amazing quality (for me at least). Very well worth the money given how much I use them, although I did get them for Christmas a couple years ago. So it's very possible that one could tell the difference in a normal situation, though it's just a matter of personal preference what quality you listen to.

      What I'd be concerned about (if I owned an iPod or used iTunes) is your ability to reencode your songs in a couple years. After all, you don't expect Apple to remain in the music industry forever, or if they do that they'll never replace AAC, right? I also hear that iPods don't last forever either, so what happens when you no longer have a player that understands AAC? The way I see it you have three choices: give up listening to your old music, repurchase your music collection, or reencode it to a newer format. Quality is critical for the latter. If it's in a lossless format then you don't have anything to worry about. In any other case I'd be hesitant to buy music that is encoded at anything near what you can notice.
    • fireaway, on 10/12/2007, -7/+6This makes me really sad. We keep losing more and more fidelity. I wish I could invite those who question vinyl over to my house to do a head to head versus CD and MP3 on an excellent stereo systerm....Not only would you choose vinyl everytime, but you would feel a three-dimensional lushness to it, a richness that digital doesn't even come close to. Yeah, you may hear a crackle or hiss....occasionally...but you would also feel the drums in your chest, the strings would be full and deep and have levels that make the digital sample feel like it's written on white lined paper for school children...and not really music, the way it should be heard....These compressed samples cut out even more of the sound. OW...No wonder so many people think rap is music.
    • insovietrussia, on 10/12/2007, -0/+12The points is, as far as Apple is concerned, the vast majority of their customers are going to be listening to the music on their iPods, on their computers and in their cars - where the difference is indistinguishable/negligible. This is the market they are trying to reach, not the audiophiles.

      If you need higher quality sampling, then the iTunes store probably isn't the place for you.
    • Azurensis, on 10/12/2007, -1/+14>The flaw here is that it's a 50:50 question. If you don't know then are just as likely to get it right.... which is why the results have all come out 50:50 across the board. It's a completely meaningless poll.

      No, if there *was* something about one sample that made it sound better to the average person, it would have showed on this poll. As it is, AAC at 128k is good enough for most people.
    • afwjam, on 10/12/2007, -0/+3@fireaway

      While I totally agree with you vinyl sounds better, its because of the warm mids not the highs and lows. CD has always sounded better to me on the highs and lows. But a majority of the music is in the mids where vinyl excells and this is why it sounds better. Or atleast that seems to be my findings.
    • randomgeek, on 10/12/2007, -0/+10It's kinda funny. Every single time I pick the compressed version. I'm not listening to the highs, to the lows, and trying to figure out which one reproduces these ranges perfectly, all I'm doing is listening to the clips and picking which one sounds the best to me. It's interesting that I've been listening to Mp3s since about '94, that's 13 years of hearing mostly compressed music. I wonder if "what sounds better" to me is simply just picking what I'm used to?

      In any case, the stated goal of the experiment - can you tell the difference - seems to have been verified, even with my strange results. Evidently I -am- hearing a difference because I pick the same one every time. I'm just not sure what it means.

      (for the record, I'm using over the ear semi-expensive headphones)
    • ipodsweatshop, on 10/12/2007, -6/+0Yeah, most people can't tell because they don't REALLY listen to music. Just look at what music actually sells.

      Also, the push for AAC is because it supports DRM. MP3 and OGG are just as good or better depending on your preference, produce smaller files, and support no DRM.
    • prockcore, on 10/12/2007, -0/+1There is a test you can do to check to see if they're recognizably different.

      You have 3 copies of the same song. 2 of them are AACs 1 is original, or 2 of them are original, and 1 is the AAC. Then you have to pick which of the two sound the same.
    • dggeek, on 10/12/2007, -7/+1Hmm, I installed the latest Quicktime on my machine so I could listen to the .m4a files and now I have a Quicktime that won't open and errors out when trying to uninstall.

      I love Apple!
    • SuperSnake2012, on 10/12/2007, -2/+2On the Crossfade song, it's easy to tell which is the original. When the vocals kick in and it's just the bass soloed, it doesn't sound as good in the 128kbps track. As a bassist my ears are sensitive to these kind of things!
    • goatman93, on 10/12/2007, -0/+2I could not tell the difference through my Creative Speakers or my set of Ultimate Ears. I think my hearing sucks.
    • yanked, on 10/12/2007, -0/+1Got the first one wrong, then I realized what to listen for -- the hiss of static, and extra crackle in the high stuff. But even speaking as a lover of HD (and someone who can't understand the people who claim not to be able to tell the difference between 480, 720, or 1080), I'm not really sure keeping that noise is actually an improvement in all cases here: a couple of the files that were "cleaned up" by the compression sounded a bit better to me. But the one of Elliot Smith (close-miked vocals) and the Liszt (piano) really do sound richer in the high-bitrate version, even with the cost of a slight hiss.
    • thegenome, on 10/12/2007, -2/+1yeah... which ever one downloads faster...
    • meshman, on 10/12/2007, -4/+6"thats significant find and congrats to Apple for such a good codec"

      Oh fer... Apple didn't make the codec. As shocking as it may seem to some, Apple didn't invent "everything".
    • nullcodes, on 10/12/2007, -0/+1@zolf

      I wasn't clear in what my explanation. The problem is that the page asks the user to "select the original" ... Let's say the original sounds like crap to you but the AAC copy sounds great to you.. then you may choose the AAC (thinking it's the original). Someone else may think the original sounds better than the AAC and choose the original (correctly).

      So there ..the AAC and the original sound different but can produce a 50/50 survey result.
    • thoughtfulclown, on 10/12/2007, -1/+2This is really more of a test of how good your ears are, not just how good the codec is.

      I'm a sound technician, and I also help train others to become sound techs as well and more than anything else its the ear of the individual that determines how good they are. Often we get rid of people just because they don't have the ear for it. Digg probably has a higher percentage of audiophiles than normal, and so the results of this test are probably skewed. Probably about 80% of the population wouldn't be able to tell a difference - especially on their average sound setups. I tried it using my cheap earbuds and could still tell a huge difference on all four.

      All in all I think by encoding at 128kbps Apple is just ignoring the 10-20% who can tell a difference. Personally, I'd love to see them start encoding at 192kbps. I'd even pay more for it.
    • Kratisto, on 10/12/2007, -0/+1I have ***** headphones, so there wasn't a difference for me. I should really get new ones.
    • cancelorallow, on 10/12/2007, -0/+0@nullcodes
      I found your post interesting. I thought it was ridiculous at first but I have a similar comparison.
      I have a Pioneer Elite LD/DVD player, a sony Wega XBR tv. Right about the time I got the Wega DVDs were just becoming popular. I had 100s of LDs and thought nothing could have a better picture. But interestingly enough the DVDs seemed to have richer colors and better contrast. Although the LDs had no lossy compression (they're analog) the digitally compressed DVDs put them to shame. But there are a few movies that do better on LD than DVD, maybe because the DVD encoding was bad or just impossible. I cannot watch Last of the Mahicans on DVD while having a superior copy on LD.

    • erikerikerik, on 10/12/2007, -0/+2@! azurensis
      you're correct.
      their is no point of reference. first the original uncompressed version should be given.
      THEN let people figure out the watered down flavor.
    • LogicBomB, on 10/12/2007, -0/+1@omatsei:

      It doesn't matter if you listen to good quality on ***** speakers because you expect *****. But if you listen ***** quality on excellent speakers it matters to some (like me). I could care less if I can't tell the difference driving in a car with a ***** sound system but when I bring that ipod home and hook it into my Paradigm 60's speaker system or bring it on the bus with my Senn's it can make a difference.
    • ericnmu, on 10/12/2007, -0/+1welp, I can't tell
    • Sabin, on 10/12/2007, -0/+1Very easy to tell the difference on my Senn HD595s and my Grado SR80s.
    • coldfusion055, on 10/12/2007, -0/+1My Shure E4c's sniff out the bad tracks with ease.
    • rovertly, on 10/12/2007, -2/+1The download time gives it away.
    • wmarcello, on 10/12/2007, -0/+1I managed 4/4 listening through average quality headphones, but it wasn't very easy. I had to listen to each file 2-3 times to be able to tell much of a difference. To a casual listener this would be near impossible, so kudos to Apple for providing good compression with reasonable file size for their songs.

      I find the difference between 128kbps MP3 and lossless to be much greater than what I heard here. If ACC files were compatible with as many players (most notably my car stereo) out there as MP3 I'd probably go with that, but since it's far from, I usually rip my cds to 160kbps MP3. I haven't done any blind tests to see if I can tell the difference, but I know for a fact that I can listen to my music without cringing.
    • wmarcello, on 10/12/2007, -0/+1Ok, I half-retract my statements. I'm currently comparing some CD lossless files to my 160kbps MP3's, and boy is there a noticeable difference, even moreso than with the 128kbps AAC files on this site. Granted they've been encoded with iTunes, but I was hoping for better. The biggest flaw seems to be keeping the stereo integrity. With the MP3's it seems like a lot of sounds wander from left to right more than they should.

      It makes me wish that some of these newer codecs were a lot more universally accepted in players... or maybe I just wish I had unlimited hard drive space to rip at any quality I want.
    • Dog_Paddle, on 10/12/2007, -0/+1I just bought $160 Sony ear cupphone or whatever that style is called. I was hoping that I would be getting much better sound on my iPod than with the headphones that came with my XM Radio

      The thing is, I can't tell a damn difference, and I've gotten all of these wrong. Either Sony sucks and I wasted my money, or my ears suck and I wasted my money =(

      I guess that's what I get for an impulse purchase at Best Buy
    • steepdecline, on 10/12/2007, -0/+14/4 with my Shure 5c's
      actually quite simple
  • pkulak, on 10/12/2007, -1/+21Here are the totals so far:

    http://duxlist.com:81/song/show_results

    It's not statistically relevant yet, but if this makes it to the front page it will be.
    • klawz, on 10/12/2007, -3/+7Yes, and so far, as you can see, most people can tell the difference in the more modern music, than so in the older music.
    • zolf, on 10/12/2007, -1/+21And statistically they are correct. If you choose randomly from two you should get around 50%.
    • Goosemaster, on 10/12/2007, -0/+3http://forums.anandtech.com/messageview.aspx?catid=38&threadid=2027423&enterthread=y

      here's another blind test but using mp3

      Basically, I found the same results that you did....
    • neuropsychguy, on 10/12/2007, -0/+4"And statistically they are correct. If you choose randomly from two you should get around 50%."

      Yep, people are only at chance level (i.e., guessing); so no, people (on average) can't tell a difference.
    • Jugalator, on 10/12/2007, -1/+4"And statistically they are correct."

      Huh? *Who* are correct? Apple, in that they're saying there's no special audible difference? Yes, it would seem so with an even distribution.
    • Mirag3, on 10/12/2007, -0/+1hmmm i got every single one wrong... which hypothetically means i can tell the difference right? I actually noticed that i often chose the one that sounded louder during bass changes, which i thought was supposed to indicate compression, but its actually the opposite, so i guess its not that i cant tell the difference.
  • TheCheeks, on 10/12/2007, -2/+152/4. If i tried hard enough and long enough i'm sure i could get better results, but honestly just getting the songs and listening for enjoyment, i would not have noticed a difference.
    • TheCheeks, on 10/12/2007, -0/+6i even busted out my Senn HD555's. i was pretty amazed how good a 128 AAC sounded too.
    • newslang, on 10/12/2007, -12/+4they were 256k and uncompressed, not 128k.
    • pkulak, on 10/12/2007, -0/+5"they were 256k and uncompressed, not 128k."

      No, I'm just an idiot, sorry.
  • Shaman760, on 10/12/2007, -7/+27I really miss the "vinyl hiss"..... ffffff-----pt ffffffffffff--------pt....So warm, so analog.
    • tempusrob, on 10/12/2007, -6/+26Yeah, nothing like *noise* to really enhance your listening experience.

      I know there are a lot of vinyl proponents out there and that's fine, but WTF?
    • CoBLeviathan, on 10/12/2007, -4/+7@Tempusrob

      I guess it's too early for sarcasm. Shame, really.
    • catalysis, on 10/12/2007, -4/+13I can't tell if he's being sarcastic. There are people who swear by vinyl and I don't get it either.
    • heyitsmedusty, on 10/12/2007, -9/+3You're thinking of older, low-quality analog setups. A high-quality turntable (~$200 and up) with a good cartridge and a clean record will reproduce music just as clearly as a CD, and some people do prefer the sound that it makes. The noise you're referring to is either from dirt and dust on the record or an old record that has been played so many times with bad cartridges that the grooves are distorted.

      You'd be surprised at the level of information that is contained on a record. Whereas a CD has to be "sampled" at various frequencies, a record groove is linear information. It's trickier to retrieve that information,
    • Azurensis, on 10/12/2007, -2/+6>A high-quality turntable (~$200 and up) with a good cartridge and a clean record will reproduce music just as clearly as a CD

      Show me any turntable that has less wow and flutter than even a cheap cd player and then we'll start worrying about sampling errors.
    • aaronm67, on 10/12/2007, -1/+4@heyitsmedusty

      It doesn't really matter how high quality the turn table, the fact is there are still flaws in analog storage/reading.
    • heyitsmedusty, on 10/12/2007, -0/+2You don't measure wow and flutter in a CD player, so I think the point you're trying to make is that CDs must be better since it's an unmeasurable quantity in digital playback. What I'm saying is that good turntables with quality vinyl minimize the wow and flutter to a point that isn't perceivable to the human ear.
    • meepus, on 10/12/2007, -1/+6@heyitsmedusty

      And a good CD player is how much cheaper than a good turntable? Getting CDs in good playing condition is how much cheaper than getting records in good condition? Who cares if it's physically possible for a record to sound as good as a CD--the value is not proportionate to the actual cost in such a way that it'd make sense to use it unless you were a DJ and wanted to scratch or something. Even then, you can do all that fun DJ crap digitally with more effects now... Really, there's no logical argument for vinyl anymore outside of the fact that you're just a fan of it because of nostalgia and stuff.
    • heyitsmedusty, on 10/12/2007, -1/+2@meepus -- I would never try to make the argument that vinyl playback can be done as well as CD playback for the same price. That's why I specified that good vinyl playback can be done, but at a higher price. All I was saying is that it's possible to get a record to sound as clear as a CD, which is true. And since it's not sampled at different frequencies then you can actually get more information out of a record than a CD if you have good enough equipment (which, again, is very expensive). On the other hand, technologies like SACD and DVDA have very closely approached the amount of information on vinyl, which means with the right equipment can sound about as good as a record. The sound that you get has a different timbre to it than digital playback, which many people prefer. If you ever have the chance to hear a high-end vinyl system, I highly recommend it. It's hard to believe any of this until you've heard it.
    • rimco, on 10/12/2007, -2/+4Kids, the point he's making is that a record is a perfect recreation of an analog signal, not a digital approximation like a CD is... the sound is far more open and real sounding... once your ears are used to that kind of realism, CDs will forever sound flat and like a tin-can.
    • tempusrob, on 10/12/2007, -1/+2"a record is a perfect recreation of an analog signal,"

      Then how do you explain the RIAA equalization curve, and other various EQ curves used on records? If it was a perfect recreation then we could use a linear amplifier on vinyl rather than a specialized one.
  • ilolled, on 10/12/2007, -4/+1Yeah, thats fine for a very simple song. There wasn't a great deal going on there.

    Try the test again with some more complicated music and i bet you'll notice the difference.
    • insovietrussia, on 10/12/2007, -0/+2Try some of the other samples, there is a good mix in there. They've obviously tried to choose songs that have a high production value.
  • nerdjesus1, on 10/12/2007, -24/+16I normally convert all my songs to 96kbps to save tons of space. I don't notice the difference between that and the original unless I try, but when I enjoy it, it's practically the same thing.
    • pathy, on 10/12/2007, -6/+6596kbps? Your poor ears.
    • TheCheeks, on 10/12/2007, -5/+3796? ow... i can tell a 128 from 192-320, i cant imagine how a 96 would sound...
    • sicapitan, on 10/12/2007, -8/+34you probably have a black and white tv and a 386SX!
    • imeddy, on 10/12/2007, -19/+6Imho under 224kbps, almost everything starts to sound crappy.
    • kaffein, on 10/12/2007, -2/+13You don't "convert" you re-encode an already encoded file... That would sound horrible on a good set of cans or monitors.
    • jupi, on 10/12/2007, -14/+3obviously your not a true music lover
    • spliffy, on 10/12/2007, -12/+2no one who is that unversed in digital audio encoding could ever have any capacity for music appreciation.
    • Obsidian743, on 10/12/2007, -16/+42You're all just a bunch of pretentious, uppity audiophile wannabes that annoy the ***** out of most people.
    • DigitalJester, on 10/12/2007, -9/+2I feel sorry for you. All my music is at least 192. I go for 320 whenever possible.
    • insovietrussia, on 10/12/2007, -1/+16I remember when I bought my Rio Mp3 player for AU$700 in 1999. Had 32MB of memory which held 10 songs, but you could fit 11 if you encoded at 96kbps. Ah those were the days.
    • Muncher, on 10/12/2007, -0/+4"96? ow... i can tell a 128 from 192-320, i cant imagine how a 96 would sound..."

      You know, you could always try it... I use 96k Vorbis on my portable and though it is far from transparent, it doesn't sound *bad* to my ears.
    • aaronm67, on 10/12/2007, -1/+5It depends on what you're using your music player for. If you use it to walk to class, using crappy earbuds or whatever, 96kb/s is probably fine for that.

      For me, I often use my mp3 player in my car, which as a fairly decent sound system in it, so I usually try for 256kb/s or higher (if it's available), and I usually wont even bother if its below 192, because it doesn't sound nearly as good,

      Not everyone is a "pretentious bastard", they probably have a good reason for wanting their music that quality.
  • CopyNinja, on 10/12/2007, -4/+1I got it! ohh yes.
  • razorgator, on 10/12/2007, -8/+84/4
    Dugg for Elliot Smith.
  • Dawson1066, on 10/12/2007, -2/+9The test uses 256k AAC files--not 128k. Of course it will be tough to tell the difference on computer speakers.
    • johnstar, on 10/12/2007, -9/+3I have a klipsch promedia 5.1 as my computer speakers :P
    • eMximeR, on 10/12/2007, -1/+1Put on headphones?
    • maebmij, on 10/12/2007, -0/+5How can you tell? The site says the compressed versions were 128kbps...
    • jimmytango829, on 10/12/2007, -4/+0actually, can someone explain the difference between 128k aac files and Apple Lossless? It says it was compressed with the Apple Lossless encoder...how is that 128k aac if it's lossless? Technically speaking, the lossless compressor should produce a smaller file that render IDENTICAL sound to the original.
    • td4guy, on 10/12/2007, -0/+4It was a 128kbps aac file reencoded into lossless format, so that you couldn't just cheat and look at the bitrate. The lossless encoder utilizes a different algorithm from the lossy aac encoder, so "technically", it shouldn't be an identical file size.
    • wolfkeeper, on 10/12/2007, -5/+1If this genuinely was a lossless format, there shouldn't be any difference at all. A true lossless format should reproduce the original sound down to the bit level.

      In practice they sound slightly different, so I'm unclear what's really happening.
  • zenmeyang, on 10/12/2007, -0/+5it was hard but I was able to distinguish some difference. I wouldn't be able to tell if I wasn't paying attention.
  • Unicast, on 10/12/2007, -1/+3i could tell the difference only by the echo, in the original it seems more lengthy
  • fireball74, on 10/12/2007, -57/+4I tested myself one better.... I only played the first 10 seconds of each clip and still got every answer right.

    Not sure what it proves except that maybe I'm a good listener of music. And for what it's worth, I am a former musician.
    • conradpaul, on 10/12/2007, -4/+33douchebag
    • logic6, on 10/12/2007, -2/+41I only read the first 10 words of your comment and knew you were an idiot.
    • rovertly, on 10/12/2007, -3/+4Indeed, you are a god among music listeners.
  • slightlyoffbeat, on 10/12/2007, -1/+26I have a lot of 96k and 128k music from my napster days, and from the days when I thought that was best to rip cds at lower quality because it saved hard drive space. My listening tastes have matured greatly since then and I can't stand to listen to those old music files anymore. They sound muddy. I suppose if you can't tell the difference then it doesn't matter, but if you CAN tell the difference then low quality music files can really ruin the listening experience.
    • vuzman, on 10/12/2007, -0/+30128 kbps is not necessarily the same as 128 kbps.... My point is that the encoder, the encoding method and the source also matter, and greatly so. I recently had a "spring cleaning" in my music files, and found several mp3 files @320 kbps that sounded like they were recorded through a telephone. I also found a LOT of semi-high bitrate mp3 files (160-192 kbps) that sounded worse than a properly encoded 96 kbps mp3 file would.

      Newer technologies, like AAC, can easily sound better than mp3 at half the bitrate. Use a good mp3 encoder (like LAME) and the quality markedly improves though.
  • scrytch, on 10/12/2007, -15/+3Problem is - if people just watch how long it takes to download each file, they can cheat as the 256k file is slower to download.

    Sorry... open to cheating, so not statistically sound.

    Regards,
    Shane.
    • scrytch, on 10/12/2007, -13/+1Who is negative digging my comment? It's valid if people are going to rely on this information to show that the way it is being obtained is open to fraudulent behaviour.

      I didn't cheat (I got 2/4), but that doesn't mean everyone won't.

      Regards,
      Shane.
    • AggieTales, on 10/12/2007, -0/+8Actually... no, both files are the same size(at least on The Shins and Franz Liszt), and both are encoded in Apple Lossless (from the 265 kbps, so its lossy -> "lossless" encoding) so there's no "smaller" file. the only slight difference I noticed was on one of the songs the "Data Rate", as Quicktime puts it, differed by 0.04 kbps (790.62 vs 790.58)
    • scrytch, on 10/12/2007, -3/+9My bad - apologies...
    • illt, on 10/12/2007, -0/+2Problem is, most of these tracks to aren't very demanding on compression, and the case with crossfade, is it's a pretty poorly produced track to begin with.
  • shableep, on 10/12/2007, -0/+10crap, 128kbps AAC sounds really freakin' good. with MP3 i can totally hear the difference between 128kbps and 196kbps. i heard almost no difference between the 2 tracks on all except the Shins. for some reason the Shins one sounded a lot more obvious. maybe try throwing in an orchestral track. with the wide range of frequencies in orchestral tracks, it might more distinctly show the difference between the too rates. would be cool to see.
    • lolgamoff, on 10/12/2007, -3/+3128kbps AAC is nice, but Apple Lossless is quite amazing. You can really tell it's better than any lossy encoded song.It's what i rip my music with. However, I'm happy with 256kbps AAC from the iTunes store.
  • sn0wkitty, on 10/12/2007, -1/+10I got it wrong, however I would argue that the difference between the two would be much easier to discern if you were using an example with a fuller frequency spread, such as a rock or dance track (with higher high frequencies and lower basses.)

    Don't quote me on this, but isn't the "Squishing" that occurs when you compress audio most noticeable in the high treble and low bass frequencies? If that is the case, a midrange-centred piano wouldn't be a good basis for comparison.

    Or maybe I'm just having a hard time admitting I got it wrong.. heh heh heh

    edit: oh, wait, there are more examples on the page. oops
  • mal1964, on 10/12/2007, -3/+1its at 50% 50%
  • ShinGnosis, on 10/12/2007, -4/+8I listened with expensive sennheisers (HD590) and very good audio card (audigy2). I couldn't hear any notable difference that made it certain that one was compressed and the other wasn't. Great job apple!

    I think it's obvious that most other people can't tell the difference either - the scores are almost 50/50 for all the music pieces. People just guessed one or the other in the end.

    Maybe with more complicated music, deep bass, and so there will be differences with (very) high end speakers but overall I think apple did a great job here!
  • Stonekeeper, on 10/12/2007, -4/+1that was really easy through a pair of BayerDynamic DT250.
  • donelson, on 10/12/2007, -1/+2It's actually amazing how good the compression is... You really have to listen hard on most earphones. Only really good earphones give you the edge. (Got all four right)
  • snek, on 10/12/2007, -0/+3I wonder what hydrogenaudio users would have to say about this test :)
  • blocguy, on 10/12/2007, -0/+3Was hard to tell a difference with my Sennheiser HD 515s.
    But to be completely honest, he could have picked a more complicated song, e.g. most modern songs today. I'm pretty sure that if you chose a song with guitar, drums, bass and vocals then the difference would be much more noticeable and the results would have been much more relevant to the widest audience.
  • Zaphodi, on 10/12/2007, -0/+9192 seems to be the limit of my ears. After that i don't hear any difference to the original. Seems to be the exact same thing as CRT refresh rates. some people are happy with 60Hz and some people see the screen flicker even at 85Hz
    • jeriqo, on 10/12/2007, -0/+3And same with FPS.
      I don't want 5000x3000 HD TV, I want more than 25ips
  • mal1964, on 10/12/2007, -0/+4the stats tell me that you can toss a coin
  • gback2000, on 10/12/2007, -0/+4I am 0 for 4. I guess I can save 30% per song by not buying the 256kps versions from I-tunes.
    • Mirag3, on 10/12/2007, -2/+1actually, you can tell the difference - 0 for 4 is just as unlikely as 4 for 4 - you just like compressed better :)
  • benskycouk, on 10/12/2007, -1/+5i couldnt get it to work, file A was broken? :S
  • gemadouble, on 10/12/2007, -0/+1It's noticeable on my monitors, but not on my headphones.

    I convert my files to 96kbps wma for my PMP and it sounds good, any lower than that than its horrible.
  • ahmerhussain, on 10/12/2007, -0/+3I wish that they put more than just 2 possible responses.

    If they could get it up to at least 4 or 5, then it could become much more than a guessing game fo most people.
  • ashmodai, on 10/12/2007, -1/+18That, that's nothing, in my day we were listening to 16 kbps and we LIKED it!
    • insovietrussia, on 10/12/2007, -0/+4What about 8bit chiptunes? Now that's hardcore.
  • BillDoE, on 10/12/2007, -1/+6The poll itself is kind of misleading. None of them are "original" 1411kbps wav format. Crossfade - Starless , the only one with more wrong votes, just sounded like crap period(I heard crackling and popping on the one you call good). I got 3 out of 4 with my old Audigy2 card with $100 5.1 Creative pc speaker system. On a decent sound system it would be even easier to tell.

    Mp3 or whatever lossy compression IMO is Temp portable music. If you like it, buy it on a disc. As long as it's DRM free ;)
    • logic6, on 10/12/2007, -0/+5@BillDoE

      "The poll itself is kind of misleading. None of them are "original" 1411kbps wav format"

      Word

      I personally got all of them wrong, using earbuds, and never having heard any of the songs before.

      The problem with compression artifacts is you can't tell if they are from the original recording or if they are from the compression. A better test would be to compare the DAT / WAV to the AAC in two compressions. Otherwise, you are just showing a preference for a certain sound.

      Remember, in science you need a control (uncompressed audio).
    • logic6, on 10/12/2007, -2/+1oops
    • maebmij, on 10/12/2007, -2/+1According to the site, the files are in Apple Lossless format, which uncompresses back to a binary-perfect copy of the original 1411kbps file.
  • kolais, on 10/12/2007, -1/+4Assuming that the website creator is the one who submitted this to Digg, this looks more like a propaganda :) Here, the test has been described as the AAC 128kbps VS the original, while the site states that the comparison is AAC 256kbps vs the original. And judging from the comments above, I see that people seem to miss that point. Buried as inacurate.
    • pkulak, on 10/12/2007, -0/+5Propaganda for what? Give me some slack man, I mistyped. I fixed it though, and thank you for pointing it out. I wrote that thing in a couple hours, and as you can see, I didn't put a whole lot of time into presentation. :D If it didn't even make it to the front page and no-one used it, I didn't want to have wasted a bunch of time.
  • Kelmon, on 10/12/2007, -1/+5I don't think I can tell the difference but the time it takes to download the files was a dead give-away. I've been downloading tracks from the iTS for about a year now and I've never had cause to complain about the audio quality so I just don't think my sense of hearing is very good - either that or those who complain about 128kbps encoding are doing so just for the sake of it. Anyway, impending higher quality audio is nice but I just don't think it will make me enjoy the music any more than I already do.
  • choopie911, on 10/12/2007, -6/+1I got 4/4
  • smackit, on 10/12/2007, -3/+42
    I used my [insert expensive audio equipment here] to listen and got (lie) correct! I must be an audio genius!!!
    • blork1, on 10/12/2007, -2/+4Honestly, who cares what bit rate your audio is? Surely the song itself matters far more, and whether you enjoy it.

      /whilst listening to MP3@320kbps
  • xmuzik, on 10/12/2007, -0/+6These debates always remind me of the software vs. hardware debates on music production forums, the sad truth that most of the heads wont admit is that most people in the general buying public never notice the difference.
  • likephatec, on 10/12/2007, -1/+30I was so good at this test the CIA and the FBI both showed up at my door and asked me to meet the president.
    Luckily, since I am a former compositor and an accomplished Yoyo-Master I head their footsteps from 2.554 miles away, and ran away in time, to my endless fields of douchebag trees and assclown bushes.
  • foxhoundadmin, on 10/12/2007, -1/+1lol. i had JUST listened to both of the files and was going back to re-listen to the first (to better compare), but then the bandwidth usage had probably been exceeded, and i couldnt! :(
  • cuoops, on 10/12/2007, -5/+4I rip all my cd's into wav files so they can be converted down to whatever I want.
    • LeonardNimrod, on 10/12/2007, -2/+9[sarcasm]Yeah! That makes sense![/sarcasm]

      If you are storing audio digitally you should be using a compressed, but lossless audio format such as FLAC or Apple Lossless.

    • dave932932, on 10/12/2007, -2/+3Why is he or she being dugg down? If you rip .WAVs, you can keep them on a extra hard drive and rip to whatever format takes the lead in the future.
  • OGTL, on 10/12/2007, -21/+2This deserves a hearty LOL, I have a $2200 headphone setup and with the AKG K701's. I can tell the difference if a song is 1kbs lower than another. (Exaggeration, but 128 VS 256 is no contest)

    Foobar, ASIO4ALL, and FLAC is the only way to listen.
    • TriaLunae, on 10/12/2007, -1/+13... and you deserve a hearty STFU.

      Where the hell is your dedicated CD transport and external DAC? Your tube-based headphone amp? Oh, right, because Foobar and ASIO4ALL (read: half-ass ASIO support for consumer hardware that does a fine job of soaking up all that lovely EMI bouncing around the inside of your computer case) are more than equivalent replacements to support your $2200 headphone setup. Or not.

      Ass.
  • Swift2, on 10/12/2007, -0/+5I couldn't get this to download. Digg Effect.

    But I did take the test with the PC World clips -- Mozart and Michael Stipe. Isn't it clear to everyone? Listen, in the Mozart, to the strings. In the rock cut, listen to the background.
  • mal1964, on 10/12/2007, -0/+0using these two words together "fraudulent behavior" is a automatic one point deduction
  • stevenlowry, on 10/12/2007, -2/+1dugg mirror they got it just in time well www.duggtrends.com
  • mathoov, on 10/12/2007, -0/+3"This file is currently inaccessible. Contact the file's owner for details."
    • pkulak, on 10/12/2007, -0/+3Sorry about that. This went to the front page while I was sleeping. It should be fine now. It really is tough to prepare for the Digg effect. :D
  • troon, on 10/12/2007, -0/+2Probability 101: here's what you'd expect if the files are truly indistinguishable:

    Each song will have a 50/50 split of votes.
    1/16th of respondents will get all of them wrong.
    1/4 will get one right.
    3/8 will get two right.
    1/4 will get three right.
    1/16th will get all right.

    Even if you got them all right, you can't draw any conclusions from that, as the only result of any value is the overall per-song results, once enough votes are cast. If you want to test yourself, you'll need to do some more rigorous ABX testing.
  • ReallyChili, on 10/12/2007, -7/+3FLAC FTW
    • KibibyteBrain, on 10/12/2007, -1/+9...if you live in the magical land of infinite storage space, especially on mobile devices.
  • Zammie, on 10/12/2007, -0/+2I hear the difference thru my iPod plugged into my Bose player.
    However, thru these garbage speakers on this garbage Lenovo, nothing has changed.
  • Mark0Pon, on 10/12/2007, -0/+2Here you can find some well-tought and well-conducted test results about audio coders:
    http://www.rjamorim.com/test/

    Roberto is the Roberto Amorim from http://www.rarewares.org

    Bye!
    • Goosemaster, on 10/12/2007, -0/+1that's acutally a good presentation...we just need more bitrates....
  • Whiskeynose, on 10/12/2007, -1/+3My headphone sounds bigger than yours
  • scottjl, on 10/12/2007, -1/+2i got all 4 wrong! i can save money and buy cheaper itunes! yay me!
  • corde5, on 10/12/2007, -0/+1Clicked the audio links: "Doh! You Received the Wrong Link!" Hmmmm
  • Fetching more discussions...
    Show 51 - 100 of 157 discussions
  •  

    Login or register to add a comment

    Create a new account or login to join in the conversation on Digg. You'll also be able to Digg stories to help promote things you like.


© Digg Inc. 2008 — Content posted by Digg users is dedicated to the public domain.
DIGG, DIGG IT, DUGG, DIGG THIS, Digg graphics, logos, designs, page headers, button icons, scripts, and other service names are the trademarks of Digg Inc.