Apple discusses AAC vs. MP3 codecs; 128 kbps AAC as good or better than 160 kbps MP3

Apple has posted an AppleCare Knowledgebase Document that compares the differences in the AAC and MP3 codecs that can be used in iTunes 4.

Yesterday, various discussion boards around the Net featured posts by Windows users who assumed the 128 kbps AAC equalled a 128 kbps MP3 and then denigrated Apple’s new iTunes Music Store accordingly.

Apple says, “AAC-encoded files sound as good as or better than MP3 files encoded at the same or even a higher bit rate. For example, a 128-kilobit-per-second (kbit/s) AAC file should sound as good as or better than a 160 kbit/s MP3 file. Because the bit rate is lower, the AAC file will also be smaller than the MP3 file. AAC files allow you to store the most music on your hard disk or iPod. The High Quality AAC setting creates files that are usually less than 1 MB for each minute of music.”

Link to the document here.


  1. >Yesterday, various discussion boards around the Net
    >featured posts by Windows users who assumed the 128
    >kbps AAC equalled a 128 kbps MP3 and then denigrated
    >Apple’s new iTunes Music Store accordingly.

    This isn’t entirely accurate. Although some posts did confuse 128 kbps, most of the informed discussion revolved around previous–and well peer-reviewed–blind listening tests that showed that Apple’s AAC, and AAC in general, faired rather poorly against the LAME MP3, MP3Pro, and Ogg Vorbis encoders at higher bit rates in creating “CD quality” audio. And it wasn’t necessary “Windows users” lobbing the criticism…it was more accurately members of the discriminating pro-sumer audiophile community. This is the community that many amatuers look to for both equipment and technology advice, the Stereo Review of the internet community, and their semi-expert opinions hold sway with a LOT of non-experts.

    Unfortunately Apple did not, even in this latest KnowledgeBase article, see the need to address this audiophile community concern; and I guarantee you it WILL become a factor in uptake. It is this community that was/is driving both MP3 usage and development. And right now, this community IS NOT buying Apple’s unfounded claims (when compared to the extent of existing listening tests) that AAC is the next best thing since sliced bread. In fact, one respected comparison site listed AAC as a technology “facing an empty grave”. Not an easy reputation to ignore, although Apple is clearly trying to do just that.

    A simpler, and more marketing savvy approach would have been to play the community’s game, releasing samples commonly used in these tests (there exists a “de facto” library of music snippets) encoded by Apple’s AAC encoder of choice. If they had, and if their encoder really IS as good as they say, they could have avoided the 3 to 6 month uphill battle against quality myths I guarantee this service is about to endure. That fact is an an unfortunate shame, only overshadowed by Apple’s complete inability to read this market.

  2. The Apple-provided 128 K AAC files sound substantially better than 128 K Mp3 files, and indeed better than VBR MP3 files with a 128 k minimum (which tend to come in around 140-150k peak bit-rate). The sound is particularly noticeable in the deep bass range, which tends to break up in MP3.

    Ogg Vorbis remains a better sounding file, and I have no opinion on MP3 Pro.

    Regardless, you still can’t get CD audio compressed to a 10:1 ratio without losing some audio fidelity. It just isn’t possible.

  3. Seriously.

    Audiophiles pushing MP3? Apple’s inability to read this market? Are you kidding me?

    Audiophiles do not listen to MP3’s. Moreover, they do not listen to music from computers, because they take issue with the low-quality outputs and possible interference that results from running an audio signal through a computer.

    The autiodphile market will play NO role in this music distribution revolution, just as they played no role in the MP3 revolution or the CD revolution.

  4. I am somewhat of an audiophile. I installed a 6oo Watt Rockford Fosgate system myself in my brand new 2002 4Runner back in Decemer. I designed the sytem to provide very accurate sound, with careful attention to low bass as well as mids and highs. The system is very good with a little deficiency in the mid-bass range. Last night I “purchased” the new Godsmack album from Apple’s music service and burned it to a CD. Today I listened to it on the way to work and was very impressed with the quality of the sound. The sound is vastly superior to most of the MP3’s that I’ve downloaded using Limewire. There is the occasional exception, but most of the MP3s out there are poorly encoded. Whether it be due to superior technology or better encoding, these new AAC files from Apple simply rock.

  5. The important thing is that AAC is part of the MPEG4 standard (as you all know). I dont’t know if the other formats you are talking about are open or not. I just think that it’s better to go with an open standard, even if it is’nt the best (it seldom is). Apple is pushing MPEG4, and I think it’s a good thing, even if it isn’t the best.

    I’ll be too drunk to hear the difference anyway ” width=”19″ height=”19″ alt=”grin” style=”border:0;” />

  6. Is there a web-link to this “audio-phile” test?<P>
    It also depends on whether it is with current AAC. In my own experience, I’ve listened to 90k AAC and it is much better than 128k MP3. No opionion on OGG, but OGG doesn’t support protection. Which this service probably requires.<P>
    I’ve seen some older blind studies that showed the audio in MPEG 4 was not as good on some audio mixes–but again, that test was 2 years ago… does anyone know if this current AAC has been refined?

  7. Franklin: My information was not dated, early feedback on the “improvements” in 6.2 are already starting to filter in; slashdot already posted one such, limited, spectrum analysis (note that dramatic high freq roll off in AAC). Further, my point was that Apple made no effort to engage the “2 year old” notion that AAC quality was bad…and that is dumb. They supposedly “fixed” it, but STILL did not decide to spent the relatively little effort to show they had. The burden of proof, at this stage, is on them.

    Any one in this forum who is declaring either a) audiophiles have no interest in quality, portable, compressed music or b) that 128AAC is acceptable (compared to MP3) obviously deserve the crap quality this service is selling. And I’d hedge another bet that those same folks are fools (who will gladly fork over $1 for substandard music) or are going to steal songs via P2P anyhow and shouldn’t even be giving their opinion.

    Recent LAME & Ogg Vorbis encoders (at higher bitrate VBR) are both more consistent and of higher quality than Apple’s 6.1 AAC encoder…the jury is still out on 6.2, but from initial reports, some of the basic problems (that would completely eliminate the chance of it besting either of the competitors) remain. Audiophiles who accept and APPRECIATE the benefits/tradeoffs of compressed audio know this. Audiophiles who don’t appreciate the benefits of compressed audio, well, they just don’t care and won’t be buying from Apple anyhow.

    As for links to these tests, do a Google search. I’m not here to do your homework. They are very available and well peer-reviewed. There may not be complete consensus (mostly because the targets are moving), but there is enough information. The time for debating the double-blind testing is over, unless Apple of course would like to enter into the fray and do the work necessary to reopen the discussion. But again, that was much my point…they haven’t…Steve simply decreed “it was so”. It isn’t. There is even considerable debate going on at the Apple Forums over the quality of the music…so it doesn’t hit much closer to home.

  8. Eric, the somewhat audiophile…I also am a Rockford man. I have had two vehicles in amateur circuit competition (no bogus SPL, only sound quality). I also have a very nice home setup, including some very good Grado and Sennheiser cans. The reason that the Apple AAC music sounds good compared to what you are getting off Limewire is the source. Many times those songs are ripped from analog sources (many Windows PCs didn’t even support direct digital ripping until rather recently) and typically use poor encoding methods–like 128kbps CBR Fraunhofer. If you are truly interested in archivable audio quality, you go lossless (like FLAC). However, some of us audiophiles are willing to forgo a bit of quality in order to save shelf space, for the convenience, for the portability, etc. and there DOES exist a true “CD quality” lossy encoder. Check out LAME, for MP3 players, or if you have the luxury (meaning you have a player that will decode them), Ogg Vorbis or MP3Pro. There is a difference. Whereas I can personally pick out the difference (double blind) between AAC (Apple downloaded songs) and CD audio almost every time, a good LAME 160 VBR high quality or Ogg Vorbis will usually stump me. And the file size is typically less than 50% larger than AAC (versus 4 times larger for CD audio). Not bad for archival quality music. And in the case of Ogg Vorbis, they have done quite a bit of work towards facilitating downsampling–taking an archived file at VBR around 180 and making a VBR 96 that sounds nearly as good as a first generation VBR 96 from it. It is THAT particulur feature that many audiophiles are most interested in. Apple could have spent some development effort to help that work or even to roll their FairPlay DRM in, but instead made dismissive remarks about open source projects at the roll out. I would recommend you look into the LAME and Ogg projects…you might find them appealing.

  9. How well do MP3s fare when subjected “Convert Selection to AAC” option in iTunes “Advanced” menu? To my ears, there is a loss at the high end for songs which have been converted in this manner. Would songs ripped directly to AAC format sound better?

    FWIW, my test song was “What’s New” by Linda Ronstadt, from the album _’Round Midnight_. The strings at the beginning seem muddier to me.

  10. If you are so called Audiophile, dont use the service.
    Simple as that. That is what the record store is for. But for rest of the population, 128bit AAC is 100% acceptable. Most of the people will be listening to them on iPod while commuting or working out, and these are not the most ideal situation for high fidelity music.

    Apple does not need to do any advertising to clear up anything, all it will do is confuse the masses. And quite frankly, they don’t need to justify this service to .1% of the people who give a rats ass about Ogg Vobis, LAME, or AAC or MP3 or Super MP3. What they need to emphasize is ease of use, and high quality of music, and selections. That is all, and if you “audiophiles” aren’t impressed, no one cares. Go back to your little Record player and play your precious music the way you want to hear it.
    No one care about your vast and highly tuned ears. Stop being pretentious dorks.

    As far as the Windows people commenting on this, How do you comment on thing that they don’t have any experience on. That just goes to show ignorance.

    There is high-end car audio, but a car/truck/suv/van is no place to do critical listening. The shape of the interior, the materials, and background noise make the whole concept past a point an exercise in futility. In fact, most auto DSP/EQ units DISTORT the audio to compensate for the background noise and image distortion caused by the interior. High-end audio in a car needs only a reserve of power to prevent clipping, and a level of THD of your average mass-market home stereo. The rest is about bragging rights (mine’s bigger/better than yours). To evaluate the iTunes music service in light of “audiophile” standards is an apples vs oranges (no pun intended) comparison. Besides, your hearing peaks in your early 20’s and then declines, especially in the higher frequency range. For most of us, that was some time ago..

  12. Whether or not the different formats and encoders vary in quality (which I’m sure they do), it seems to me that AAC existed with DRM capabilities, while having an improvement over MP3, most people’s reference. Therefore, it was less work for Apple to implement AAC in iTunes Music Store than have to shoehorn DRM into OGG or MP3Pro or what-have-you. I’m sure over time the quality of their AAC encoder will improve, just as the MP3 encoder has, and really, the majority of people who download songs over P2P I’m sure are not that discriminating on quality, almost by definition.

Reader Feedback

This site uses Akismet to reduce spam. Learn how your comment data is processed.