Slashdot Mirror
AAC vs. OGG vs. MP3
asv108 writes "Yesterday, Apple unveiled their new music service claiming that the AAC format "combines sound quality that rivals CD." Here is a little comparison of lossy music codecs, comparing an Apple ripped AAC file with the commonly used MP3 codec and the increasingly popular OGG codec. Spectrum analysis was used to see which format did the best job of maintaining the shape of the original waveform." Wish they had WMAs in there too. And for the spoilage, it looks like OGG comes out on top.
I've got a nice pair of Bose headphones, and I listened to an Apple Store AAC file and an OGG version of the same song. I don't consider myself a real audiophile, but it's damn near impossible to tell the difference between the two; though I can definitely hear the improvement from MP3 to AAC or OGG.
Some decent quality properly blinded listening tests would be more interesting than a graph though.
When VHS established dominance of the video market, there were high barriers to change - your player and media were committed to that format.
There are far less barriers to change in the ripped audio format, although there will still be some inertia, but there is nothing* to stop ogg vorbis becoming the dominant format.
Where's my ogg pod then?
* apart from the silly name.
Humorous signatures are over-rated.
"the increasingly popular OGG codec."
sadly, I don't think OGG is *currently* known to anybody except nerds or IT pros.
Most people who use ogg do not use it for it's quality. All that matters in that respect is that Ogg is comaprable to other formats at similar bitrates.
The important aspect of it is that it's free. There are no patents (at least as far as we know of) preventing anyone from using it, and it's made quite clear that the code can be included in open and closed source software without royalty payments.
Ogg is a container format. I could in theory put an ACC audio file into an Ogg container.
.ogg because it is inside an Ogg container.
The audio format you're babbling about is Vorbis. Usually
Hell, it's not just a silly name problem, it's an entire naming convention issue.
I don't think graphs are all that useful for comparing lossy sound compression.
Microsoft likes to show how their wma looks better than the other compression methods... it does look beautiful in graphs, but it sounds all tinny and horrible.
I don't care if the compressed frequency response graph looks nothing like the original frequency graph, as long as my ears are unable to tell the difference between the two.
Regarding the AAC->CD->MP3, I burned a couple of Music Store tracks to CD, then re-ripped them (using iTunes, no less) using VBR High, and they sounded indistinguishable from the original Music Store files (albeit being significantly higher average bitrates).
Regarding DRM, it appears that your Music Store file is locked to your Apple ID, and you have to Register up to three computers that you want to be able to play songs associated with your Apple ID. If you sell a computer, you have to unregister it before you can register a replacement computer. This appears to be the only restriction on usage -- you can still burn the songs to as many CDs as you want, copy them to as many iPods as you want, and streamthem to as many other Macs (and TiVo) as you want using Rendezvous.
The Xiph folks have signed up to add Ogg support on the Neuros audio handheld. Its a firmware upgradable handheld which currently supports mp3, but will probably have Ogg support by mid-late summer.
Check out the highlights.
http://www.neurosaudio.com/
MP3 this, OGG that, AAC somewhere in the middle... Sorry, I don't use any of the above. I encode all of my music into Musepack. At high bitrates, it's the best lossy audio codec, period. For more information on Musepack, see Case's Musepack Page</a>, or <a href="http://www.hydrogenaudio.org/index.php?act=S T&f=11&t=1927&">List of Recommended Musepack Settings</a>.
Musepack encoders and decoders are available for both Windows and Linux, with Winamp plugins available. The only real downside to Musepack is there is currently no hardware support. But having tried each of the codecs mentioned in this article as well as Musepack at the Quality 8 setting, Musepack is music to my ears each time.
"We are all in the gutter, but some of us are looking at the stars." - Oscar Wilde
Understandably, most of the discussion here is about the pros & cons of various compression formats. But the first thing that jumped out at me when I clicked on the apple.com link was:
"Preview any song for free, when you find a song you want, buy it for just 99... It's what music lovers have been waiting for: a music store with Apple's legendary ease of use, offering a hassle-free way to preview, buy and download music online quickly and easily."
FINALLY, a business model for downloading music that makes sense! (Now if only I could afford to switch to Apple products.)
let's compare video codec image quality by streaming the data thru a hex editor in realtime. :)
Just raise the taxes on crack.
...and I gotta tell you, having played that trumpet and serving as Music Director for the Celestial Choir since the Dawn of Time, I know Audio, and MP3 is the way to go. I've analyzed OGG, WMV, AAC, and this cute l'il analog thing which that wack job Orpheus put together Back in the Day, and I must say, nothing beats MP3, in your or anyone else's universe.
Of course, I'm logging in here under a pseudonym, so you'll just have to trust me. But hey, would a member of the Heavenly Host lie to you?
Last night I downloaded a bunch of tracks off of Apple's Music Store Service. I then played them (along with several tracks I already had in OGG and mp3) through my computers $9.95 speakers while holding my portable cassette recorder very, very close to the speaker (For the technical out there I was holding it close to the LEFT speaker and even turned the TV down some to get the best possible sound) and then replayed them all back on the same portable cassette recorder.
My conclusion is that all three sound like complete shit.
Appended to the end of comments you post. 120 chars.
AAC/MP3/OGG are all based on psychoacoustic models. Comparing their decoded spectrums is pointless. The spectrum isn't supposed to be faithfully reproduced. Frequencies that your brain wouldn't fully hear aren't fully stored.
The only value I can see in a spectrum comparison would be to find obvious errors in the encoder or decoder. Like the 16kHz spike in the Xing encoder. But how likely is that going to be these days?
The only proper comparison involves a good hi-fi, a sensibly furnished room, and a comfortable chair. It is called "golden ear" testing and it's the ONLY way to compare psychoacoustic models.
Or at least it's the only way until the research scientists work out how the human brain works.
Learn why you shouldn't use spectral analysis to determine lossy codecs' quality.
The most respected technique is double-blind testing using an ABX tool such as PC ABX, WinABX or ABC/HR.
More info on conducting blind tests can be found at the PC ABX site.
Use ISO 8601 dates [YYYY-MM-DD]
Modern compressions schemes are supposed to make sound that sounds as much like the original as possible, not looks like the original on an FFT.
The only way to test this is to use double-blind listening tests. The spectral analysis stuff is absolutely useless for finding out how good the music actually sounds.
autopr0n is like, down and stuff.
At low bitrates, AAC is very weak, at 128kbps it was the worst of all:
Study
I was one of the 3000 participants, btw. And my ranking which I gave (blind, I did not know which sample was which) confirms pretty much the results, at 64kbps, AAC was unbearable, while ogg was not distinguishable (by me anyway) to the original.
The only test where AAC didn't fail miserably was the "expert test" with only 8 listeners.
OGG has beaten all other codecs consitently at all bitrates.
Let's see. Given the task of creating a codec de novo and the financial and political means to have access to the original source material rather than a version sent through a horribly non-linear sampling mechanism out of your control and beyond your specification, which would you choose?
I'm sure most Slashdot readers will be familiar with the Nyquist limit and understand the complete inability to represent information above the limit, but how many are familiar with the degradations that occur near the Nyquist limit when you have non-infinite signal lengths? This is why oversampling is so important. In general, if you have a signal at frequency f that you want to accurately capture, you should be sampling (by rule of thumb) at 5f or greater. If you sample at lower frequencies, the distortions in phase and amplitude are difficult to predict and statistically analyze as they tend to have uniform rather than Gaussian distributions.
So again, I re-pose the rhetorical question: given the task of creating a new codec rather than rewriting an old one, wouldn't you want to use the least-filtered signal possible as a source, especially when the extant filtering is non-linear, and be able to select by design which parts to encode and which parts to ignore? I sure would.
Put my fist through my alarm clock with its ding-dong death inside my ear. - The Blackjacks.
from macslash:
AAC comes with a significantly lower number of b*tching [\.] users than ogg
Spectrum analysis was used to see which format did the best job of maintaining the shape of the original waveform
Will people ever stop doing that. It's complete bullshit and certainly not the way to evaluate a codec. These codecs use perceptual weighting of the noise. That means that the idea is to distort the signal as much as possible in any region of the spectrum where it won't be heard at a certain time. That means that you see a big distortion in the spectrum and think the codec is worse than the others when in fact it's better because it realized that it doesn't matter.
The only way to correctly evaluate a codec is to listen to it. I write codecs (see sig), so I know a bit what I'm talking about. I use spectral analysis sometimes, but only to identify problems which I've already heard before, not to say that my codec is good.
As a aside, I'd say it probably wouldn't be hard to write a codec that does better than any other on those spectrum analysis. They would sound like crap because their psycho-acoustic model would be all wrong.
Opus: the Swiss army knife of audio codec