AAC and MP3: Space

AAC is about 4 MB and an MP3 file is about 7 MB!

Hope This Helps!!

AAC is about 4 MB and an MP3 file is about 7 MB!

I'm tempted to pick this to pieces. But, let's stick with the facts.

AAC and mp3 take up quite the same space.
If you get different sizes, it means you have them encoded in different bitrates or the tracks aren't at the same length.

An average track of 4min 30sec, bitrate 128kbps, is about 4.2MB in size. And that applies for both formats!

You can change the encoding rate to make the
MP3 files smaller.

In Preferences, Advanced, Importing, import using - MP3
then select the setting to minimize the file size

For Talk Shows, 64kbps stereo to 112 kbps stereo settings work just fine.

I encode most of my speech files at 56kbps mono (the 112 kbps stereo), all the sound is clear.

Speech still sounds ok down to 40 kbps (80 stereo), but you start to blur 'S' sounds below that... (as in Sammy)

Note that if you're downloading MP3s, you may need to reencode them.

Right click on your Library and you will notice the ability to list bit rate and encoding settings, file type and size.

Do that, then take one of your shows and encode it at different levels all the way down to the lowest settings.
(So you make 8 or 10 files of different encoding levels)
Then pick the smallest file that still sounds good, or one setting above that to make sure all the sound is good.

'My victories are few and far between, but they are Glorious!" - Anon.

you possibly encoded the aac as lossless which are much bigger files.

Supposedly you can get the same quality in AAC at a lower bit rate than mp3 (96 aac = 128mp3) which is where the notion that AAC takes up less space comes from

128 aac = 96mp3

Don't you mean the other way round?

sorry yes gpood catch edited

Katie,

Unfortunately, you've been fed some incorrect information. MP3 and AAC are "audio codecs" or compressed audio formats. A 128kps MP3 file will take up exactly the same amount of space as an 128kps AAC version of the same song. The difference is that AAC is a more efficient format so it can pack better sound quality into the same file.

The default setting in iTunes is 128kps AAC, and independent studies have shown that most people can't tell the difference between songs encoded using that and the original CD. I tend to import at 192kps AAC myself.

I'm glad you clarified the "incorrect information" for Katie, as it was bothering me also. Don't you just hate that <g>?

I do want to question, however, the validity of the statement "independent studies have shown that most people can't tell the difference between songs encoded using that and the original CD." With all due respect, I find that hard to believe. Yes, I think I have pretty good ears (and, accordingly, most of my music is encoded either in Lossless or 320 kbps), but I believe that even the most casual listener can discern the audible distinction between an unadulterated wav or aiff file and a compressed version of that file one-tenth as large. Yes, sometimes the distinctions may be subtle--other times, more pronounced. Largely, that's dependent upon the many variables involved. But, by definition of the compression process, there almost has to be an audible difference. The point is whether the listener finds that difference objectionable enough to want to do something about it.

My two cents.

I believe that even the most casual listener can discern the audible distinction between an unadulterated wav or aiff file and a compressed version of that file one-tenth as large.

Well, your belief is wrong. Using a good quality encoder, most people can't tell the difference between the original and the encoded file. Some can, of course.

But, by definition of the compression process, there almost has to be an audible difference.

Ummm... no. The process is called "psychoacoustic encoding". The bits that are removed are supposed to be the bits you cannot hear anyway.

Take two tones. Play one really loud and one really soft. Now, although both are still existing in the reproduced sound, your brain does this trick where it filters out the soft one. You don't hear it at all. The idea is to therefore remove that softer tone, since you can't hear it anyway.

Not being able to tell the difference is called "transparency". That's the point at which you can't tell the difference anymore. You prove this using a double-blind test (mostly people use ABX testing for this).

With a high quality encoder (LAME), MP3 can be transparent for 99%+ of the population at around 200kbps. Naturally this varies depending on the specific audio in question, of course, but 200 seems to be about average. It's generally higher for classical music and lower for rock music.

AAC can be better than MP3, in theory, although there's no actual quantitative data on how much better that I am aware of.

Otto,

If I wanted to test your thesis and try encoding with LAME, how would you suggest I go about it?

Thanks...

Using LAME 3.90.3, try the --alt-preset standard setting. As in:
LAME --alt-preset standard input.wav output.mp3

Using a newer version of LAME (3.96 or later), use the V2 setting, like so:
LAME -V2 input.wav output.mp3

These are generally considered transparent for most music.

320 (aka --alt-preset insane) is overkill for nearly all samples.

Once you have encoded your samples, you can do an ABX test to see if you can reliably tell the difference. There's several ABX testing programs out there. The gist of an ABX test is that you do several trials on two different versions of the same sample. One is labelled A, one is labelled B, and one is chosen randomly to be X. Every trial, you can listen to A, B, and X. Then you choose whether X=A or X=B. Later, rinse, repeat. If you can tell the difference, then at the end, your guesses will be more correct than incorrect. If you can't tell a difference, then your guesses will be no better than 50/50 over many trials. You need at least 8 trials to have any good usable result, but usually you want more to be certain.

Anyway, a proper ABX test proves that you can indeed tell the difference (or not), not how big the difference is.

Trust me,

Listen to any iTunes you buy off the internet,
and a CD version of the same song,
on $300 noise canceling headphones,
and iTunes sound like wet mush...

128 AAC files were fine for old ear bud headphones,
but now with people having 5.1 or better sound systems in their home, and hi-fi headphones, 128 sounds bad, at least to my ears.

Of course, AAC was never designed to Exactly copy CD quality music, it was designed to compress it and still sound good.

If you've been blasting your hearing out listening to whatever, perhaps the hearing loss makes up for the difference?

But if you have healthy ears, and great pro-headphones, iTunes downloads are no match for CD music.

For any quality music, classical, vocal, acoustic guitar, etc. I buy the CDs or Audio DVDs and would recommend using MP3 at 320kbps or even better - Apple Lossless encoding. Much better than iTunes downloads.

iTunes' 128 AAC file format was good for grabbing low quality pop songs, or grabbing a wide collection of songs that are not locally available to you.
(such as 100 different Jazz artists, recordings from the 1940s, or your favorite selection of 1980s one-hit-wonders).

But, with all due respect to Apple, the 128 AAC format has got to go.

Apple can't say it's a bandwidth or server load problem anymore, because they are now pushing 50MB video files with ease.

256 Stereo is what iTunes AAC should be set at for modern 2005 type equipment. Afterall, making the songs bigger will sell more 60 GB iPods! And for Shuffle users the software can down convert it to fit the tiny 1GB Shuffle, and for the new iPods they can handle the larger music files quite easily.

Anyway, I think iTunes music is long overdue for a musical format upgrade. Considering most home systems now come with laser optical cable hook-ups as standard, and even remote home users are downloading over cable or DSL, Apple has to raise the bar and meet the new standards in technology.

Trust me,

Thanks, but I'll trust actual objective listening test results instead.

Exactly what I said.

Test your music for yourself on high quality equipment.

Within your own limitations of your own hearing, if the lower bit rates sound good to you, then you're good to go.

Sound, and especially music, is a subjective experience.

If you listen to CD or DVD sound on studio quality headphones, and then to the same song from iTunes or Napster or whatever, and you CAN hear the difference, then pick what sounds best to you.