Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: iTunes AAC VBR vs true VBR (Read 30291 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

iTunes AAC VBR vs true VBR

Okay, I have a mac and upgraded to Leopard. I notice the update in afconvert, and you can convert using a true VBR mode. iTunes uses VBR_Constrained instead.

In order to compare these to see which setting would be the best choice, I converted to both.

iTunes 256kbps VBR (max vbr setting itunes allows) outputs a 4.8MB file with 256kbps bit rate.
afconvert VBR setting at max quality outputs a 2.6MB file with a 153kbps total bit rate.

The difference in the size of these 2 files leads me to believe that this vbr_constrained setting that iTunes uses is FAR from true VBR. Im assuming that these 2 files have similar quality, but the vbr_constrained slightly better but with much less compression obviously - but I can't test because I don't know of any abx software for mac.

In comparison to a true vbr mp3 at highest quality (v0), the mp3 was 3.1MB at 148kbps total bit rate. Nothing too shocking, the AAC equivalent is smaller in size and higher quality. What surprises me is this setting that Apple labels as VBR when it is clearly not, this setting is pretty much useless.

Here are the comparisons:

MP3 V0 (vbr): 3.1MB; 148kbps
AAC VBR 127: 2.6MB; 153kbps
AAC VBR_Constrained 256k: 4.8MB; 256kbps

This is the syntax I used to create the vbr aac file:

afconvert -f m4af -d aac -s 3 -u vbrq 127 track01.wav

iTunes AAC VBR vs true VBR

Reply #1 –
Quote from: _banana on
MP3 V0 (vbr): 3.1MB; 148kbps
AAC VBR 127: 2.6MB; 153kbps
AAC VBR_Constrained 256k: 4.8MB; 256kbps



Sorry to say, but the test you concluded is both, innacurate, and uninformative. Let me explain:


3.1MB / 148kbps = 175 seconds . Did you mean 3.1MB / 184kbps = 141 seconds?
2.6MB / 153kbps = 142 seconds
4.8MB / 256kbps = 157 seconds . Did you mean 4.3MB / 256kbps = 141 seconds?

Note that: 2.6MB < 3.1MB, but 153kbps > 148kbps


Next, you seem to ignore what VBR_Constrained is, let me enlight:

AAC modes (at least in Quicktime/iTunes):

CBR: keep a constant bitrate over a slice of time, varying the bitrate within a small margin from the target.
ABR: keep a constant bitrate over a slice of time, varying the bitrate within a bigger margin from the target.
Constrained_VBR: Start at a given bitrate, and allow to *increase* it if needed by the complexity.
True VBR: Allow variation of bitrate (i.e. allow to increase or decrease, depending on the complexity)

What i mean with this is even if you want to encode pure silence (ok, haven't tested this exact case, but the point remains) with Constrained_VBR at 256kbps, chances are high that you'll get a 256kbps file.

The real point in testing iTunes/Quicktime encoder right now is to find out both, the transparency level of the True VBR setting, as well as find out if it is stable, or the quality varies significantly with the same setting.


About ABX software... the JAVA ABR/HR program should run in Mac OSX  (I guess there's java in it)

http://www.rarewares.org/files/others/abchr-java-0.5b.zip

iTunes AAC VBR vs true VBR

Reply #2 –
Why even have constrained vbr ?
Is it some 'security' for an unproven or buggy true vbr ? What is it with apple anyway ?

I am really surprised. At least mpc, vorbis got it right from the start.

iTunes AAC VBR vs true VBR

Reply #3 –
FYI

If you are using iTunes to display the bitrate of the final file, iTunes masks the real average bitrate and instead displays the target bitrate that you set.

In Leopard with iTunes 7.5 it displays the actual bitrate, however.

iTunes AAC VBR vs true VBR

Reply #4 –
Quote
Sorry to say, but the test you concluded is both, innacurate, and uninformative. Let me explain:


3.1MB / 148kbps = 175 seconds . Did you mean 3.1MB / 184kbps = 141 seconds?
2.6MB / 153kbps = 142 seconds
4.8MB / 256kbps = 157 seconds . Did you mean 4.3MB / 256kbps = 141 seconds?


I simply encoded the same wave file 3 different times to compare them. Those are the file sizes and total bit rates recognized by OS X for each file. I know this is NOT the most accurate way of comparing by any means, but I did it to get a general idea on how much is actually "constrained".

I fully realize what Constrained_VBR does, but I said it was pointless as iTunes labels it as just "VBR" when it is actually not. If you are trying to get the best compression possible, you want to use VBR and that is not what it is - very misleading.

iTunes AAC VBR vs true VBR

Reply #5 –
For those running OS X 10.5 or 10.4 w/QT 7.3, XLD has now been updated to offer all four flavours of QT AAC encoding.

iTunes AAC VBR vs true VBR

Reply #6 –
Just wondering if all itunes songs are Constrained_VBR ?

iTunes AAC VBR vs true VBR

Reply #7 –
Quote from: iwod on
Just wondering if all itunes songs are Constrained_VBR ?


I know this thread is almost a month old but I can answer.  All VBR AAC songs encoded using iTunes are constrained, iTunes does not use the true VBR AAC encoding scheme that QuickTime has to offer.  I think the reason is that people would have been really confused.  Imagine someone selecting 128kbps VBR and yet the file came out to have an overall average bitrate of 100kbps or 150kbps.  People would complain that their 128kbps is not 128kbps.  Or someone selecting 320kbps VBR and having the bitrate come out to 180kbps VBR, they would complain about the extra bits that they probably wouldn't even hear anyway.

Remember that iTunes is setup more for the masses.  It was designed for those people who purchase an iPod, install the software, pop in an audio CD, and don't ask any questions.  People who have more knowledge can change the settings around for their liking but iTunes was designed to make everything easy.  I wish that it had an option to pick the true VBR mode but then I am sure many people would get confused about picking a quality level (instead of known bitrate) and hammer Apple will e-mails asking what is the right setting to use.