Question regarding converting AAC to AAC
Feb 14, 2009 at 5:34 AM Thread Starter Post #1 of 7

nway

Head-Fier
Joined
Feb 14, 2009
Posts
76
Likes
17
Ok, this is a puzzling question, but for those sharp enough, bear with me.

AAC carries greater quality than mp3 with the same bitrate and size.

So if you have 254kbps mp3, and you convert it to 224kpbs AAC, it's essentially the same quality.

However, once a file is AAC, you seemingly lose this benefit when you convert it to a lower-bitrate AAC, right?

To make it clearer, I'll describe my particular situation:

Throughout the past year or so, I've been converting mp3s into AACs of equal or slightly less bitrate, so as to not lose any quality. In other words, If I had a 237kbps mp3, I'd convert it to a 224kbps AAC. If I had a 192kbps mp3, I'd convert it to a 192kbps AAC.

However, now I'm thinking all I really need are 128kbps AACs. BUT, my entire music library is already in AAC, not mp3. So, I'd be converting from 160+kbps AAC down to 128kbps AAC.

My question, then, is that if I'm already STARTING with, say, a 192kbps AAC, and I convert it to a 128kbps AAC, will this 128kbps AAC still have the same quality as a 192kbps mp3, which is what it would have had if it had been converted from an mp3?

I guess the fundamental, implied question I'm asking is:

If I converted all my mp3s into equivalent or slightly less AACs, and then converted all those AACs into 128kbps AACs, would those last files be of lesser quality than had I just converted all my mp3s directly into 128kbps AACs?

I realize this problem is rather hard to grasp, but if you understand what I'm saying, PLEASE help.
smily_headphones1.gif


Thanks in advance!
 
Feb 14, 2009 at 7:27 AM Post #2 of 7
Yeah, never encode a lossy file with a lossy codec a second time.
Lossy encoders (like AAC, MP3, Ogg Vorbis, ...) run the data through a psychoacoustic model during encoding, throwing away "not audible" audio data. When you use that file as source for a second lossy encoding it will run the data through a psychoacoustic model again, throwing away "not audible" audio data.

You are much better off re-ripping straight to the preferred codec and bitrate.
 
Feb 14, 2009 at 7:33 AM Post #3 of 7
Mm, that's what I thought. Unfortunately, I've got thousands of tracks, all manually labeled, and I couldn't possibly redo all this. Especially the classical works - ugh, relabeling all those would be a nightmare. -_-

How significant would the loss be? After all, is there not a reason why the audio data is regarded as "not audible" in the first place?
 
Feb 14, 2009 at 7:48 AM Post #4 of 7
192 mp3 -> 224 AAC has less quality than source -> 192 AAC. So you would be decreasing the quality even more.


As to how significant the loss, try a few tracks and listen. In all likelihood you will notice a difference, the question is if it would adversely affect your experience.


For tagging ripped tracks check out MusicBrainz Picard. It should be able to automatically, and correctly, tag your tracks to your specifications.
 
Feb 14, 2009 at 7:54 AM Post #5 of 7
Quote:

Originally Posted by nway /img/forum/go_quote.gif
Mm, that's what I thought. Unfortunately, I've got thousands of tracks, all manually labeled, and I couldn't possibly redo all this. Especially the classical works - ugh, relabeling all those would be a nightmare. -_-


Yeah, its a lot of work.
You should have seek help one (1) year ago before you began transcoding your MP3's to AAC.
wink.gif


Quote:

How significant would the loss be? After all, is there not a reason why the audio data is regarded as "not audible" in the first place?


Only you can tell!
Cause we all have different set of ears and different gear. Hence the difference may be very audible to one, while transparent to another.
 
Feb 14, 2009 at 8:01 AM Post #6 of 7
Quote:

Originally Posted by ThePredator /img/forum/go_quote.gif
192 mp3 -> 224 AAC has less quality than source -> 192 AAC. So you would be decreasing the quality even more.


As to how significant the loss, try a few tracks and listen. In all likelihood you will notice a difference, the question is if it would adversely affect your experience.


For tagging ripped tracks check out MusicBrainz Picard. It should be able to automatically, and correctly, tag your tracks to your specifications.



Well, I'm talking more along the lines of:

192 mp3 -> 128 AAC vs. 192 mp3 -> 192 AAC -> 128 AAC

Basically, how much worse the latter 128 AAC would be than the former. I'll definitely try out some listening tests tomorrow, though my equipment isn't very good, and indeed the whole reason why I kept high-bitrate AAC's was for when I get better sound equipment in the future. The reason why I may not be able to distinguish between the two in a listening test might have more to do with my equipment than my ears. :X

And...I am extremely anal about how my music library is organized. This may sound psycho, but I never, ever let an automatic tagger touch my babies. Plus, I developed my own labeling system that doesn't really fit the protocol of "mass tagging".

Quote:

Yeah, its a lot of work.
You should have seek help one (1) year ago before you began transcoding your MP3's to AAC.
wink.gif


Indeed. I have too many regrets in life.
frown.gif
 
Feb 14, 2009 at 10:02 AM Post #7 of 7
All you have to do is to encode some sample tracks, from the same WAV file ripper from CD. Example:
1. Direct to 192kbps AAC
2. Direct to 192kbps MP3
3. Direct to 128kbps AAC
4. Direct to 128kbps MP3
5. First to 192kbps MP3, then transcode to 192kbps AAC
6. First to 192kbps MP3, then transcode to 128kbps AAC

Then perform a listening test, to verify if there are an audible difference. Using your ears and your gear.
wink.gif
 

Users who are viewing this thread

Back
Top