capmorgan wrote:
Hello all,
I'm trying to understand something about compression.
I know that CD tracks are pretty big files and when they get imported into iTunes (with AAC compression) the tracks get "shrunk" to pretty small file sizes.
From what I read, compression is achieved by eliminating "unnecessary" bits of an audio track. To my mind if certain portions of an audio track are eliminated they're gone forever .. right?
----------
That's right.
----------
However, when I create an audio CD using AAC tracks from iTunes, on the CD the tracks become full size files again.
So, my question is if compression was achieved by eliminating portions of the audio track how come when the same audio track is burnt onto a CD, it becomes a full size file again?
----------
A compressed file can be expanded to _simulate_ the file before it was compressed--except for.....(read on). For example, an mp3 file can be "converted" to a much larger wav file. Note the quotation marks!
----------
Do the eliminated bits somehow come back to life in the CD burning process?
----------
NO, NO, NO, a thousand times NO!
----------
Isn't compression an irreversible action?
----------
YES, with the exception of a special type of compression called "lossless" compression. Lossless compression is analogous to .zip compression, which elminates dead space but does not remove any information from the file. Typically, lossless compression of a sound file leads to a file which is about one-half the size of the original file, where as a "lossy" (normal) compression might shrink the file down to one-tenth (or even greater) of its original size. Two common types of lossless compression are FLAC and APE. Although I have used FLAC (a free, open-source compression method) in the past, I am switching to APE (a free, privately developed encoding method), since APE has more flexibility and can compress a file in much less time than is required to compress a file using FLAC. To checkout APE, go to www.monkeysaudio.com. I am really impressed with this lossless codec, which is freely available for download with no strings attached--except you will not get the source code.
The major use for lossless compression is in archival storage, because the file can be restored to its original size (as a wav file) without any loss in quality. This pristine file can then be edited, etc., without the loss of quality that takes place when a file compressed in a lossy manner is edited (see below). This is a BIG advantage of shrinking a file for storage purposes using a lossless compression method. I use it regularly for archival storage of all of my music files, even though I store these files in mp3 form for _playing_ them in a media player.
Lossy compression removes information which can be removed without a terrible degradation in the sound quality. The higher the compression, however, the more the quality is degraded. Remember; once those "extraneous" bits are removed, they are gone forever, and converting the file back to a "pseudo-wav" file will not get them back. If you recompress this pseudo-wav file, you will just degrade it more.
Furthermore, any time that you edit a compressed file with an editor such as WavePad, the file is first decompressed to a pseudo-wav file, without telling you that this is what is happening. So you can edit as you see fit and then save the edited pieces in the same format as was the file before you loaded it into the editor, BUT these edited pieces will be of a more degraded sound quality because they actually have been compressed TWICE by the time that the leave the editor.
----------
Is all of the above specific to AAC and CDA or is that how compression works in general - ie unnecessary bits are not really eliminated, they're just kinda hidden?
----------
They are not "just kinda hidden". They are GONE. Sorry about that.
----------
This brings me to another question - when an AAC track is played back - am I listening to the compressed version or the full version?
----------
By "full" version, I assume that you are talking about what you get when you "convert" a lossy, compressed file back to the original (pseudo) file. Except that it is _not_ the original file, since you have lost essential information.
In other words, you are listening to an audio file in which certain "dispensable" pieces of information have been removed--unless the amount of compression used is so great that these removed pieces of information have become INdispensable (to _your_ ears) . Get the picture?
----------
I've read quite a few articles on compression but they are all so technical and talk about the algorithms of compression (whatever the heck that means)
----------
Compression (when talking about a sound file) means just what it says: squeezing down the file to make it smaller, whether or not the squeezing loses any information that needs to be recovered to maintain the sound quality (i.e., lossy versus lossless).
If your question is about the meaning of an algorithm, this is simply a series of steps used to produce some particular result--such as compressing a sound file, for example.
----------
but I just can't seem to find an article that ties together in layman terms compression, playback and "uncompression" if you will.
----------
No I won't :-) It's "decompression".
----------
Any light shed on this will be greatly appreciated.
----------
You've sure come to the light place!
Musikone