views:

298

answers:

4

Like it is the case with jpeg images for example? What's about mp3 or ogg vorbis audio files?

+8  A: 

It all depends on the definition of "loading" and "saving".

If the program in question actually decodes mp3/ogg upon loading and encodes in back when saving (this could be detected by noticeable time it takes to save a file), then yes - you could experience degradation in quality. Especially if the encoding uses different settings (lower bitrate, etc).

On the other hand, a loseless mp3 splitter, for example, could take mp3 and split it into several smaller mp3's without re-encoding them.

So you need to consult a manual of the software in question, there is no generic answer.

ADEpt
A: 

Any time you use a lossy compression, you're going to lose some quality. If you're lucky, most of what you lose the second time was already lost the first time.

If you're curious, try it yourself and see - take the same mp3 file and load and save it repeatedly, then compare to the original. I'm sure you'll know soon enough.

Mark Ransom
+2  A: 

If you're using a lossy codec (ogg, mp3, and jpeg, among others, are all lossy) and you're recoding the data when saving, then yes, there will, over time (or rather with repeated recoding), be a loss in quality. If you're simply "copying" files, then you're getting an exact copy of the lossy coded data, and thus it should be the same as the source file. If you're concerned about the loss of audio fidelity, consider a lossless format like FLAC.

theraccoonbear
A: 

Any time you use a lossy compression, you're going to lose some quality. If you're lucky, most of what you lose the second time was already lost the first time.

That was my assumption. The settings are the same every time. So if high frequencies are removed for example, it should make no difference at all if the algorithm tries to remove them a second time. However, there is more to audio compression than just removing frequency ranges and that's why I ask this question.

And yes, the material is re-encoded every time.

Roger Ween