Original or Copy

When copying, converting or editing audio, something is always lost despite the use of digital technology.
Does anyone have an idea how to tell which is the original file when you have two audio files?

  1. a broader spectrum
  2. less distortion

Are there any more clues?

Your presumption isn’t correct.

If I copy a file from one folder to another it’ll be identical. Same if I copy/paste an event on the timeline. Editing the second event by cutting off the second half means the first half is still the same as the first half of the “original” that I copied it from, even though one now has been edited.

So it just is not true that “When copying, converting or editing audio, something is always lost”

I didn’t mean that kind of copying either. If it is only about copying the binary number values from one place to another I agree with you. If I would copy a CD, then I would be already not so sure whether on the copy the sectors are again at the exact same place. Also the mp3 format, or just all formats that compress, are changed afterwards. Songs that are distributed on the Internet are usually copied in compressed form. It is these copies that I am concerned with.

Ok, so imagine then that I have an original .wav file at full resolution, let’s say 16 bits 44.1kHz. I then take that and make it into an mp3 at a pretty low rate. Then I bring that back into Cubase or Nuendo and resave it as a .wav file at 16 bits 44.1kHz again. We now have the original “A” and the new file “B” (“B” is not a “copy”, so we should use a different word probably).

If you want to tell the difference between the two then the first option is to just listen. You’d often hear a difference.

Another option would be to look at the file using a spectrogram. It’s possible that the file that went through lossy compression got a bunch of stuff thrown out and that stuff wouldn’t come back. So you could possibly see a lack of high end in that spectrogram when comparing the two files.

It’s also possible that loudness and peaks have changed. As for “distortion” I wouldn’t be entirely certain that you would see that. I think it’s possible, but not necessarily the case.

Normally if I brought that stuff into software to analyze I’d guess that whatever has a more extended frequency response is the original. It’s not always true I think, but most of the time.

Was that what you were looking for?

So basically you are saying after an audio file is processed, the resulting file sounds different. Well that is kind of one of the main reasons folks process audio.

How to keep track of which is what? The best approach imho is to come up with a good naming scheme that keeps track of whatever you think is important to keep track of and always use it.

When moving audio on the Internet if the compression is lossless like flac the resulting audio should be the same as a copy of the original file. But if it is lossey like mp3 they will be different. File size is often a good clue for sorting this out - the smaller the file the greater the loss.

Can you give us an actual use case so we can better understand what you are looking for.

I think your ideas are along the same lines as my thoughts were.
It will probably be like a painting. You can hardly distinguish a good copy. A very good copy might even surpass the original.

I copied your post. Had anything lost?

I don’t understand.
I have not lost anything.

Did my second post answer your original question?

If you still have questions, what are they?

In fact, I’m still thinking about it.
For example, I wonder if there is a way to measure the amount of distortion in a recording, or if I have to rely on my hearing. After all, distortion would be an indication that it must be a copy.

What’s the practical goal of all of this? What’s your specific situation and problem you’re trying to solve?

I have a feeling that you’re not going to get any answers that will satisfy you.