Copy link to clipboard
Copied
I was recording about a 20 minute track, usually speaking for 5 seconds, then stopping and hearing my clips over again to check I sound good. This is significant because I know that my microphone was working and audition got a clean copy of my audio, at least initially. I did a lot of edits on the spot too, meaning I cut and silenced lots of audio, along with saving my progress every hour or so.
Then I went from the start to look for any final edits I could make when I realized some parts of my audio was randomly silenced, with a spectral view that looks like this:
In addition to the silencing some parts of my audio got distorted too, like turning what was once a four second clip of me saying something into me repeating the first second of that clip four times (and losing the rest). Again I need to reiterate that there was a point in time where I played back my audio after recording and made sure that it sounded ok so I don't think it's a hardware problem.
Unfortunately I don't think there's a way I can restore my audio since all the data is destructive, but does anyone know what caused this and how I could prevent it from ever happening again? It seems hugely inconvinient.
Copy link to clipboard
Copied
This is not unknown, and is one of the downsides of recording in Waveform view. You have to consider what happens when you do that; basically your recording isn't going directly to an editable file, but to a temporary file. Every time you do an edit on this recording, it creates a new version of it in another temp file, and nothing gets saved to the final file until you save that version specifically; until that point your file is extremely vulnerable. Why so? Because by default, the temp file location is in a space that's shared by the OS, which has priority on the space. And unless you've taken steps to prevent it, it can alter the size of the file it's using in that folder for its own purposes.
I've said it before, and I'll say it again - recording in Waveform view is not a good idea, and certainly not when there's a far safer alternative available in Multitrack view. Why is it safer? Because any recording made in Multitrack view is written straight to the final file - no vulnerable intermediate file that you can't get at, but the OS potentally can. If you are recording something that simply can't be retaken (live public performances, etc) then it would be an extreme folly to record it in Waveform view - you're taking a real risk.
Other things you can do: Go into Preferences and select a different location for your primary temp file. Create a folder in a completely different location (better still on a drive that doesn't contain the OS if you can) and use that. Also it's safer to make a series of recordings and not edit them as you go - that way you don't keep rewriting the temp file and the available space isn't being filled anywhere near capacity - which it might be if you are editing hour-long takes on the fly.
The temp file system is used for recording because it gives you the ability to do multiple undos - but that comes at a cost. Multitrack recording appears on the face of it to do something similar, but actually it's not. Every ime you hit the record button, you start a new file. Even in punch-in mode - they all get kept separately. Many people regard this as something of a pain, but once you get the workflow sorted out, it's fine. And way safer, because you have to take deliberate steps to destroy things.