Let's start with the JPEG. The artifacts in a JPEG file are predictable. Consequently it is possible to mitigate them. Unfortunately the same thing is not possible with audio at all, because all artifacts are different, and so there is no 'correct' reference point. Basically, distortion cannot be fixed, and won't ever be. People sometimes ask the same thing about MP3 files, but the situation with those is worse; information is actually discarded because the stuff you are not supposed to hear gets 'masked'. In reality what that means is that it's encoded at a much lower bit depth, so there's no 'correct' reference point for restoring it.
And unfortunately a lot of that also relates to the horrible noises you get from Skype and Zoom - there's no reference point to use as a basis for restoring any form of normality, because the data is either missing or corrupted. But that does give a clue as to what to do, strangely enough. For the audio you care about, get the person at the other end to make a local recording using a decent mic at the same time as the Skype/Zoom conversation, and independently send you a copy of the file. In Multitrack view it's then a relatively straightforward process to replace the horrible audio with the local recording. We've tried this and with care it can work remarkably well - it's like your remote contributor is in the room with you!
1
Reply
AdChoices