I'm working on a cooking show and some of the audio sounds clipped and over modulated but when I try and fix it in Audition (please note that this is the only audio editing program I have access to at my place of work) it doesn't recognize that the distortion is there because the audio never goes into the red. Is there a way to fix this? I can't even issolate what's causing the distortion, it happens randomly with no clear trigger.
he original raw audio might be clipped too, though. imho isotope rx does declipping a lot better, but here's the audition version.
tips for getting good audio(if you can't reshoot with expensive sound equipment that is ☺
1. clean up diagnostics audio. de-clip 1st ,de-reverb 2nd, denoise always last, (watch for DC offset errors and phasing errors)
2. equalize, removing wind, rumbling and enhancing vocal range raising 95hz 2.5 dB Q 2 width, and 972 hz 4.5dB Q2 and 1306 Hz 4 dB 2Q
3. studio reverb size 1; decay 200ms;width 19.9;high freq 161hz, low cut 20hz, damp 50%; diff 50%; dry 100%, wet 78.6% to get warm microphone sound.
4. multiband compressor-fiddle with this one's settings so it sounds good to roll off dynamic range loudness
5. match loudness ITU-R BS.1770-3 -24lufs for correct audio levels
Distortion is almost impossible to get rid of unfortunately. It's like once you've over exposed your whites, or under exposed your blacks, you'll never get that missing detail back because there is no information for the editor to work with.
Fortunately Audition has a Declipper as you have stated, but being built in, it might be quite limited in it's applications. The only other Declipper I know of, so far, is the one found in iZotope RX, but that's an expensive piece of software. Even then, there's only so far a declipper can go to repair audio, as it is "guessing" at what the audio might have been like before it was destroyed. The more heavy the distortion, the more information was destroyed, the more the algorithm has to "guess", the less perfect the result.
Although, there is something you could try. It could be that the reason why Audition isn't recognising it is because the levels may have been reduced at some point after being distorted. Audition may rely on the audio being maxed out to recognise the problem.
So, in short, try Normalising the signal then try the Declipper again. See if it will recognise it after that.
Other things you can try is, a reshoot, obviously. You could try to find someone who owns a copy of RX or equivalent. You could consider making a new edit and removing the part with bad audio while covering the edit point with a cutaway or a scene change. It's possible you can steal sound elsewhere in the piece and replace the bad audio with that. Doesn't really work for dialogue though, of course.
Another thing to consider is ADR and just have whoever come to replace bits of dialogue. Then grab bits of ambience from elsewhere in the main audio, run that beneath the ADR. Then do a bit of audio treatment to match the ADR track as closely to the original captured audio.