I do it kinda backwards, and export the scratch audio from the camera over to my audio editor (audacity) where I manually sync the good audio to it, & trim to the exact length.
Then I replace the scratch audio with my prepared version.
I'd like to be better at the video side, where I could perhaps add or delete frames, rather than tampering with the audio so much...
Heard Pluraleyes is good for camera sync but never got it to work myself.