Timeline examples or learning resources for audio levelling of live event dialogue and audience reactions ? (e.g. stand-up comedy)

Question

For best audio quality from a live audience event, complete with audience reactions (laughter, cheering etc.) I understand that it is necessary to perform audio levelling manually, via clip gain points/nodes/keyframes/envelopes. I realise that automation tools (like RX) can help, but here I am specifically interested in the "hand-tailored" approach.

Although I have some years of experience in doing this, I have always worked alone, and I have always wondered how my approach compares to that of others, indeed whether it could be improved in quality and/or productivity (eg keyframe density). To me this kind of work seems like a fine art, but despite much web (including video) searching I have never found any examples going into detail on real-life timelines with envelope-based levelling, shaping the audience reactions and presenter comebacks so that they sounds realistic etc. Any pointers/tips ?

Answer 1

I have just found a broadly relevant example - though film-set based (hence lacking audience reaction elements) - in one part (as per the start-time in the link) of the following Youtube presentation: Tutorial: Dialogue Signal Processing with Matt Yocum.

Based on this (admittedly singular) example, it appears to be reasonable practice to have as many as one or two level changes per word (e.g. for particularly loud syllables). At least when such a degree of work is justified.

This pattern is consistent with my own clip-gain/envelope/keyframe -based manual levelling practice.

I had (for a long time) been nagged by uncertainty (in my work-isolation) about whether or not my levelling practice had been over-done (excessive keyframe/node densities). It seems not. Consequently this particular uncertainty will nag me no more!

Possibly this (trivial but vital) insight will likewise reassure others in a similar position.

Hopefully I will eventually locate one or more further publicly available examples, this time including the use of levelling edit techniques specific to audience reaction. If/when this happens, I will link to it via a further comment to this comment-thread.