Which first: comping or timing correction?

Question

I have a number of takes (usually 5-8 are usable) of the whole band, all recorded to click that I would like to comp (merge together into one performance). It is no-brainer what to pick if some part is significantly better than in other takes. However, I have problems deciding when some take is slightly better musically, but it has worse timing, while the some other take is the other way around.

Normally, I would say that one should rely on their taste. The issue I have is this: lack of good synchronization might cause inferior choices, on the other hand timing correction affects how we perceive the performance, in particular different parts might require different corrections.

Ideally, I would have time to time-correct every take to match any other given take, but for 8 takes 4 parts each it creates over 200 different tracks to edit for one song (although in reality that would be more like 80-100).

My question is, what is the best approach:

• comp first and synchronize later,
• synchronize first (to one of the takes? to grid?) and then comp?

The music is something between jazz and funk, so even slightly bad timing would instantly kill the groove. Fortunately the musicians were good, so timing within-one-take is great. Unfortunately, I cannot use just one good take (each has some flaws), cannot re-record individual parts and neither can I invest time to do all the 200+ edits per song (lots of edits in those recordings would be rather hard to do seamlessly).

I suspect there is no easy answer here, but I would appreciate any suggestions.

Answer 1

If I were in this situation, I would comp first.

You aren't really in a position to fully appreciate timing issues (or lack thereof) until you hear the performance in context. A musician never plays in perfect time and this is often desirable - otherwise, what would be the use of the "humanize" function?

After the initial comp, listen to the whole track without the click and see how it grooves. As long as there are no major slips at the moment you transition between takes (which you will have some control over in choosing your comp points), any timing issues within the individual takes will simply be the human element of the performance.

The transition points will be the trouble spots, but your correction workload will be reduced dramatically - you only have to make sure the groove transitions smoothly between the takes you have chosen, as opposed to syncing all the takes (which should probably be done to the click track, in order to keep things consistent).

Moreover, the corrective work you need to do will inform your choices as to which parts of which takes to pick. The goal, at the end of the day, is to have a complete track that sounds the best it can, given what you have to work with. Perhaps this will mean having to discard elements that are in some ways superior, but cannot be made to work with the rest of the material. Them's the breaks.

If you're worried about committing to a bad decision, hold off the final choice for as long as possible. Do a number of rough comps, experimenting with different transition points and try to determine what is going to sound best and require the least amount of work. Only after you've tried out a number of alternatives will you have a good idea of what parts are worth fighting for and where you should just let it go.

Given that you have said correcting everything isn't feasible anyway, comping and then correcting only that which really needs correcting seems the only way to go.