Order of operations for repair of live music recordings?

SpectraLayers
1

So I have some videos of my band where the performance and cinematography are great but audio recording is not and I don’t have multi tracks.

In one case I have:
a decent board mix for vocals ( but there are almost no drums)
3 Camera mics with different amounts of excess room reverb and different balances of instruments due to the position of camera
A recording from the Stage on a zoom H4N.

I can see two ways to proceed here.
Method 1
Use a current AI tool like SLP or LALALA or UVR to separate the vocals out of the board mix

  1. mix the various sources above together using EQ and multiband compression to get the best possible mix then
  2. bring that mix into SL to unmix the elements and then make the mix better
    OR
    Method 2
    Use a current AI tool like SLP or LALALA or UVR to separate the vocals out of the board mix
    Then listen to each source separately and then hear that for example the drums sound best in Camera 1 - so use unmix to get the drums ( and room reverb out of that recording and then use that recording only for drums
    Then find that the sax and guitar are clearest on cam2 so use SL to un mix sax and guitar from Cam 2

Any advice from folks who have done this sort of work would be most appreciated.

2

Yes Spectralayers is definitely capable of unmixing. I have unmixed entire orchestral performances(from each opera singer singing triplets/quartets in unison singing the same harmony to unmixing Trumpets and cellos and even stacked claps and snares layered on top of each other despite overlapping one another) and done stuff like THIS and have done other more complex stuff(where I unmixed similar timbre instruments closely related in frequency like pulse signals mixed with pulse square frequencies).

If you need assistance don’t hesitate to private message me.