Optimal sequence - first comping or first ARA processing?

Hi everybody,

after having recorded vocals by doing a cycle recording (singing for example a chorus multiple times in order to do comping later), I want to work on two next steps:

  • clean up the recording using ARA (to clean up the recording, let’s take as example sonible prime:vocal)
  • comping the vocals

Is there a certain sequence recommended?

  • First ARA incl. flattening afterwards, then comping?
  • or first comping incl. flattening afterwards, then ARA?

I explicitly chose prime:vocal as an example, since there is no additional effort in ARA-processing ALL takes/lanes (as opposed to manual effort one would prefer to save, and thus decising for first comping; that would be the case for example with manual pitch correction using Melodyne, where I surely would first comp).

The main point I refer to, is, whether the “cycled takes” in the lane in reality are ONE audio file, and whether ARA still would deal correctly with it - especially when using “make ARA permanent” afterwards.

Any experiences or hints?

Thank you in advance.

Here’s what I do

  1. Comp the Vocal
  2. Render the comp so it’s a single clean Audio File
  3. Do the ARA Processing
  4. Make the ARA changes permanent, which will create another clean new file

Thanks, @raino