[Frank Stearns]

"Peter Larsen" writes:
Hi Guys,

here is how to get a spot microphone just right if you again forgot measure
or record a clap near it:

1) know the order of magnitude of the required delay so that you start out
close to right

2) pan suitable main pair microphone fully left or right

3) pan spot mic fully right

4) get volume reasonably balanced

this does it!

5) adjust delay for proper center image of spot microphone sound source

/this does it!


Not sure I follow... Step 2: you mean, just one of the main pair mics? What flavor
of stereo pair are we talking about? AB? ORTF? NOS? XY? MS? (etc)

Step 3: should the spot pan be opposite pan of the selected main pair mic (assuming
you're just using one from the main pair).


If I understand it correctly I like the approach, but anticipate some little
gotchas with this method. Want to make sure I fully understand it and then try it.

I used to take clicks and clacks but found that it's just easier to start with your
steps 1 and 4, then single-sample step a delay on the spot channel and listen for
the best results. Once found, the spot level is often pulled back, lessening any
remaining artifacts.

Often as I do the initial step-through of delay values there will be more than one
delay that will sound good. I'll note them all then do a second pass to winnow
those delays to the absolute ideal.

Once in a while I'll even automate a delay factor as the same delay might not sound
as good in one piece as in another (featured instrument/voice picked up by the spot
has shifted their position, changing the spot/main pair geometry).


6) add a few milliseconds if spot microphone sound source protrudes when
properly mixed.

Yes, you can do this to push the spot out of comb filtering. Just don't hit the Haas
boundary, unless you want the effect.

Note: adjustment sensitivity is single milliseconds. May for some be easier
when listening via headphones and for some be easier when listening via
loudspeakers, not fully verified what works best, try both!

If it's a good room with good monitors free of their own time problems, speakers
generally are a little more revealing for this.

I find millisecond steps can be too large; single sample steps can be better. Quite
often the ideal delay is only 3 samples wide (44.1K); go one sample out on either
side and you lose the good sound.

Thanks for the idea; will be interested in clarifications.

Frank
Mobile Audio
--