[Frameworks] Optical soundtracts at Colorlab - all good?

[Scott Dorsey]

The "frequency parameters" are not really a fixed thing.

There is a frequency response at nominal level, which is one thing.
Then there is a frequency response at an elevated level or a lower
level, which are going to be different.

ALL of the optical sound cameras are going to be spiky in the midrange
and there's nothing you can do about it.

But they are going to be slew-limited... the faster you try and move
the ribbons, the less distance you can move them.  So you will get
high frequencies at low levels that you will not get at high levels.

This is course also limited by the MTF of the printing process and
the original film.  It is _possible_ with a Mauer camera and very
careful printing to get a -3dB point at 15kc on a B&W print, but
you will not get that on a color print and you will not get that at
nominal reference level.  Or anything even approaching it.

But the REAL problem is intermodulation distortion which tends to
blend all these sounds together.  

Your job as mixer is trying to get a mix without too much high end
or low end at high levels but with enough high and low at low levels
and with sounds separated enough.

Some people who mix for 16mm make the assumption that their films
will be shown on crappy classroom projectors and roll off all the
bottom end.  This may or may not be a good assumption depending on
your target audience.  But it is not something you have to do.
--scott