AW: [sdiy] Crazy idea to improve transient response?

[Glen]

At 05:36 AM 9/25/03 , Nils Pipenbrinck wrote:

>Anyways, if you fade between two inputs based on one input signal you'll
>call for trouble. The notches will start to move around. What you do is to
>mix two amplitude modulated, phase-shifted signals together. This will give
>distortion.

I got the idea from looking at the waveforms from two different mics that
were recording the same source (a glockenspiel). One mic had a better
transient response, but a much lower overall output level. It looked like
an interesting idea to "graft" the impulse transient of that mic onto the
signal of the other mic. I realize that technically speaking this creates
all sorts of opportunities for distortion to creep in, but it would only be
happening on large and brief transient peaks. Under those circumstances, I
think the ear might be fairly forgiving of the errors. 

It would seem much more accurate than having a signal with practically no
transient response at all (the unmodified signal from the
transient-challenged mic). It would also seem more accurate than boosting
the volume of the mic with the better transient response, and suffering the
increase in noise levels that accompany such a boost. If the mic with the
good transient response is a noisy condenser mic, the noise problems are
even more of an issue--boosted or not.

I know that aligning the mics and balancing the mic preamp circuits might
be critical, but I still think this idea might have some merit to it. I
think I'll breadboard something and see what I can get. At the moment, both
of my mics are dynamic, but I want to try this with one mic being a very
cheap condenser. I want to see if I can get a good transient response from
the cheap condenser and use the dynamic mic for the majority of the signal,
especially the smaller signals that would benefit most from lower noise
levels of my dynamic mic.

I also don't want to simply mix the signals together. That wouldn't give
the noise reduction effect I want to achieve from the cheap condenser, and
switching from one signal to the other should avoid many, if not all, of
the phase cancelation effects you mentioned. Much of what you mentioned
depends on mixing two signals together of different phase. I won't be doing
that. I'll be switching between two different signals, based on their
amplitude. I know there might still be some phase problems during the
crossover period, but I think they are probably minor issues compared to
what you described.

I forgot the company's name and the model number of the mic, but someone
actually makes a combination dynamic and condenser mic already. I don't
think they have an amplitude-driven active switching circuit in theirs, as
I am thinking of doing, but they *are* using more than one type of capsule
inside the same mic. I think they compare their idea to that of a two-way
speaker system. They are probably splitting the signals from the two
capsules according to frequency, and I want to split them according to
amplitude.


later,
Glen