Quote:
Originally Posted by just.sounds  When i use omni microphones in a A-B-C (three in a row) i hear three phantom images.
one between L-R one between L-C and one between C-R To avoid this i should make better channel separation. but when i put the microphones wider the hole in the middle of A-B setups hits me THREE times. The way to make better channel separation would include directional microphones witch is for acoustic recording a compromise IMHO.
So that is why i stay with stereo for the front for now.  does someone have other experiences?? |
I find that I'm quite bothered by these extra phantom images. Reducing them was one of the key goals when Gunther Theile developed the OCT technique. I've made what I consider some very nice LCR recordings using OCT arrays.
Augmenting OCT for surround is simple, and the rear-facing cardioids reject much of the direct sound. More details
here.
Of course, OCT does not use omni's. To accomplish low-crosstalk 3/2 surround using omni's, the only way I know is to use the
Trinnov SRP system. This is shown in my picture at left (at least until I get bored with it).
If one sticks with a stereo front array for recording, it may still be desirable to use the center channel on playback in order to improve image stability. But common LCR panning techniques don't work for this. The late Michael Gerzon studied this problem and patented a class of upmix (and downmix) matrices. They cover not only the common 2->3 and 3->2 cases but arbitrary n->m conversions. They are described in (excruciating) detail in U.S. patent number 5,594,800.
I've been wanting to try such an upmixing matrix for a while now, but I haven't been able to find any commercial implementations. You'd think that someone could build a VST plug in, and charge enough for it to pay the license fee, but so far it hasn't happened. Too bad.
David L. Rick
Seventh String Recording