First off, just to be clear - this project isn't really a main focus for me at the moment - I've bandied around ideas of Matrix Mixers for several years (there was the 4x4 in the modular line and I almost got going on a very big mixer collaboration project a few years back..). So, don't yet count on this actually ever reaching fruition - and anything will likely be some way off anyways. (did I kill the buzz enough already?!). As I said, partly I'm interested to see if people are interested!
I've seen a few PASSIVE Matrix Mixers appearing of late. But for proper fidelity and control, I really think ACTIVE is the only way - just as with regular mixers.
How big? Well, you'll almost always want more inputs than outputs (so some can be used as effects returns most likely).
Does a 6 x 4 provide enough, but not too much?
This initial iteration has:
- input preamps for line level signals (fine with modular though AC coupled, of course) + simple clip level leds (at this stage, keeping it simple)
- 24 knob matrix
- output buffer / level controls
Having tried this, I do feel already there is need for:
- improved signal level indication - 2 or more leds per input and maybe the same for each output.
- still doubtful about whether phase and/or boost/pad switches may be good
- maybe impedance balance the outputs.
- scribble space (white rectangles) on the panel to note input/outputs.
One other thought would be to treat the lower two rows differently - change one row to PAN and the other to LEVEL. Saying that, same could be done for the upper two rows too. It is a subtle but perhaps useful difference.
Oh and I haven't yet tried any form of zener/diode level limiting - something I've used a bit of in recent designs & which could well be useful here - - definitely something for me to try in the next stage.
The build size is twice the length of my current standalones (ie. 6 FracWidth) with internal DCDC power conversion. Decent, but not over-the-top-$$$ opamps inside.
I don't think anything really exists along these lines at present?
(sorry for the slightly shitty picture)
sounds very, very interesting to me!
6 x 4 means i can patch it as 4 x 2 with 2 channels of feedback (that can be routed through a pair of vcas, for example). that should be sufficient for my purposes, although i wouldn't if there were another pair of inputs.
Yes to banana and 1/4", balanced outs, and some scope for inversion/feedback.
:hail: hooray, tho i would also be just fine with the original design shown on p1 (esp if items such as click/thud-free operation of the switches adds signiciantly to product cost and development time - knowing 
tho, i think i can solidly hope for clever solutions).
