By
m56p87
Thu Apr 04, 2013 3:21 pm
I've actually messed around with this idea and has some success with minimal effort. I know there's probably a more complex and ingenious way of doing it, but I basically just ran a buttload of test tones and a few different noise samples through a 12bit sampler then viewed it all on a spectrum analyzer/oscilloscope. The white noise viewed on the spectrum gave me a great reference for that high end roll off. Then the sine wave test tone tones can give you a reference for the bass end. A really slow sine wave will actually show you the bit depth quantization.
The tricky part was to take into account the non linear qualities of old AD/DA. To increase the effect it's not a matter turning settings up like you would think but actually making different versions of the entire effect chain itself for different "amounts".
Basically run as much as you can, samples, drums, and most of all test tones at different input volumes. It was actually pretty fun and a great way to understand what's happening when you sample into old gear. Why read about AD/DA when you can actually see what's going on. I had todo it with someone over the net who owned an mpc60 since you have yours in front of you things should be easier.
All I can give you in terms of advice, is that multi banding should be a big part of what your building. Multiband bit crush, multi band distortion, etc... It's not as simple as just slopping stuff on in the right amount but simulating the actual response as best you can.
Also one last piece of advice, go with what you see on the oscilloscope.. Of course the mpc60 is "12 bit" but analyzing the signal it's actually more like 13-15bits depending on what the source is. Also distortion will begin on the high frequenies before it does on the low end, so you end up getting distortion on the highs but its more like compression/saturation on the lows hence my suggestion on multi banding your effect chain as much as possible.
