Gabriel M. Beddingfield wrote: > BUT... on the TR808 kit (where we just fixed the clicking noise) -- this > patch makes it very easy to bring the clicking noise back when editing > the instrument. :-)
I don't fully understand. Do you mean that with 100k it's too easy to make the release stage too short (which causes clicking?) It will be even shorter if you decrease the max time from 100k to 50k :) It's not a problem of too high max time, I think, it's a problem of quadratic scale. There are problems with all "trivial" scales: - linear: most of knob range is covered by long times; if you increase the max from 10k to 100k, 90% of the knob will be covered by values that were out of range before the change - quadratic: the middle-to-max range covers 25% to 100% of the max time, but the lowest 10% or so covers some extremely low values, and you can also argue there's not enough precision in the most useful range of 1k to 10k samples (which is accessible in the "10% to 30%" range of the knob now) - with exponential, the zero value, which was previously a legal value) becomes unreachable I'll try to come up with something that would fulfill those criteria: - 0% would be 0 samples, like it's been before - lower 20% would be for short values (but not "mostly insanely short" ones) - 20% to 80% would cover the typical range for kicks/snares etc (say, 0.1s to 1s with 44100) - 80% to 100% would cover the typical range for cymbals (1s to 10s, with better precision for lower values than for highest ones) Is that OK? Any suggestions? Or maybe clamping release at "safe" time like 30msec would be a good enough "low-tech" solution for 0.9.4? > How about from 10,000 to 50,000 samples (about 1.1 sec. @ 44kHz)? That > patch is attached. What do you think? That was my initial proposal, but I've tested it with cymbals in the GM kit and even 100k wasn't long enough :) Still, it was a good compromise. On the other hand, sure, 50k is better than 10k :) The most universal (and probably least confusing) max value would be the related to max layer sample length, but that means that the knob range changes when samples are added/removed/edited. Some extra food for thought: the whole concept of "max envelope time (in second) is dependent on sample's original sample rate" may case problems in certain situations. Open hats and cymbals are sampled at 44.1k+, so max decay time is short. Kicks and snares may be sampled at lower rates, and yet they get twice the max decay time. And what if different layers are sampled with different rates, will the envelope times change as well? (too long post again :| unfortunately, the envelope stuff isn't as simple as one could think it is) Krzysztof ------------------------------------------------------------------------------ Apps built with the Adobe(R) Flex(R) framework and Flex Builder(TM) are powering Web 2.0 with engaging, cross-platform capabilities. Quickly and easily build your RIAs with Flex Builder, the Eclipse(TM)based development software that enables intelligent coding and step-through debugging. Download the free 60 day trial. http://p.sf.net/sfu/www-adobe-com _______________________________________________ Hydrogen-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/hydrogen-devel
