My first attempts at ‘clap’ sounds created on the Super 6
p6_Claps_v1.s6 (35.6 KB)
p7_Claps_v2.s6 (35.6 KB)
Both examples use noise generated on DDS1, with a short decay on ENV2 (D=45-50% w/all other ENV2 parameters set to 0). Both employ a moderate amount of delay (LEVEL 50%, TIME 33%, FEEDBACK 0). The Filter ENV control is set to use both ENV1 and ENV2 (1+2). Both have the filter set as a band pass (VCF HPF ‘FIX’) with the HPF set just below 50%. VCA dynamics in the recordings is set to 1/2 - full dynamics makes the sound ‘squelchy’ at much lower velocities
The idea was to carve out a small section of the basic sound created above by using ENV1 with a small amount of ATTACK HOLD to give the impression of a flam, or two distinct sound ‘peaks’ in the output separated by a very short period. DELAY is then applied to increase the number of ‘hands’
Clap V1 uses a positive ENV profile, using a short DECAY only. This creates a higher frequency peak in the timeline a little after the initial transient. The envelope decay can be adjusted by using the self-modulation from ENV1 on ENV1-D to either a more convex or concave profile. ENV1 can also be used to modulate the other active envelope parameters on both Envelopes in order to realize the most distinct peaks.
Clap V2 uses a negative ENV profile and thus uses the ATTACK phase of ENV1 to create a lower frequency ‘dip’ in the timeline, which is then immediately restored to the initial frequency via the zero DECAY value. These two parameters can also be modulated by ENV1 - again to emphasize those peaks!
There is a fine balance point between the ENV modulation amount and the VCF FREQ that makes the flam stand out the most, but it is really elusive - one moment you have it, the next it’s gone!
I also tried modulating the VCF Frequency with (either and both) LFO1 in HF (non-tracking) mode and DDS2 (not in LFO mode) for a little chaos at the transient, with varying degrees of success
Anyhow, here are two examples without external processing:
This was more of a challenge than I expected, and I’m not certain if I’m completely happy with the results, but maybe by putting this out there someone else can refine the method (or devise an altogether alternate approach!)