[sdiy] ADSR Trigger/Retrigger behaviour

Tom Wiltshire tom at electricdruid.net
Sat Jun 28 23:00:05 CEST 2014


On 28 Jun 2014, at 21:25, rburnett at richieburnett.co.uk wrote:

> While we're on the subject of analogue envelopes, and in particular the attack phase, does anyone know if there is any sort of convention regarding how far the capacitor is allowed to charge during the attack phase, before we switch over to the decay phase.

<snip>

> Is there any convention for this, or has anyone even given the implications of this any thought!?!?

I don't know about a "convention" but for some reason both SSM and CEM chose to charge a 5V envelope towards 6.5V (SSM2056 and CEM3310 data sheets). That's about 77% of the max value when it switches over. Maybe it's a fluke, or maybe they had some reason, or maybe one copied the other. Dunno.

I've used the same value in digital envelopes, and it provides a noticeably curved attack, but loses the "nearly flat" part that you get at the end of a charging curve. This was significant for me because I needed all the table entries to differ so that I could interpolate between them to increase the resolution. With 8-bit values that implies at least a change of 1/256th between table entries.
With better resolution, I've moved towards using values closer to 100%, since the flatter portion at the top of the attack curve significantly increases the average amplitude of the overall attack, and this gives the sound more "punch" with very short attacks. Using the full curve is not dissimilar to including a short "Hold" stage before the envelope enters its decay.

HTH,
Tom





More information about the Synth-diy mailing list