# VCA enveloppe release cut abruptly

Hello Shruthi users

The release of the VCA enveloppe has a very abrupt end (even with a high value, of course the release last a long time but the end of the sound is always cut suddenly and audibly). I’d like to know if it’s an issue with my Shruthi, or is it general?

Thanks

Patrick

Could you please post a .wav file showing the problem?

I think I’ve found something :

In the modulation n°9 page (env2 -> VCA), when the amount is max (63), the release is fine (slow decay until 0). When the amount is under 63, the tail of the sound is suddenly cut. Something else, when amount is 63, sound seems to be weaker than with a lower amount (must be the opposite, right?).

http://dl.free.fr/eP1fJhkkw

first note: amount=63, second note: amount=40

Hope this can help.

Patrick

When you set the env->VCA amount to 0, you want the envelope to have no effect on the sound - ie it is constantly at full amplitude. When you set the env->VCA amount to 63, you want the envelope to have full control on the VCA. When you set it to 50%, you have the exact average of those two behaviours - the mix between “sound at full amplitude” and “sound controlled by the VCA”. In other words it’s not like the modulation amount is balancing between the source and nothing - it is balancing between the modulation and “full amplitude”. This is better understood with velocity: with velocity amount set to 0, the sound is always at full amplitude ; with velocity amount set to 63, the sound is at an amplitude linearly mapped to the velocity.

So I believe there’s nothing fundamentally wrong with the way the modulation matrix works.

Where there’s a problem is in the following “hack” which is there since the early version: the oscillators go off when both envelopes have reached 0. This saves computation time when the Shruthi is “at rest”, and keep it silent, since the VCA on the SMR4 board bleeds a bit. Now I think that the real criterion for deciding whether the oscillators should go quiet is not the envelope values, but simply whether the VCA control signal has reached 0. I have this change in preparation for the next firmware.

I take your message as an additional vote for the “digitally kill the sound only when VCA gain is 0” fix.

Thanks Pichenettes,

From your last quote I think I understand. It sounds like oscillators turns off when enveloppe reach a certain point, different from 0 because the amount is not max. Can make sense after what I’ve heard. It will be interesting to have this change.

Patrick