attack/decay paramter curve

[st3rk]

Hey Thomas,

Enjoying a lot the last firmware. I'm using the sweet sixteen over I2C. A small issue I have is the attack/decay parameters of the envelops : it affects the time in an exponential way. I get it's a very good way to have a lot of precision for times of a few ms, while being able to reach very high times when adjusting the value directly in the UI. However, it's far less usable when using a fader to adjust the value.

My workaround for now is to divide the fader value and add a constant, to have a better precision. Would it be an option to add a more linear parameter? Or to add "ln" in the calc functions, but I guess even using an approximation it would be CPU intensive?

[proxemics]

Hey Thomas,

Enjoying a lot the last firmware. I'm using the sweet sixteen over I2C. A small issue I have is the attack/decay parameters of the envelops : it affects the time in an exponential way. I get it's a very good way to have a lot of precision for times of a few ms, while being able to reach very high times when adjusting the value directly in the UI. However, it's far less usable when using a fader to adjust the value.

My workaround for now is to divide the fader value and add a constant, to have a better precision. Would it be an option to add a more linear parameter? Or to add "ln" in the calc functions, but I guess even using an approximation it would be CPU intensive?

I'd be curious if Thomas' development was more bound by CPU or memory: lookup tables are a quick way of going linear to exp and back if you have a bit of memory. I ran into this issue as well...

[XORadmin]

CPU and Memory.

The parameter values themself are both linear, however the envelope times are using a lookup table which is exponential-ish.

I can add other calc functions but most are not usable as the results can be only integers. So everything after the decimalpoint would be lost and many of the functions result in 0.x values = unusable.

[sarukas]

Hi,

I am adding here a feature request for ability to switch between linear / exponential envelope response types to all envelopes. This would be a great improvement to the modulation capabilities of Nerdseq IMHO.

An further additional imrpovement would be DADSR and Decay only envelope types.

[XORadmin]

Hi,

I am adding here a feature request for ability to switch between  linear / exponential envelope response types to all envelopes. This would be a great improvement to the modulation capabilities of Nerdseq IMHO.

An further additional imrpovement would be DADSR and Decay only envelope types.

Still there is not enough processing power and/or memory anymore to have exponential envelope responses.
There will be also no additional stages to the current envelopes. Maybe you can get more variations by cascading them?