ADSR settings creating clicks on rapid retriggering

P-box

Hi All,

I am reconfiguring the Resynth module by James Walker Hall. It's a brilliant device. What I cannot understand is why, when the attack is too low, or the release is too long, it produces a prominent clicking on rapid retriggering (8th to 16th note speed).

I can work around it by setting the A and R values into a range that avoids the clicks, but it significantly limits functionality.

Resynth.ens

Any Ideas why?

The ens is his original - it comes without samples added. It is still in the User Library.

Thanks!

colB

https://community.native-instruments.com/discussion/49275/adsr-settings-creating-clicks-on-rapid-retriggering

Any Ideas why?

Without looking at the ensemble, there are some basics to watch out for.

Any disontinuity in an audio stream will cause a click (A discontinuity is when the value changes abruptly/instantaneously).

If the attack is very short, this can/will cause a thump, pop, or click, depending on the sound, and how fast the attack can get.

If there is sound playing on the that voice when the envelope resets to zero (to start a new sound), then the existing audio dropping to zero is a discontinuitly, and will cause a click

One option to fix this is to use multiple voices and cycle between them, so that a new sound can start without cutting off an existing one. There are various ways to implement this. It's not foolproof though. If the user sets a very long release time on the envelope, and the sample is long, then its easy to 'use up' all the safety voices until one has to be cut off while still in use, causing a click.

Another option is to use a very short fade out if a voice is active while being reset. Problem there is that you introduce latency, so not ideal depending on the type of sound. Great for pads, not ideal for percussion.

Another is the 'analogue' approach. This is just not resetting the level of the envelope, instead, restarting from the current active level. This works great for virtual analogue sounds where the oscillator doesn't reset on a new note, but if the oscillator resets, or if you are using samples, then although the ADSR level doesn't reset in this case, there will be a discontinuity if the instantaneous value of the previous sound, and the starting value of the new sound are different.

So, as you see, its not a bug, just a law of the game. There are only compromise solutions, no one-size-fits-all simple fix.