Can you hear the clicks?

[JerGoertz]

Hey all,

I'm noticing high-pitched clicks in the attack portion of a Synth Bass patch for a certain (unnamed) synthesizer. Both the amp and filter envelopes are set at 2 ms. I can hear the clicks on many of the notes, although not all.

Please listen to this file and tell me whether you can hear the clicks, and whether you think it indicates something is wrong with the synthesizer or patch, or whether this is expected behavior.

(In my experience, I think it's possible/expected for a synth patch to have a fast-ish attack without having a click like this.)

https://soundcloud.com/user-748763066/c ... 8aa5f47194

[BertKoor]

With extremely short attack times clicks are to be expected.

Some synth envelopes just are clicky where others are not. The strategy to avoid it me thinks is to tell a lie in the GUI: value zero is not really 0 ms attack time. Also a bit of a curve can help, just as syncing oscillator phase as opposed to having free running oscillators.

You may reveal the name of your suspect for others to confirm.

[JerGoertz]

Does 2 ms count as an extremely short attack time that would create a click like these?

[Unaspected]

Maybe try increasing the sample rate of your system and see if that improves things.

At 44.1kHz, 2m = 88 samples (rounded down)
At 192kHz, 2m = 384 samples

[JerGoertz]

Maybe try increasing the sample rate of your system and see if that improves things.

At 44.1kHz, 2m = 88 samples (rounded down)
At 192kHz, 2m = 384 samples

That seems like a reasonable thing to try, I'll give it a shot here...

[Unaspected]

Of course, it will only improve things if the envelopes operate at audio rate or if control rate is in some way tied to sample rate in that synth.

[JerGoertz]

Of course, it will only improve things if the envelopes operate at audio rate or if control rate is in some way tied to sample rate in that synth.

Of course, that was my hope as well.

Unfortunately, my shitty soundcard only goes up to 96k. Going from 44.1 to 96 didn't seem to improve things, so... I dunno.

[BertKoor]

Here's a screenshot of a part of the audio you shared, zoomed in ofcourse:

Screen Shot 2021-12-07 at 09.48.59.png

I highlighted a section being 2ms long there, being your attack time. You can also see that the oscillator is free-running.

The very sharp bends at the start of the note and end of the attack are discontinuities, and you hear those as clicks. Some notes start close to a zero-crossing of the oscillator, then there is hardly any click.

The attack envelope is implemented quite naive and thus expected to click. Simple design fault if you ask me

[JerGoertz]

Here's a screenshot of a part of the audio you shared, zoomed in ofcourse:

Screen Shot 2021-12-07 at 09.48.59.png

I highlighted a section being 2ms long there, being your attack time. You can also see that the oscillator is free-running.

The very sharp bends at the start of the note and end of the attack are discontinuities, and you hear those as clicks. Some notes start close to a zero-crossing of the oscillator, then there is hardly any click.

The attack envelope is implemented quite naive and thus expected to click. Simple design fault if you ask me

Thanks for the analysis. I agree with your conclusions. How unfortunate that the developer decided to code things this way.

[Unaspected]

That is unfortunate. It's not caused by an extensive release time? If the envelop doesn't always restart at zero and has no option to switch behaviour, surely it should settle back to zero eventually?

[JerGoertz]

That is unfortunate. It's not caused by an extensive release time? If the envelop doesn't always restart at zero and has no option to switch behaviour, surely it should settle back to zero eventually?

It's not related to release time (just checked).

[Unaspected]

That is unfortunate. It's not caused by an extensive release time? If the envelop doesn't always restart at zero and has no option to switch behaviour, surely it should settle back to zero eventually?

It's not related to release time (just checked).

Last resort would be using envelopes provided by your DAW. Just leave your oscillators running and attach an envelope controller to an appropriate gain parameter in the synth.

[BertKoor]

Another solution is to put a separate EQ plugin after it set to low-pass.

Or shop for another synth with non-clicking envelopes. This seems a very basic sound that almost all synths can produce.

[JerGoertz]

All good suggestions, I'll see what I can do here. I might just send up editing the waveforms to remove the clicks after rendering as the first step before mixing.

[DJ Warmonger]

There are some tiny clicks - at random, which comes from random oscillator phase.

I don't think this will be a significant issue in a full mix, but if you start a track with this bassline alone it will sound wrong.