Fastest envelope attack time? How to remove clicks effectively?

#1

Hey :smile:

Was wondering which envelope has the fastest attack time? some of the envelopes has got 2.41 ms attack time as shortest attack time and for example the adsr m envelope attack goes to 0. Is 0 the same as 2.41 ms?

Anyway, sometimes I encounter clicks when scanning through wavetables and was trying to remove the clicks with an envelope. In this context, 2.41 ms is kind of a lot. So was wondering if there are another way to obtain shorter attack times than 2.41 ms? Other objects I should look into?

Thanks

#2

"env/d" has "zero" attack time.
"env/adsr m" has the same scalings as "env/adsr" but misses the real-world unit annotations, that needs further development.

#3

Cool.

I kind of figured the adsr m was same as adsr, but nice to have it verified :smile: and nice to know about the env d. Will check it out.

#4

Tried with those 2 envelopes last night and no luck. There is still clicks when scanning through the wavetables. Have tried that many times before, actually. Still havent found an effective way of removing clicks from wavetables. The env/d has no attack time and the adsr has minumum 2.41 ms attack time. Had no luck with those and I not really convinced that using envelopes is the way to go.

But I can gladly annouce that the additive wavetable creator I am working on is basicly click free. I have not heard any clicks from that patch yet. Great. But When I for example import wavetables from Serum to axoloti there is often a lot of clicking, in most of the files that I can not get rid of.

Anyone else progressed in removing clicks in other situation, not just wavetables? I would be interested in trying different techniques, but dont know where to look anymore, casue all I have tried have not worked that well.. That includes envelopes and overlapp add shifting( i know it is not intended for this but tried anyway). And still no progress really.

Any ideas on how to remove clicks effectively?

Thanks.

#5

Did you tried to use math/window object?
Like it was done here:

#6

Yes. I have tried that. It doesnt remove the clicks. It colours the sound, but does not remove clicks.

I thought to begin with that windows(overlap add shifting) are used for click removing, but they are more for everything time based, compensating for pitch/time. Alteast that is where I am at now. Need to experiment more with it to find out if they have other functions.

But then again there are many different window types, but we got only a hanning window. Maybe some of the othere can be used for removing clicks. I think I saw a hamming window mentioned in context with click removing. But to be honest I dont know that much about them yet... and we have hanning only. So looking into other windows might not be relevant since we dont have them.

Thanks for the input :wink:

#7

ANyway, I think my solution wil be to look through the wavetables I have that creates clicks and compare them to the ones that dont.

What I have noticed is that for example audio files from movies can create a lot of clicking when trying to convert them into wavetables. I think effect like compressor etc. used in mastering chain, has a huge impact on this. The more the masetring is squashed, the more cliking you will have. So getting as CLEAN files as possible is important. Or maybe some decompression could be done to samples BEFORE loading them into Serum and converting them so Axoloti wavetables... Going to test this idea....