Strange Keytracker device behaviour? with phrases

Help, Support & Bugs
1

In the attached xrns there is simply a kick playing on track 1.

The chain goes:

Kick Sample, (with or without phrase activated) -> doofer on: Amp Envelope preset (LFO->Key->Tracker->Gainer)

The phrase for the kick is simply a C-4 being triggered, so switching it on and off should make no difference in playback. This appears to be the case until the doofer is activated. When this doofer is activated and the phrase is activated then it sounds as though some of the attack goes missing. There seems to be some funny interaction going on, it`s like the key tracker gets fired too late after the note event.

–mini explanation in instrument list too:

5811 Doofer Phrase Attack Odd Behaviour.xrns

edit: I`ve just noticed with both activated, the first note played back seemd to be fine, it is just the subsequent notes that lose the attack… .

2

There’s something happening with the TPL hereI think.

3

That phrase timeline does seem to be jumping around rather erratically if that is what you mean?

4

I’ve automated the tpl on the master track and when the tpl is low, with the phrase on, the envelope is too slow, with the phrase off regardless if the tpl is small or not, it sound almost the same, as the resolution is still high do to the bpm/lpb.

But intuitively it would make sense that a smaller TPL should have no effect like this on the phrase.

5

I can hear some distortion in the attack with the phrase off aswell in your example, on two out of four beats.

It` subtle but if the count was 1,2,3,4 ; 1 = most distortion, 2= less, 3 and 4 sound clean.

This could be explained by the automation? but in my original example there is none.

Something funny is going on either way.

6

Well my file doesn’t use the gainer, it uses the floor on a ducking gate, so with those settings it will distort (I prefer to use it instead of the gainer,because 9 (among otherthings)it’s ‘faster’ as the input curves are different and the range is smaller, getting a gainer to do silence rapidly requires a much larger tick resolution as I’ve tested in 2.8)

Edit: after closer inspection any automation of TPL (at least with a meta device) can cause cuts

7

There does seem to be a weird interaction when the phrase enabled. We’ll have to investigate that.

However, the main problem with the dull attack is not really related to phrases or doofers.

Quite simply, there is some ramping/smoothing being applied to the gainer whenever you change the gain level, in order to prevent nasty clicks in the output.

Automating the gain (via LFO or any other means) from total silence to some audible value will never happen instantly. There’s always some ramping applied that will smoothen out the transition, which in this case is softening the attack portion of the kick sample.

I dunno what you had in mind for this particular sound, but if you only wanted to shape the kick drum then I would suggest doing it via the instrument’s sample modulations instead.

8

1st, i don’t know why or how to fix.

The behavior is a little bit different for me. If Phrases is disable, i have the same behavior like you. If Phrases is enabled, all notes are missing the attack, even the first one.

yes I`ve noticed it is sometimes not 100% consistant here too.

There does seem to be a weird interaction when the phrase enabled. We’ll have to investigate that.

However, the main problem with the dull attack is not really related to phrases or doofers.

Quite simply, there is some ramping/smoothing being applied to the gainer whenever you change the gain level, in order to prevent nasty clicks in the output.

Automating the gain (via LFO or any other means) from total silence to some audible value will never happen instantly. There’s always some ramping applied that will smoothen out the transition, which in this case is softening the attack portion of the kick sample.

I dunno what you had in mind for this particular sound, but if you only wanted to shape the kick drum then I would suggest doing it via the instrument’s sample modulations instead.

Yes the report is not complaining about any behaviour of the doofer envelope, just that it differed when the phrase was enabled. I ran into this unexpectedly while scripting so did some hunting for the cause, though sometimes it can be nice to have different methods to do amplitude control too.

Anyway at least it is confirmed and can hopefully be addressed now.