Hi this bug has been affecting my workflow for a while (and i’m sure many others) -
With oscillator phase randomisation is set to 0, & Voices are set to 1 (mono), long release times affect the consistency of the attack/phase of new notes.
PLEASE resolve this as it’s a big functionality issue, in which the only current ‘workaround’ is having very short/fine tuned release times, and using long notes that don’t touch in the DAWs piano roll - far from ideal!
This issue isn’t resolved by adjusting any voice override settings or other that I’ve tried and doesn’t appear in other synths (eg Ableton Wavetable, Xfer Serum), so I strongly feel that it’s a bug as opposed to any form of intended behaviour at all.
Vital’s envelopes aren’t super consistent anyway. How does the release time affect the attack phase? Could you describe it in more detail and perhaps provide an example patch? Would probably help the dev eventually.
Set the Amp AHDSR Release to a long value, eg 3-4seconds.
Input three short notes in a single bar in any DAW.
The attack of each new note is now affected adversely by the long release time, you can clearly see this using an oscilloscope or by rendering the MIDI to audio and viewing the waveform.
In legato mode the 0 phase randomisation is effectively nullified, creating randomised oscillator phase reset.
Alternatively with legato mode disabled, the attack phase of each new note is also no longer clean and consistent, as the release tail of each note prior affects the phase, amplitude etc of each new note, rather than the release being cut off cleanly.
The voice override mode ‘Kill All’ should counter this behaviour, but it doesn’t.
Any ‘808’ style patch or other basic sub bass patch will display this behaviour.
The envelopes seriously need a subtle but crucial design overhaul/tweak to fix this issue, otherwise Vital is my (and many other producers and sound designers) daily driver for sure.
What I can see is that with voice override at kill and legato off the previous note isn’t cut off immediately when the new one arrives. With voice override at steal or legato on despite the override mode the new note cuts the previous off right away.
Aslo, it seems that at the onset of a new note the phase of the oscillator is rotated 180 degrees, but somehow it seems as if the new note affects the previous note before its onset, bounced in Bitwig. Dunno if this is a Bitwig quirk, but surely it’s weird,
This picture’s reverse saw should start at down ramp from zero line at note onset:
Somehow it’s starting at 180 degrees instead of 0 degrees in Vital terms. Also, something strange happens to wave before that’s octave above before the new note hits.
Is this what you mean?
edit: actually it seems like that when Vital receives a new note it’s trying to keep the oscillation consistent and avoid sudden jumps in phase, thus avoiding clicks I suppose. This seems like a design decision and would be preferable in many cases ig. So if you need the starting phase to be consistent at note onset ig you need to make sure the previous voice’s oscillations are finished before a new note is started. That’s what you’d ought to do even if Vital didn’t do that, otherwise your 808 sub patch might do clicks and pops.
Unfortunately no I will show you the issue shortly in several screen shots, along with confirmation that Serum and other synths don’t behave in same way (based on surgically identifying the issue in an oscilloscope, I will post all additional info here.
I can also refer you to an earlier post from November 2020 that reflects the same issue that I refer to,
Part of the issue is described in this post - the legato on observation - However a related issue is displayed with legato off as I will show in upcoming screen shots of the oscilloscope with descriptions of settings applied within Vital.
After doing further analysing, I can confirm that Serum set to mono with oscillator phase randomisation turned off behaves the same in both legato and legato off modes - this is actually the answer - Vital’s legato off equivalent leads to amplitude increases with each new note, however, which could potentially be avoided (though as you say dependent on whether it will create clicks or not (maybe/maybe not) as displayed here)
However I really feel like this could be improved, which potentially would lead to it having functionality that would be better than other synths, especially as the functionality is almost there - but that additional volume from the compound of both the new and last note is the actual ‘issue’ observed.
What happens if you change the oversampling (quality, in advanced) of Vital up or down? Wo dering if that may improve (or maybe worsen) the issue?
I’d recommend a Dash Glitch video on YouTube be has one about this effect as it does occur in others synths as well.
It’s just because the time factor required for the env is so small, there are tiny imperfections as the current waveform rls occurs and the attack of the new both happen within that 0.001s interval
Oversampling doesn’t alter the affects of envelopes or LFOs as envelopes unfortunately, but they can improve the aliasing of the wavetable - though not relevant when we’re discussing static simple waveforms.
I’ve watched Dash Glitch’s videos in the past for other purposes, they’re great but not relevant to this scenario (yes his vids on Vital are cool and deep, and I respect his content, along with several others that cover amazing use of physical modelling techniques with exciters, comb filters etc)
Also unfortunately I was incorrect in my last response, the issue is real and still persists, and is also related to the functioning of the sampler between Legato and non legato modes (1 shot sampler mode doesn’t behave as expected/exist in legato mode), LFOs as envelopes etc. Serum’s envelopes/LFOs as envelope behaviour is tighter, but Vital is still my go to for most sound design uses regardless, as it’s ease of use and undo/redo are highly beneficial.
Either way, Matt Tytel, please address the envelope behaviour with the next major update, as well as sampler 1-shot mode with legato enabled.
the only workaround that I’ve used myself, and that anyone else covers is effectively, - draw or play long notes that don’t meet up, and don’t use the release phase in the envelope, thus preventing the release phase from affecting the following note. This doesn’t change the other issue I mentioned in my last post of sampler 1-shot mode not functioning in legato mode however, and other synths are capable of the desired behaviour without short release times.
Nice, the workaround you found is the one Dash presents in his video, so all good there. Best of luck!
I only mention oversampling if the lil spot is clipping vital as well then perhaps quality might then let you control that clip more for an attempted workaround, despite the release issue
