Author |
Message |
Oortone
Joined: Mar 07, 2011 Posts: 30 Location: Sweden
|
Posted: Thu May 28, 2015 3:08 pm Post subject:
Can I control which voicenumber to activate? |
|
|
I have a patch where I assign different settings depending on voice using "In/Out Stat>Voice No."
This is for live use and I would benefit greatly if I knew which voice is activated when a key is pressed.
I assumed when starting to play and all voices are off the G2 would begin with voice no. 1 but it seems random.
Am I being right? Are they assigned randomly or can I control voice selection order somehow? Last edited by Oortone on Sat May 30, 2015 6:47 am; edited 1 time in total |
|
Back to top
|
|
|
varice
Joined: Dec 29, 2004 Posts: 961 Location: Northeastern shore of Toledo Bend
Audio files: 29
G2 patch files: 54
|
Posted: Thu May 28, 2015 3:48 pm Post subject:
|
|
|
I don’t think there is a way to select which voice number will activate in a polyphonic patch.
The polyphonic voice allocation does not seem to be random, but instead seems to follow a “rotate” mode. A discussion about the G2 polyphonic voice allocation is here:
http://electro-music.com/forum/topic-15459.html _________________ varice |
|
Back to top
|
|
|
Tim Kleinert
Joined: Mar 12, 2004 Posts: 1148 Location: Zürich, Switzerland
Audio files: 7
G2 patch files: 236
|
Posted: Thu May 28, 2015 4:25 pm Post subject:
|
|
|
The voice assignment algorithm of the G2 is programmed to prevent cutting off already sounding voices as good as possible, taking into account the voices that are currently played on the keyboard (or via MIDI) and the ones which have been released. Thus the sequence of voices becomes scrambled and appears to be random, but isn't at all.
Overriding the internal G2 voice assignment scheme has been a recurring subject and a much desired feature.
A simple scenario is mapping specific keys to specific voices, eg. 32 voices assigned to 32 specific keys on the keyboard. This isn't straightforward, but possible with minimal compromises (reduction of update rate due to keyboard scanning). I've used this kind of approach, albeit for a different purpose, in my user-definable arpeggiator algorithms (found in the building-blocks section).
A dynamic scenario, where any voices will be re-assigned to any key they where last used for -if they have not have been re-used for other keys in the meantime due to polyphony restrictions- (referred to as "repeat" mode) is a lot more complicated and remains an open problem on the G2 to this day.
Actually I conceptually know how to do this in a reasonable way, but the fact that the interslot busses have latency makes it tremendously difficult, because the FX section (containing the main engine) ideally needs to be able to control the voices in realtime in order to make it work. So it would mean having to implement some kind of additional latency compensation or whatever.
And one has to keep in mind that these kind of new voice allocation schemes have to be efficient enough to leave enough resources for the actual sound generation patch, otherwise it becomes pointless.
cheers,
t |
|
Back to top
|
|
|
Oortone
Joined: Mar 07, 2011 Posts: 30 Location: Sweden
|
Posted: Sat May 30, 2015 6:43 am Post subject:
|
|
|
OK, thank for the info.
I actually had little hope it would be possible.
In my case one solution might be to treat the lowest key pressed in a separate way, since it's the bass note so to speak.
Any advice on how to determine which is the lowest key, that is; how to make a comparsion between the voices? |
|
Back to top
|
|
|
Tim Kleinert
Joined: Mar 12, 2004 Posts: 1148 Location: Zürich, Switzerland
Audio files: 7
G2 patch files: 236
|
Posted: Wed Jun 03, 2015 4:09 am Post subject:
|
|
|
Oortone wrote: | OK, thank for the info.
I actually had little hope it would be possible. |
I "seeded" my brain with the problem. Parts of the solution I already have in my head, others are a lot harder -interslot latency, voice stealing... Who knows. I still have one G2X here (one already sold) if I get somewhere.
Oortone wrote: | Any advice on how to determine which is the lowest key, that is; how to make a comparsion between the voices? |
If you just need the lowest key, the Monokey module in "lo" mode will give you just that.
Sorting voices top-down or down-up is more complex but possible. I did that in my DIY arpeggiator building blocks. |
|
Back to top
|
|
|
Oortone
Joined: Mar 07, 2011 Posts: 30 Location: Sweden
|
Posted: Sun Jun 07, 2015 8:55 pm Post subject:
|
|
|
Tim Kleinert wrote: | [...]
Sorting voices top-down or down-up is more complex but possible. I did that in my DIY arpeggiator building blocks. |
Thanks, I'll have a look at that.
I too have experimented with interslot solutions and latency is clearly a problem. And in a live situaion I don't want slots to depend on each other since my way of working is patch based rather than performance based.
Lowest key was easy I realize now. I seldom use the mono key.
I so much wish Clavia could release a Modular G3. It would be absolutley amazing with modern DSP in such a creative tool. |
|
Back to top
|
|
|
Tim Kleinert
Joined: Mar 12, 2004 Posts: 1148 Location: Zürich, Switzerland
Audio files: 7
G2 patch files: 236
|
Posted: Mon Jun 08, 2015 1:40 pm Post subject:
|
|
|
Oortone wrote: |
I too have experimented with interslot solutions and latency is clearly a problem. |
As I mentioned, I've given the topic of totally overriding the inbuilt voice allocation some serious thought over the years.
However, since most of the approaches require DIY keyboard scanning and also require interslot latency compensation at each roundtrip, the timing of the voices would be compromised. So I dropped the problem, as it is not worth the hassle.
Oortone wrote: | I so much wish Clavia could release a Modular G3. It would be absolutley amazing with modern DSP in such a creative tool. |
Imagine a G3 with their new floating point DSPs and the resulting radically improved sound quality (my favourite is the NordLead A1, it sounds absolutely amazing!). Would totally take the cake. But.... oh well. |
|
Back to top
|
|
|
|