From the original SynthMaker forum (2004–2014), with its members’ names and dates.
stw ·
Related to the discusion mentioned HERE i urgently request an improved voice management. This topic has many aspects and can be expanded in different directions like ENV behaviour etc.
But the most important improvement is an alternative how SM handles stolen notes.
For now we only got a simple kind of oldest voice stealing method after reaching max polyphony. I don't know ANY other Synth which does it that way and simply for that reason SM Synths can't compete in the real world IMHO.
When voice reduction comes in it is required to have more control over it or at least let do SM what other instruments do by nature -> HOLD notes which are still pressed and kill released ones first!
There've been some discussions about voice management in the past and obviously some features made it into SM. But unfortunatly not that one.
matti ·
stw wrote: Related to the discusion mentioned HERE i urgently request an improved voice management. This topic has many aspects and can be expanded in different directions like ENV behaviour etc.
But the most important improvement is an alternative how SM handles stolen notes.
For now we only got a simple kind of oldest voice stealing method after reaching max polyphony. I don't know ANY other Synth which does it that way and simply for that reason SM Synths can't compete in the real world IMHO.
When voice reduction comes in it is required to have more control over it or at least let do SM what other instruments do by nature -> HOLD notes which are still pressed and kill released ones first!
There've been some discussions about voice management in the past and obviously some features made it into SM. But unfortunatly not that one.
+1
Well i guess i was the first to mention this anyway 🙂
It's pretty important for people who want to play the synths they create..
infuzion ·
While I see your point, why do you even need to worry about stolen notes? Why not set polyphony to a ridiculously high number like 32, so you don't have to worry about stolen notes? Maybe I'm missing something here...
stw ·
infuzion wrote: While I see your point, why do you even need to worry about stolen notes? Why not set polyphony to a ridiculously high number like 32, so you don't have to worry about stolen notes? Maybe I'm missing something here...
Simple...
a) my synth is a CPU hog
b) i play sounds with long releases
c) i play with damper pad etc. etc.
there could be any reason why voices have to be reduced.
One of the most important reasons for an improvement on this: It's the common and most (IMHO only) reasonable way of voice allocation! That shouldn't be a feature that should be self-evident in an enviroment which is called "SYNTH Maker" 🙁
Tzarls ·
I don´t know if this is of any help, but I saw around here a module to keep track of the notes being played, in the order these were played. Maybe that could be used as a foundation for a voice management tool, to overcome SM´s "limitations"?
Acrobat ·
That's a very important issue, what can I say, stw, I just can say you are damn right and I hope for a solution soon...
it's years I claim for some RED-V (voices) primitives to edit voices, or some other solutions to the task... 🙁
matti ·
infuzion wrote: While I see your point, why do you even need to worry about stolen notes? Why not set polyphony to a ridiculously high number like 32, so you don't have to worry about stolen notes? Maybe I'm missing something here...
Wow.. What supercomputer are you using? No matter how much power i've ever had under the hood of a computer at a studio i've been, it's always running out at some point. With this trick that limit would be reached even sooner. 😕
infuzion ·
matti wrote:
infuzion wrote: While I see your point, why do you even need to worry about stolen notes? Why not set polyphony to a ridiculously high number like 32, so you don't have to worry about stolen notes? Maybe I'm missing something here...
Wow.. What supercomputer are you using? No matter how much power I've ever had under the hood of a computer at a studio I've been, it's always running out at some point. With this trick that limit would be reached even sooner. 😕
Running out of voices in a single synth?!? Just copy the same synth into several tracks, & put some notes on one track, & other notes on another track. I've very very rarely used more than 8 voices in a synth since I split up different parts into different synths. Even if I'm playing chords & melodies on the same piano sound, I'll still use separate plugins; easier to handle dynamics & editing that way. I can see if you're using a sustained sound like harp or piano you may think you need tons of voices, but perhaps longer sustains like this are better suited using reverb at a certain point
Sorry, I still believe there are plausible (though extra work) workarounds. & I still think advanced users who know Code/ASM can make their own voice management.
matti ·
infuzion wrote:
matti wrote:
infuzion wrote: While I see your point, why do you even need to worry about stolen notes? Why not set polyphony to a ridiculously high number like 32, so you don't have to worry about stolen notes? Maybe I'm missing something here...
Wow.. What supercomputer are you using? No matter how much power I've ever had under the hood of a computer at a studio I've been, it's always running out at some point. With this trick that limit would be reached even sooner. 😕
Running out of voices in a single synth?!? Just copy the same synth into several tracks, & put some notes on one track, & other notes on another track. I've very very rarely used more than 8 voices in a synth since I split up different parts into different synths. Even if I'm playing chords & melodies on the same piano sound, I'll still use separate plugins; easier to handle dynamics & editing that way. I can see if you're using a sustained sound like harp or piano you may think you need tons of voices, but perhaps longer sustains like this are better suited using reverb at a certain point
Sorry, I still believe there are plausible (though extra work) workarounds. & I still think advanced users who know Code/ASM can make their own voice management.
Ah, sorry i see you missunderstood what i ment. I was talking about the fact that if every synth would use 32+ voices for something that was easily achievable with 8, it would chew up all cpu resources a bit too fast.
Getting the voice management to work better would be cool. I might look into it(if it's even possible) when i have more time..
infuzion ·
matti wrote: Ah, sorry i see you missunderstood what i ment. I was talking about the fact that if every synth would use 32+ voices for something that was easily achievable with 8, it would chew up all cpu resources a bit too fast.
Getting the voice management to work better would be cool. I might look into it(if it's even possible) when i have more time..
Yes, I did misunderstand, sorry!
"Max Voices" does not work the way I think you're thinking about. Whenever you play one note, it will use a single poly stream, regardless if you set "Max Voices" to 32, 8, or 99. If you play 4 notes, still only 1 poly will be used. When they play the fifth note at the same time, a second poly stream is started. When they are playing some mega-chord with 30 notes (with Max Voices set to 32 or more) at once, they'll be using 7 poly streams. If that is too much CPU% for them, tell them not to play 30 note chords! 😉
So "Max Voices" = upper limit, not all turned on. SM will only turn on the minimum poly streams as needed.
matti ·
infuzion wrote:
matti wrote: Ah, sorry i see you missunderstood what i ment. I was talking about the fact that if every synth would use 32+ voices for something that was easily achievable with 8, it would chew up all cpu resources a bit too fast.
Getting the voice management to work better would be cool. I might look into it(if it's even possible) when i have more time..
Yes, I did misunderstand, sorry!
"Max Voices" does not work the way I think you're thinking about. Whenever you play one note, it will use a single poly stream, regardless if you set "Max Voices" to 32, 8, or 99. If you play 4 notes, still only 1 poly will be used. When they play the fifth note at the same time, a second poly stream is started. When they are playing some mega-chord with 30 notes (with Max Voices set to 32 or more) at once, they'll be using 7 poly streams. If that is too much CPU% for them, tell them not to play 30 note chords! 😉
So "Max Voices" = upper limit, not all turned on. SM will only turn on the minimum poly streams as needed.
I don't want' to sound rude, but i do understand how voice limiting works(im familiar with the term "limit"). You, sir, haven't understud how the voices work when you play sounds with long release times. If you can limit the amount of voices playing at one time you can play things that sound pretty much the same as with higher voice counts, if the oldest released notes are the ones that cut out first. Even tho it's originally a limitation of synths with finite amount of polyphony, it's still useful if you don't have infinite amounts of processing power on your DAW.
So currently when the voice management cuts out the oldest note that was triggered, even tho it's still playing, you can't limit the voice amount without glitches when playing two handedly..
Yoy should try it. Take a sound with a long release time in an SM synth that has a voice limiting option. Start with something relatively low limit, like 8. Now hold a bass note with your left hand and play other notes with your right. What happens when you play more than 7 short notes(different ones) right after each other?
Now play with just one hand, with a limit of 16 voices or 32 voices. Then with 64 voices. Do you notice a difference in sound? Do you notice a difference in cpu usage?
exonerate ·
If this is expected behaviour (I don't play keyboard much myself) then it should be added rather that us trying to hack our way around this problem. It does seem like perfect common sense that if you are holding notes you want to keep them held!! they shouldn't just cut out, so I see this as being a major issue actually.
I tried to think of someways round this, I think it *may* be possible, but even if it is it'll just be an ugly hack that is adding to CPU, this should be an option in the midi to poly.
I can't see this being a hard thing to add, just keep a list of all the notes currently in release, when a note is stolen kill the voice of the oldest note from that list.
Cheers 😉
Exo
stw ·
exonerate wrote: If this is expected behaviour (I don't play keyboard much myself) then it should be added rather that us trying to hack our way around this problem. It does seem like perfect common sense that if you are holding notes you want to keep them held!! they shouldn't just cut out, so I see this as being a major issue actually.
I tried to think of someways round this, I think it *may* be possible, but even if it is it'll just be an ugly hack that is adding to CPU, this should be an option in the midi to poly.
I can't see this being a hard thing to add, just keep a list of all the notes currently in release, when a note is stolen kill the voice of the oldest note from that list.
Cheers 😉
Exo
Hi Exo your thougths are almost a perfect blueprint of mattis, Acrobats, mine and others requsts and statements about that issue. Welcome to the club 😉
stw ·
may i throw in a little reminder here.... 😈
I still think this is major issue and should be adressed as soon as possible
stw ·
ok 1.1.3.1 is on stage. So it's time for a reminder of a still fundamental issue!
BUUUMMMMMMP
Acrobat ·
stw wrote: ok 1.1.3.1 is on stage. So it's time for a reminder of a still fundamental issue!
BUUUMMMMMMP
and BUMP again,
Reading some notes by Exo here about notes HELD in SM, I must say one thing I would do is lets go see how OLD things worked, and how they worked well: GATE is the word.
In old electronical instruments a note is a channel interrupted by a gate, like in many others HI-FI electrical components, but we do now think the fu**ing opposite, that a note is something you turn ON... 😉 😉 😉
Wrote something here? To have a post removed or your name hidden, write to [email protected].
SynthMaker