preenfm Forum
PreenFM => preenfm2 and preenfm3 => Topic started by: Toltekradiation on November 11, 2023, 02:02:21 PM
-
Hi,
lately i tried to make some choir patch on the preemfm and quite failed at it ::)
i guess it's not that easy to be done with fm, i had to find something else...
Fortunately, there is great informations on the internet !
https://electro-music.com/nm_classic/015_workshops/Clavia/NordModularWorkshops&Threads/WerkMap/WorkShops/Hordijk1999-2000/VOSIM.html (https://electro-music.com/nm_classic/015_workshops/Clavia/NordModularWorkshops&Threads/WerkMap/WorkShops/Hordijk1999-2000/VOSIM.html)
https://scsynth.org/t/fake-resonance-aka-windowed-sync/2720 (https://scsynth.org/t/fake-resonance-aka-windowed-sync/2720)
so i tried to add new algo with windowed sync oscillators :)
here is some quick snippets done with it :
http://sndup.net/tmfx
http://sndup.net/v7wy
Code start here :
https://github.com/pvig/preenfm3/blob/75144a2b58bacd852ae444163dde0cbc9ffc3432/firmware/Src/synth/Voice.cpp#L3621
There is a color code on the preen screen for synced osc.
I chose simple algos : 3 & 4 osc, based on Algo2, Algo 6.
Suggestions and comments welcome about that !
(firmware attached, built on top of 1.02)
-
Those samples sound awesome!!
I cant wait to try out your firmware tonight.
-
Any chance you have some suggested starting points for the IM settings?
-
i'm still learning how to use it as well, it's a trial and error process at this stage.
i guess sidebands created by IM add new formant peaks but alone it's not enough to make a lively voice articulation, one need many matrix modulation.
For now i do simple pseudo voxy resonant thingy patch, it's fun !
Below is the patch bank i currently use, if you want to have a look.
-
perfect. thanks.
I'v played with the three new Algorithms a bunch. I like Algo 30th best so far. I've been able to make a bunch of crazy sounds, but none were vocal or choral like yet. I think I'm not modulating right/enough, so your patches will help me in that direction.
-
i should try different algo, yes algo30 is the most versatile, with its second modulator.
any idea is welcome to help finding new interesting topology :)
-
@Toltekradiation These patches are exactly what I wanted to do. After reviewing the patches, I realized I didn't have nearly enough modulation going on. Thanks for the help. :-) I'm probably going to spend all weekend playing with these new algorithms.
-
Here is a new version, with another algo, which let the master sync osc to be modulated.
And i found a trick to get a much larger pitch range modulation for the slaved osc, pitch modulation from the matrix is now x 16 for those.
It helps to get the fake resonance like heard in the supercollider thread above :D
-
re the additional modulation, thanks. Very cool. I just started playing with the new algorithm.
-
Another version :
Algo29 is reworked as it was too close to algo31 : it's now a 4 osc, with two carrier, a classic one and a synced one.
Matrix modulation is now squared for synced oscillators : sounds more natural and possibly more extreme !
-
as the preenfm2 have not so much love these days, i did a pfm2 firmware version with the lfo shapes from the pfm3 and the 4 vosim algo as well...
it seems to work well, these features does not need much cpu nor ram.
fw attached
-
Awesome. I still use mine all the time. I will update it with this OS to try out the new features.
-
That's really cool. Thanks :)
My work pulled me away from preenfm firmware the last couple of months.
It should much better in a week or so.
I cannot wait to take a look at that.
-
Hello Xavier,
i have a github branch for this, if you want to check it out :
https://github.com/pvig/preenfm2/tree/vosim (https://github.com/pvig/preenfm2/tree/vosim)
-
this sounds super interesting, and a massive thank you for including the Preen2!
edit: just had a go, and it's brilliant, love it!
Just had a super fun patch pop up using it while randomising. Will see if I can build a whole bank to share.
-
Great to hear this ! :D
do not hesitate to report any wish for improvement.
-
Thanks again Patrice for this new algos.
They really add something to the sound possibilities of the preenfm.
And as they are separate new algos, there's not reason not to include them :)
I finaly have a PR, can you take a look ?
https://github.com/Ixox/preenfm3/pull/47/
Also, let me know what you think of the commit i added on top yours (https://github.com/Ixox/preenfm3/pull/47/commits/655da0b1836c590024d2212e77bbd02c011332f7).
I don't think index should be 0 when we sync. Index should correspond to the phase of the master osc. No ?
Don't hear any obvious change in most presets, but i think that could be more "hearable" when waveform are smaller.
Please confirm the changes :
. 4 New vosim (synced windowed modulation) algos
. New midi CC to control paraphonic filter (FX2)
. small adjustment to Filter2 "BODE"
Anything else i miss ?
Can you explain you change in Osc.h (i added a commment in the PR)..
I don't understand.
Thanks :)
-
Hello Xavier,
i answered about Osc.h in the PR : it's about a small cpu optimization..
About the phase reset, yes why not using the master osc phase, it should sound a bit better, thanks !
Another small change :
i removed the bias in the sigma filter as it don't bring anything significant and can possibly cause click
https://github.com/Ixox/preenfm3/pull/47/files#r1527499583
-
Hello Xavier,
i answered about Osc.h in the PR : it's about a small cpu optimization..
I thought there were a problem.
I looked again and it looks good. Thanks :)
-
OK i clean up a little as localEnv were not used anymore in getNextBlockWithFeedbackAndEnveloppe.
Let me know if it's OK for you.
If yes can you "approve" the PR. I'll then know it's good for you.
Thanks
-
Just tried the new master on my preenfm3, works fine !
:D
-
Thanks.
I need to add a little documentation somewhere about the new algos..
And will do the preenfm2 version very soon.
-
About documentation, i would mention the first part of the Rob Hordijk page about vosim :
https://electro-music.com/nm_classic/015_workshops/Clavia/NordModularWorkshops&Threads/WerkMap/WorkShops/Hordijk1999-2000/VOSIM.html
which describe well what is done for the windowed sync part.
then for each algo, the comment of the voice.ccp should do :
ALG29 :
Osc 1 synced by 3
IM3
<----
.---. .---.
| 3*| | 4 |
'---' '---'
|IM1 |IM2
.---. .---.
|*1 | | 2 |
'---' '---'
|Mix1 |Mix2
ALG30 :
Osc 1 & 2 synced by 3
.---. .---.
| 3*| | 4 |
'---' '---'
|IM1 \IM2 IM3/ |IM4
.---. .---.
|*1 | |*2 |
'---' '---'
|Mix1 |Mix2
ALG31 :
Osc 1 2 3 synced by 4
.---.
| 4*|
'---'
/IM1 |IM2 \IM3
.---. .---. .---.
|*1 | |*2 | |*3 |
'---' '---' '---'
|Mix1 |Mix2 | Mix3
ALG32 :
Osc 1 synced by 4
IM4
---->
.---. .---. .---.
| 2 | | 3 | | 4*|
'---' '---' '---'
\IM1 |IM2 /IM3
.---.
|*1 |
'---'
-
@Xavier,
as i added the latest modifications of the vosim branch to the preenfm2 version, here is a PR to make things slightly easier for you : :)
https://github.com/Ixox/preenfm2/pull/18
-
Thanks a lot.
I think i'll have time to look at it this week-end.
EDITED : i didn't do it.... But i keep that in my radar.