New filters

[Toltekradiation]

hi @maki, i guess you would like to have the parameter 2 available in the modulation matrix ?
if so, i tried to, but it failed.. (exception crash) because of my lack of understanding the structure of some part of the code (matrixParameterRow), and, according to the code comment( https://github.com/pvig/preenfm2/blob/master/src/synth/SynthState.cpp#L314), 8 destination target were removed in firmware 2.0, i think for a good reason, so i gave up...
for now, i use my digitakt to control param2 by CC, so i don't think i will dig this, sorry.
or maybe i misunderstood, what do you mean by tracking generator ?

[maki]

Tracking generator, as implemented in Oberheim Matrix, Alesis Ion, Alesis Micron, Akai Miniak, called "table" in Alesis Fusion and "edir user curve" in Control assign in Yamaha Montage/Modx is a user defined respons on input in modulation matrix. In this case, instead of having one breakpoint and a few functions left and right form the break point, user could define whatever 8,16 or 32 points response curve, giving much more detailed response. On my Miniak I can distribute fm over the keyboard on many more ways than cust according to a few math curves.This function alone gives tremendous variety of flavours. It looks very simple, like an array of 8,16 or 32 places, each containg a user editable number in range from -127 to +127.

[baxgas]

@Toltekradiation definitely dive in to sound design, you deserve a bit fun also after the hard work on engineering ;-)
I hope these improvements will also get in to the "official" OS. @Xavier?

[Toltekradiation]

@maki : thanks for the info, i wasn't aware of this. It may be possible to add a 4 points fm shaper on each slot, would have to make a poc.
Anyway, it's too much work for me at the moment, maybe in some months.

[maki]

Toltekradiation, if tracking generator is a to big task now, but you still have some extra energy to work, why not implement 1)paraphonic properties in Preenfm2? It seems simple: a counter counts the number of held/active notes and this value changes something in a modmatrix output.
2)Or you could subtract incoming notes and use those values further in modmatrix outputs.
3)Or you could use the time difference between held notes to change something in modmatrix output.
4)Or you could measure the speed of incomming cc events to change something in modmatrix output.
There are so many unused, simple synthesis ticks we musician could use but are not available else in the world, waiting for a developer to implement.

[Toltekradiation]

@maki : thanks for the ideas, these may be fun to try for some pseudo algorithmic music
i would also like to have polyphonic glide, maybe some sort of unison, and step sequence advance with note received.. i may try all those things later.. for now, i prefer to stay in the filter/fx context, to fullfill this part before starting something else.

[superfake]

Hello, @Toltekradiation!
Glad the work continues. 
I apologize if I tear you from the sound design. 

I did some tests ...
I understand that I am probably asking for the impossible... 
So ... is it possible to return the sound of SIGM and FOLD effects from firmware 3108, while maintaining processor resources, as in firmware 0509?

I don’t know how the other guys have, but for me these effects in the latest firmware do not sound as tasty as in 3108.
3108 is cool, but it's too easy to make the processor cry with it...

In the demo below, the SIGM effect from version 3108 sounds better in the lower range, but at the same time, the sound is interestingly compressed - the attacks are smoothed out. And I really like it.
I could not configure the SIGM effect in firmware 0509 so that it sounded the same as 3108.
The FOLD effect from 0509 sounds quieter and somewhat strange.

But, at the same time, I want to have both versions of the SIGM effect. From firmware 3108 and 0509. Because on some other sounds, the new version also gives an interesting result.

WRAP, in my opinion, also sounds strange ...


applying the result of my little test.
In tests, effects start from the same values. The remaining settings, including volume, are also the same.


Fold from 3108
https://clyp.it/rcbwg5sv

Fold from 0509
https://clyp.it/lz3nmtoj

Sigm from 3108
https://clyp.it/to1ho1l3

Sigm from 0509
https://clyp.it/fak0frz3

[maki]

Toltekradiation, I am jealous on programmers ability to change an instrument from the ground up, the way they want. I have many artistic ideas but no programming skills. Could you, please, describe your "working enviroment", tools and software you used to change Peenfm2 the way you want. Thank you.

[Toltekradiation]

@superfake , thanks for the feedback,
yes you're right, i prefer the previous version as well   
in this version the 3108 sound is back on fold and sigmoid ! and it's not so cpu hungry after all
i also fixed a problem on lp2, hp2 when extremely modulated, the filter went off..

@maki, you don't need much software to try things on the pfm2 :
https://code.visualstudio.com/ to edit the code
arm-gcc version 4.7 (see https://github.com/Ixox/preenfm2 )
and git would be helpful but not mandatory (https://git-scm.com/)
but if you're not much into programming, i would suggest you to test puredata, or easier, get an axoloti (http://www.axoloti.com/) and have a ton of fun creating crazy things  !

[Toltekradiation]

adcIn version

[superfake]

Merci beaucoup!

[superfake]

Hello, @Toltekradiation!
Something is wrong... 
SIGM and FOLD in 0709 do not sound like in 3108.
SIGM does not have such energy in the lower range as in firmware 3108.
FOLD also does not sound like 3108 - it still has clipping, plus there was a problem with panning when the FREQ parameter changes.
 

Here is the test result:

Fold 0709
https://clyp.it/au51xft2

Fold 3108
https://clyp.it/04tw0a2d

Sigm 0709
https://clyp.it/b2diewso

Sigm 3108
https://clyp.it/vniflitk


Of course, without effects, patches sound the same in both firmware versions.

Anyway, thanks for your work. 

[baxgas]

@Toltekradiation, thanks, I flashed the new OS, and it seems working nicely (no lags). I didn't test every FX thoroughly so can't really tell the difference for now, but if it improved – so better

Now what I would find an improvement, if we have so sophisticated FX section, is to have a bit more interaction with the Envelopes in the Matrix (eg. I always use Env1 for interaction with the Filter Freq on Note-Events), but maybe it is more a task for Xavier?
I think it would be nice to have modulation destination to control Env1 speed, globally or by parameters (Attack, Decay, Release). This way the Not1 could be used for controlling Env1 speed – good for Keyboard tracking of Env1.
Also would be nice to make available the Filter Parameter2 (eg. Resonance) in the Mod-Matrix. It is available on CC#72 btw.

[Toltekradiation]

@superfake : sorry, my bad, the rollback was not well made, now it should be ok.

[Toltekradiation]

adcIn version