some thoughts/suggestions

[insane]

im not sure but the modules seem to be fixed, i mean their position, like racks, i would enjoy an option to display them freelly..
thanks.

[sonigen]

im not sure but the modules seem to be fixed, i mean their position, like racks, i would enjoy an option to display them freelly..
thanks.

Hi, that has been requested before and it is on the list of things to investiage, but it's low priority.

[SoulState]

Please make "constant" fader also in knob view. And ability to rename modules (or write some tips on them). And when mouse cursor at mod. amount knob - is shows a tip from which module this mod. input came.

+Limit of 32 voice is too small, pls make at least 64 or, better, 128 voices max. (32 voices limit - it is only 6 notes in 5-voice unison mode, it's too small for leads with long release or pads)

+Want to see ability of oversampling 2x, 4x, 8x, 16x ...

+SuperPulse osc, or better, SuperPulsaw osc and SuperWavetable osc... ))))

+At least 8-voice unison... (in pair with increasing voicing limit to 128 voices)

+Precission control of values when assigning modulations to modules - 0.01 step with holded on SHIFT button, also writing values numbers from keyboard.

+Mute (or on/off) button on oscillators, LFO's and noise generator, and gain knob! It is the MESS when modulating 10 osc's with noise, and need to increase or decrease mod. level - i need to tweak mod. level in all 10 osc, instead of one knob in noise gen... LFO has level control - so it is good to have this control in all signal generators.
Also ability to modulate gain of generators - like in LFO module (It's may partially solve the problem with difficult realisation of modulation of mod. levels on inputs of modules).

+Raising down saw waveform in addiction to raising up saw in LFO module.

== Found A bug - then switching saved patch with unison on (in my case 5 voices) to same patch with unison turned off (and contrary) - pan position on voice mixer always loads at center (pan knobs turned in right positions when patch loaded, but all sound is mono) - if in my patch two voice mixers with pan set lo full left and full right - loaded patch sounds in mono instead of stereo - but little touching pan knobs again helps...

[sonigen]

Renaming modules is likely to appear at some point.
Increase in voice count will happen too.
SuperWavetable, Mute / Bypass, and Increased precision on modulation controls are all on the "to investigate" list.

Regarding the 'gain' on generator modules. Would a simple gain module be sufficient? A module with a single input and output and a gain fader. I think a gain control on every generator module would end up redundant in the vast majority of cases as either a generator goes to a single target or the gain is different for each target. So I think a gain module would be more efficient CPU and GUI wise.

Thanks for the bug report. I will look into it.

Thanks,

[SoulState]

Regarding the 'gain' on generator modules. Would a simple gain module be sufficient? A module with a single input and output and a gain fader. I think a gain control on every generator module would end up redundant in the vast majority of cases as either a generator goes to a single target or the gain is different for each target. So I think a gain module would be more efficient CPU and GUI wise.

Stop stoppp, i can use an amplifier module for this! No neet to make all this what i wrote ooh.... i should think a little better lol)))) my bad... yes it is no need to make gain on osc's if it resulting more cpu usage... an amplifier is like "simple gain" already... so is solves the problem. But gain on Lfo's is pretty usable and space saving on patch panel (:
Ok this feature not so necessary than others, so maybe later or don't needed for now at all...

Ok let's go further -

- Found another similar bug - the "restart phase" button on Pulsaw osc wont' saves with the patch, all buttons loads unpressed... on Wavetable and Syncrosc worked fine.

- Issue with changing project samplerate (i have Cubase 5) - if i change it from 96k to 44100 hz or other rates, Sonigen just slows down or raises pitch according to new samplerate, - instead of working at new samplerate with right pitch. It is necessary to reload patch to work right at new sample rate. (Maybe write all current patch parameters in some buffer and recall it after sample rate change, i'm not a programmer so don't know )

- Problem with GUI display - sometimes when project loaded with some plugins (maybe depending on memory - in some projects with big or moderate memory usage from 40%-50% of 4 gb ram) - the GUI wont' loads - it is empty window with all gray color, but sound in working fine. To solve this problem needs to close and restart cubase, just close project and starting new - won't helps. (Similar problem was with some Arturia plugs - Moog Modular and others, but only with heavy loads of ram about 90%).
i just loaded 4 instatces of Sonigen Modular in empty project, played with them few minutes, and closed project, then created new empty projrct, loaded one instance of Sonigen modular - and GUI wont appeared (window all gray)...

- How about writing automation? It is not implemented for now? (automation won't writes automatically when record automation is pressed in cubase 5) for automation i can use only controller parameters. (but can't change any random controller number except few from list)

Some more suggestions -

+Separate Keytrack Controller Module (Want to use for key follow LFO's)

+Raising down saw waveform in addiction to raising up saw in LFO module. (i edited my last post for this, so maybe you won't notice)

+Drive or saturation parameter on filters


And about supersaws/and other "super" osc's -
It is hard to make a stereo supersaw/square whatewer-table osc? (With stereo-spread control and adjustable voice count)
If one mathematical "super" osc costs only 15% of cpu compared to many separate similar osc modules - it is very good, but it is useless when i need control over stereo spread of each super osc separately, global unison also won't helper in this type of patch... (and a two mono supersaw osc's panned in left and right gives not proper result too...)
I have many cool trance patches from synths like sylenth1 or zebra, but recreating it for now or making something similar is a difficult thing in terms of cpu usage - for each one proper emulated "super" osc i need 8 saws/squares/other waveform... my patches usually uses all 4 oscs with 8x unison for each - it is 32 separate oscs in Sonigen - my CPU just go down on one simple chord , considering long release of sound - polyphony for one chord melody is about 512 oscs total

Somewhere on forums one guy wrote what stereo spread in mathematical super-saw osc is not so big deal...

And for spread formula, somewhere here on forums one guy write detune coefficient for Roland's JP supersaw (7 saws) :
" Ronald Supersaw
One in center and:
1st pair: f*(1 +/- 0.019)
2nd pair: f*(1 +/- 0.062)
3rd pair: f*(1 +/- 0.108)"
- this values resulting very smooth sounding detune, but don't know values for stereo spread... Also don't know values for Access Virus Ti Hypersaw (9 saws), but i think it's something close...
A proper emulation of JP supersaw can be found in this plug - jp6k http://www.adamszabo.com/jp6k/, really smooth and pleasing detune sound and special non-linear control knob of detune amount for more precise and smooth control at low values of detune. (heh, but sound quality is not as good as real JP )

Maybe make super osc with adjustable values of stereo and detune spread for each pair of voice (or each voice) by entering amounts from keyboard with three digits after point accuracy for detune... and ajustable count of voices from 1 to 9.

P.S. If it all can be implemented with your forthcoming signal handling update - then it is very good, and i too much worrying about this super-unison-stuff , but dont' know how it be in terms of cpu usage compared to dedicated stereo super-osc's...

And few words about your global unison algorithm - it is the best sounding unison i've heard so far realy, realy cool and pleasing, sounds just awesome, as well as whole synth)))) - sound quality from Sonigen can be compared with hardware monsters like Nord Lead/Modular and Virus, really cool, i don't need anymore save thousands of bucks for these metal boxes)))) Finally here is a softsynth with such high sound quality and it's modular!

[SoulState]

Forgotten one question about multithreaded processing - does one instance of Sonigen modular dynamically uses many cores in multicore processors or just one core per instance?
As far as I can see it's only one core per instance at this moment.
Is it real to make multithreaded processing for one instance of synth? With some complex patches synth's power (polyphony) stucks in one core of cpu

[SoulState]

And one more about drag'n'drop cables idea and pop-up menus for this...
I think a pop-up menu when dropping cable (after dropping cable on module), for cable destination is not really good, but pop-up menu BEFORE dropping cable on module is a great simple solution! Let's menu appear when mouse with cable goes over module - simple!
But this is ever more easy solution -
i have an idea just to drop cable straight on module controls - it very simple) and to be intuitive to new user - just lead round the red colour or highlight this controls/knobs when cable hangs on mouse, and mouse with cable goes over current module. And use for different type of controls different highlight colours - e.g for all knobs affecting pitch - red, for pw - blue, etc.

[sonigen]

But gain on Lfo's is pretty usable and space saving on patch panel (:

IIRC I included it on the LFO because i felt it was needed when you have a level modulation input. If you are modulating the LFO level via something else you might want to modulate up from zero, or down from 1, for example.

How about writing automation?

I'm working on basic automation and MidiCC atm, but it will be pretty basic for now. I want to have more comprehensive automation / midicc support but due to the work involved it's low priority ATM.

Separate Keytrack Controller Module (Want to use for key follow LFO's)

There's a new feature in the works that will add keytrack and other controllers as either sources or modifiers for input signals. But a more complex keytrack module is also on the todo.

+Raising down saw waveform in addiction to raising up saw in LFO module. (i edited my last post for this, so maybe you won't notice)

You can set the mod amount to negative on the signal input, you'll get the inverted shape. Will look at adding falling shape saw too.

+Drive or saturation parameter on filters

Its on the todo but it will likely be a separate LPF module.

my patches usually uses all 4 oscs with 8x

Would 5 voice unison, with unison panning, and 1 supersaw not work? I suspect you wont be able to tell the 35 saws are not perfectly spread across the stereo field.

Supersaw voice count is on the todo.

Somewhere on forums one guy wrote what stereo spread in mathematical super-saw osc is not so big deal...

Its not, but Sonigen Modular doesn't have stereo at voice level so it would have to have two outputs and the user would have to patch left and right signals separately.

I will look into it though.

Forgotten one question about multithreaded processing - does one instance of Sonigen modular dynamically uses many cores in multicore processors or just one core per instance?
As far as I can see it's only one core per instance at this moment.
Is it real to make multithreaded processing for one instance of synth? With some complex patches synth's power (polyphony) stucks in one core of cpu

Yeah unfortunately that is true. And it's actually a large and difficult task to make it spread across many cores. Not because multi-core is hard particularly, but the nature of the routing in a modular synth makes it hard to split it into things that can be done in parallel.

Not saying it wont happen, but its not in the immediate plans.

And thanks for the bug reports!

[SoulState]

Would 5 voice unison, with unison panning, and 1 supersaw not work? I suspect you wont be able to tell the 35 saws are not perfectly spread across the stereo field.

No, it's really good when used as a single osc or i need to detune all osc's...)) My main goal is ability to make complex patches like this - mostly for phat basses: one or several detuned and stereo super-osc for tops (which i can make with unison at this moment) and one pulse/saw osc that goes 1 or 2 octaves lower - for low end, but not detuned at all with global unison and goes straignt on center of panorama, because detuned low frequences is going to out of phase, and low-end bass punch dissapeared sometimes with flanging due to phase substraction. One obivious thing for reaching it is ability to control participation in global unison for each osc, if it real, then, basically, it solves this problem)

Of course i can always use two instances of modular for upper bas and low end separately, but all in one wolud be more comfortable)
Oh! A good idea to make layers too! Layering is really handy for this purposes, just two or more separate synths in one (or maybe with cross modulation, but this is not so necessary) like in Nord mod. or other synths with layering. And ability to use multiple stereo outputs for layers can be good too...

[sonigen]

My main goal is ability to make complex patches like this - mostly for phat basses: one or several detuned and stereo super-osc for tops (which i can make with unison at this moment) and one pulse/saw osc that goes 1 or 2 octaves lower - for low end, but not detuned at all with global unison and goes straignt on center of panorama, because detuned low frequences is going to out of phase, and low-end bass punch dissapeared sometimes with flanging due to phase substraction.

Hmm, ok I see.

Selectable participation in unison is a not technically viable.

Dual layering is likely to appear at some point, although I cant even guess at a time frame right now. Not in the near future anyway.

The main issue I see with stereo voice modules is that there is no stereo signal path in the voice processing. I did have it in to begin with but I removed it because it added a lot of complexity and overhead not just CPU wise but also in terms of keeping the GUI simple and intuitive. I really don't see that decision being reversed at any point.

So a module with stereo output would require separate patching of the left and right signals.

For now you could try no unison and two supersaws, one panned left, one panned right. You'll have to patch each one individually to it's own voice mixer and then use the pan on each voice mixer to pan them. FWIW at some point individual inputs to the voice mixer will be panable, so you wont need to use extra voice mixers. It's a fair bit of extra work but it might get you closer to what you want for now.

Thanks,

[SoulState]

Thank you for answers, anyway, your synth is the best of all softsynths - after trying Sonigen, i simple can't use another vsti synths anymore, because they sounded like plastic crap compared to yours

[randomelement2`]

I was just listening to the audio demos linked on the website. A couple of pieces (3 and 4) in particular really made me go 'Wow'. Not your average synth sounds by any means.

[Jesse J]

I was just listening to the audio demos linked on the website. A couple of pieces (3 and 4) in particular really made me go 'Wow'. Not your average synth sounds by any means.

Yeah, Chris has something really awesome going on here! Glad you liked the demos!

[randomelement2`]

It seems that most times I listen to demo sounds, I keep thinking that I've heard those sounds 1,000 times already (and wonder why people rave about the synths). What I heard here sounded genuinely fresh.