One Synth Challenge #192: Six Sines by BaconPaul (johnkhf Wins!)

[Hanz Meyzer]

So there's a bit of an aesthetics of the project here.

Totally get your point, YET it would not hurt to add some noize I mean, why not? Focus should be IMHO making a nicely usable synth, over aesthetics of computer scientists.

Also the synth already is quite the best freeware fm synth available, more capable than a dx7.

[IV!]

Except in OSC, you hardly have to use one synthesizer for everything) Then you use FM for FM, and for noise... something else. Although I am also sinful in thinking that a filter would be very appreciated here)

[empphryio]

So there's a bit of an aesthetics of the project here.

Totally get your point, YET it would not hurt to add some noize I mean, why not? Focus should be IMHO making a nicely usable synth, over aesthetics of computer scientists.

Also the synth already is quite the best freeware fm synth available, more capable than a dx7.

Would you like a noise patch? It makes noise perfectly fine.

[baconpaul]

So there's a bit of an aesthetics of the project here.

Totally get your point, YET it would not hurt to add some noize I mean, why not? Focus should be IMHO making a nicely usable synth, over aesthetics of computer scientists.

Also the synth already is quite the best freeware fm synth available, more capable than a dx7.

Would you like a noise patch? It makes noise perfectly fine.

I was thinking the same

The self modulation turned all the way up sounds like any other noise source. And as it breaks sounds like no other one.

The filter argument i am thinking more about. But it's not a CS thing. Adding filters would take me a couple of hours. It's a musical design constraints thing! This is an audio rate modulation based synth not a subtractive one!

[whatevergeek]

btw, is it ok to use built-in patches in six sines for the contest? or do we have to create/design each patch we use?

[RichardSemper]

btw, is it ok to use built-in patches in six sines for the contest? or do we have to create/design each patch we use?

You can use built-in patches or your own. Sometimes others will post useful patches which you may also use.

[ElVincente]

So there's a bit of an aesthetics of the project here.

Totally get your point, YET it would not hurt to add some noize I mean, why not? Focus should be IMHO making a nicely usable synth, over aesthetics of computer scientists.

Also the synth already is quite the best freeware fm synth available, more capable than a dx7.

Would you like a noise patch? It makes noise perfectly fine.

I was thinking the same

The self modulation turned all the way up sounds like any other noise source. And as it breaks sounds like no other one.

The filter argument i am thinking more about. But it's not a CS thing. Adding filters would take me a couple of hours. It's a musical design constraints thing! This is an audio rate modulation based synth not a subtractive one!

For what it's worth, I love to use high shelf filter after fm and.audio rate modulations, as it allow you to get a little crazy with mod amount without getting too ear piercing. Per voice low cut and high shelf would be welcome for sound design.
talking about clean eq style kind of filter without 'caracter' curve non linearities or distortion.

[Hanz Meyzer]

+1 for filter, too

[baconpaul]

Well I’m definitely not adding filters during the osc. I think that would obviously be unfair to late comers

But the interesting question is where to place them. I was actually thinking of a filter per op and then a filter on main (so 7 filters just like there are 7 vcas) but the filter per op being in matrix or post matrix is something I’m pondering

That is if op 4 has a low cut and modulates op5 is that filter just in the audio path or also in the mod path

Anyway yeah we have a super high performance snappy neutral svf in the code base. Like I said this isn’t a tech challenge it’s an instrument design one

And no matter what it’s post this osc and post 1.1

[Taron]

BaconPaul, your flow is beautiful and should not be interrupted by anyone, seriously. I love it and support it completely. This is literally fun & games here, but your excellent synth goes FAR beyond this here. So... don't let anyone put a stick into the spokes of your powerfully rolling wheel! 🛞

[baconpaul]

Thanks! Ha! But it's really just 9 days. Anyway I want to ship 1.1 and that has a few other changes I want to hold until after the OSC mostly involving voice management and note expressions.

I have been doing some performance improvements over the weekend thanks to some higher cpu patches a few folks shared with me. A nightly later today should be quite a bit better for patches which use lots of nodes but don't use their LFOs and a few other such conditions.

[Taron]

Has someone been able to send you some crash logs? I've had a few too many crashes today, this time using MuLab instead of Reaper, but apparently it make no difference in this case. When it "crashes" it actually hangs up either DAW so that not even the console receives a crash log.
What I have noticed, however, is that Unisono appears to heighten the likelihood if not even being responsible, but I can't say for sure. However, there is a fairly severe problem. Oh, hmmm... would your menu's log option provide potential answers?
If I knew how to handle XCode, I would've tried to run a debug, but thus far I've only done terminal compiles.

[baconpaul]

If you have a crash log send it over on GitHub
I got a report of a crash when sending Mpe in non mpe mode. But I haven’t tracked it down or repro4uced it

[Taron]

I know, I saw that in the issues. Wished I could get a log, never gets there, though, guess it hangs before it can generate one. Tried the console, of course, nothing in there. But yeah... I know, it's frustrating. Thing I found interesting in MuLab was that a crash literally repeated what seems to have been left in the buffers for a few cycles before hanging dead. Curious thing.

[baconpaul]

I just found it and have a fix coming

If midi events on the same controller came very quickly and it was the only controller the smoother could get in an infinite loop and wedge the audio thread.

Really sorry about that. "Code a linked list" is an interview question I've given a hundred times, but apparently can't pass myself in critical code. Sigh.

New nightly in ~30 mins