Madrona Labs Sumu

[Michael L]

If you’ve used Aalto and Kaivo, then you’ll be used to the workflow. ML does what they do very well, but they don’t deviate from it. I’d imagine Sumu will be popular with sound designers and West Coast synth intelligentsia

I am a connoisseur of Aalto factory presets Hot mess, Wild donks, Adorable gleeps, Muppet dungeon, and Beyond the valley of the wubs.
Also Kaivo factory presets Take your vitamins, Playing with a rubberband, Pots n pans, Smelly tongues, and Arf!
I hope to be partial to similar in Sumu.

[foosnark]

But I'm still kind of bothered somehow with its limit of something like 64 partials per voice, such as if we 'compare' (so to speak) it with other additives with many more.

I think comparisons to other synths -- especially in terms of raw numbers -- are meaningless. But if anything, it's more of a 128-operator FM synth (with one algorithm) than a 64-partial additive synth.

[Borbolactic]

But I'm still kind of bothered somehow with its limit of something like 64 partials per voice, such as if we 'compare' (so to speak) it with other additives with many more.

I think comparisons to other synths -- especially in terms of raw numbers -- are meaningless. But if anything, it's more of a 128-operator FM synth (with one algorithm) than a 64-partial additive synth.

Thanks. Unsure I quite understand that, but if it has some power/depth, then it should show soon enough as it gets fleshed out and/as more people put it through its paces.

[crawlingwind]

Saw this today

https://youtu.be/TtAsnxruIoU?si=m0oJVRqiqo9qpDSt

[Borbolactic]

Saw this today

https://youtu.be/TtAsnxruIoU?si=m0oJVRqiqo9qpDSt

The video was a bit lumpy and what with Randy being half in the frame at the end and muting a sound that kept playing too.

He mentioned getting Sumu out, come hell or high water, by the end of the month... Perhaps that might be a good idea, even if it's a bit lumpy, if it allows people to smooth it out or tame it as that seems to be what it needs. And in the mean time Randy can then work on it some more, and in the process of more of a response/two-way-dialogue from users and not just beta testers, and release fixes and mods and whatnot in the process.

Normally, I wouldn't necessarily suggest early release of something that its creator seems rushed about but in Sumu's case it might actually be good for it/everyone, especially if people start learning it well and even making YT videos of what it can actually do.

[EvilDragon]

I think comparisons to other synths -- especially in terms of raw numbers -- are meaningless. But if anything, it's more of a 128-operator FM synth (with one algorithm) than a 64-partial additive synth.

I would definitely say it's a 64 partial additive synth that has more advanced partials that can be spectrally a lot more interesting than just a sine wave.

[Borbolactic]

Crony-Capitalism: Talent & Technology Grabs & Enclosures & The Slow Murder Of Music Software, Among Other Things

I think comparisons to other synths -- especially in terms of raw numbers -- are meaningless. But if anything, it's more of a 128-operator FM synth (with one algorithm) than a 64-partial additive synth.

I would definitely say it's a 64 partial additive synth that has more advanced partials that can be spectrally a lot more interesting than just a sine wave.

Sonically or auditorally or whatever the word is, I'd like to hear it match or preferably exceed (at least to my ears), something like Alchemy, Harmor, Razor or Loom 2-- all of which have become problematic in different ways.
Then we will have what I consider an improvement over the years those crony-capitalist profit-motive-degraded four have been around.
Loom 2, for example, seems to have been captured, exchanged and quasi-abandoned by too many questionable corporate hands (leading to a few problems with the software that I'm discovering), while Harmor and Alchemy have been captured and enclosed ('land-grabbed') in DAWs ('gated communities').

That's in part why I increasingly appreciate FLOSS software and software otherwise shared...

If Vital/Vitalium or Surge XT could do a certain level/quality of resynthesis & additive synthesis along the lines of the aforementioned (assuming they cannot already) I'd be there in a pinch. In fact, I'm about to inquire at their sites about that shortly...

I've even been looking into and considering PlugData, if you can believe that.

The Murder Of Love

[Andreya_Autumn]

Sumu will be quite different from those other synths you mentioned. Capable of some similar things for sure, but not direct replacement. And also capable of some totally different stuff.

Vital has an additive-style editor for its wavetables. But it's a wavetable synth at its core, not an additive. Try it for sure, though I don't think you'd find it equivalent either.

Surge has nothing that's even close. There's a 16-partial additive mode in the Alias osc. And well... it's in the Alias osc.

[foosnark]

Sumu will be quite different from those other synths you mentioned.

That was my point, really. Having 64 partials doesn't mean it's a worse version of Alchemy, because it's not trying to be Alchemy (or any other "standard" additive synth).

[Neon Breath]

Arturia Pigments’ additive engine has up to 512 partials.

[gentleclockdivider]

Reaktor sine bank can go up to a whopping 10,000 partials , all good if you want to accurately reconstruct-mimic analogue style stuff or just weird stuff in general . .( like razor does)
Otoh the old synclavier had ONly 24 harmonics,and I am having tons of fun with synclavier Go !
More doesn't necessarily mean better , I am sure Madrona labs guy knows what he's doing

[pandafish]

It’s easy enough to synthesize and sum sine waves - the big challenge of additive synthesis is user interface. How do you enable mass editing of partials in a musical, fun, and creative way? Sumu seems like a nice paradigm and also fits into the X+Y concept of prior ML plugins (Kaivo = granular + physical modeling, Sumu = additive + FM).

[foosnark]

Xaoc Odessa can do 2160 simultaneous partials (and their frequencies can be bent inharmonically and/or multiplied/divided by integer values).

Make Noise Spectraphon can do 64 (created from a single sinewave through Chebyshev waveshaping, so they are always exactly harmonic partials and can't be bent).

This does not make Odessa 33.75x better than Spectraphon. They are both awesome, they do different things, they each make sounds the other one can't even dream of. If I had to pick just one it'd be Spectraphon, but I'm glad I don't have to choose.

It is pretty easy to predict there will be people who don't like Sumu. But I also predict, none of those people would change their minds if it had 256 partials instead of 64 (even if that didn't mean needing 4x as much CPU).

[Borbolactic]

Wow you guys rock. And given this is a music site, maybe in the more literal sense too.
I'll have to check out some of the mentions, some that I've never heard before and maybe get back to you, maybe along with some experience with Sumu, who knows. (I hope Madrona Labs offers a trial version.)
I wonder if Pigments does resynthesis or can import the result of another software doing so, rather like how Sumu uses Vutu, if I recall the name correctly.

Edit:

Xaoc Odessa can do 2160 simultaneous partials (and their frequencies can be bent inharmonically and/or multiplied/divided by integer values).

Make Noise Spectraphon can do 64 (created from a single sinewave through Chebyshev waveshaping, so they are always exactly harmonic partials and can't be bent).

Ok, I just looked those two up and was thinking only of software synths, ideally that can plug into a DAW and is/are 64 bits. Both of those seem to be hardware.

I have heard of Cube 2 (the more recent 64 bit upgrade) also, which ostensibly does both resynthesis and additive, but am unsure of its quality as a software, given its age. Along with Pigments, I might check a trial version if Virsyn offers it. It may be 512 partials.
I am also aware of Syne but it looks questionable/funny. Any ideas/experience?
I have LinPlug's Spectral and, while I like it and it's a keeper, it just doesn't seem to have that particular 'additive sound' that I'm looking for that both Razor and Loom do have, perhaps along with another one-- Parsec-- that's nevertheless locked up in a DAW that I'm unwilling to acquire and deal with.

Why is there such an over-wealth of subtractive synths, along with their analogue emulations, all of which have become cliches, but a relative dearth in additives? Even wavetables and maybe granulars seem to be in overabundance.

[Andreya_Autumn]

I think what Pandafish might have something to do with that. Subtractive have some really well-established, well-understood and well-liked UI/UX paradigms which are easy to build on.

Additive UI/UX isn't as "solved", if you will. Which means it takes a lot more thought to do well.