Korg multi/poly native - reimagined Mono/Poly Synthesizer plugin

[frag]

That being said, I think this plugin is a good candidate for it.

You think so?
And until yesterday everyone was against it, some people laughed at my suggestion that every VSTi should have multicore switch
They said "Get a new computer!". Well there you go, now even a 2-3 years old machine is outdated. It's ridiculous to run 2048 oscillators on single thread, and implement GUI elements which require latest WIn11 builds. All we need is sound, not fancy graphics.
Mulitpoly would run like a charm on 3rd gen Xeons, even older. In fact, it might run better on these many-thread CPUs than on its own hardware.

But no..... they had to make it single thread in 2025!

I hope someone from Korg reads this. They will have to implement multithreading, just like Arturia did with Pigments.

[Gam456]

Less than 3% CPU usage with a 4 layer, 4 ocs, 8 unisson, full of EFX, 4 seq
Not so bad with 64 buffer/ 48khz on a Ryzen 7950X 96 Go

[tumface]

That being said, I think this plugin is a good candidate for it.

You think so?
And until yesterday everyone was against it, some people laughed at my suggestion that every VSTi should have multicore switch

I do laugh at that suggestion, though.

Well there you go, now even a 2-3 years old machine is outdated.

No idea what you're talking about. I'm using multi/poly on a 7 year old computer. I can run over 10 instances of it, all of them with multiple layers enabled and playing multiple notes.

It's ridiculous to run 2048 oscillators on single thread

Er, why? Works fine. I could make a 8192 sine wave synth that runs on one thread if I wanted.

But no..... they had to make it single thread in 2025!

I remember replying to you about this in the past. You have some fixation with internally-multithreaded audio plugins and how every plugin that doesn't spread itself out over threads is deficient and lazy devs or whatever you said before.

and implement GUI elements which require latest WIn11 builds.

I'm using it on Win10. It doesn't require Win11.

In fact, it might run better on these many-thread CPUs than on its own hardware.

Its own hardware is just a 4 core ARM CPU. It's not fundamentally any different than what your computer probably has in it, except it's lower power and probably use the ARM instruction set instead of AMD64/x86-64 and probably isn't as fast. But it has full control over the hardware and doesn't have to try to play nice with whatever else is going on in your PC.

[frag]

Wow! You're awesome!

[tumface]

Wow! You're awesome!

You attempted to derail this thread weeks ago using this same talking point, ranting about lazy evil devs and conspiracy theories about CPU makers. I replied to you multiple times, including with this in depth post which was upvoted by multiple people: viewtopic.php?p=9029703#p9029703

You responded by flaming out and leaving the thread. Now you're back and doing the same thing, saying the same things, like you're stuck in a time loop. I will be blocking you now, since I have nothing further to add.

[frag]

You actually remember things posted on internet forums? I have no idea who you are.

Btw, I can run at least 24 instances of Multipoly on 12 year old Win7 machine.

Did I win?

[Hanz Meyzer]

Hi,

I am trying to automate the "note on advance" button of the pitch motion sequencer with a note gate signal. Goal is to restart the pitch sequencer with every key press. This works manually, but seems to fail with the gate signal automated at "1".

Any idea how to achieve the goal? The also is no "Note-Off" source...

EDIT: Also automating "Loop End" doesn't seem to work as expected.

[frag]

ranting about lazy evil devs and conspiracy theories about CPU makers.

This tumface is your typical "stick head in sand" guy.
Who the f*** blocks something on internet forums? It's not real life. Grow up.

But it's good he reminded me about these earlier posts!!!

Multipoly and some other new VSTis are literally designed to make you want to buy new PC. It's shocking how many PCs, mobile phones and other devices are thrown away, even though they are in great condition. So yeah, it is "conspiracy", but only if you don't understand the way market and companies function. They live from your money. Not using multicore even though hardware uses it, is a joke on the users.

Anyway, apart from some ridiculous multi-layer sequences with long release and dynamic voice allocation, I can optimize every patch and play it on "old" machine. So that's nice to know.

[frag]

Less than 3% CPU usage with a 4 layer, 4 ocs, 8 unisson, full of EFX, 4 seq
Not so bad with 64 buffer/ 48khz on a Ryzen 7950X 96 Go

Yes, well, I am also getting 3% CPU usage on a CPU from 2012.
Report ASIO buffer usage on 7950X with preset Constellations.
That's what we want to know.

[HTT]

I am using a 2020 Macbook Pro Intel i5 2GHz. Per Geekbench's posted results, with my limited understanding, my CPU is singe-core blessed (upper i7 level) but multi-core performance is average i5 level.

On my system, one instance of Multi/Poly Native is maxing out the CPU with Ableton Live 11, significantly worse than any of my other AUs/VSTs. With Logic Pro, there is moderate CPU (25%) use with one instance, about Diva-level. For both DAWs, Multi/Poly Native is significantly more demanding on the CPU than either Modwave Native or Wavestate Native.

I do quite like the sound of Multi/Poly but the softsynth does not really differentiate itself compared to Modwave or Wavestate (or Diva). Add in the Multi/Poly's inordinate demands on my system, and I am a pass. I will investigate once the bugs/inordinate demands are fixed and there is a larger sound library that gives me something that I do not already have covered.

[db3]

Multipoly and some other new VSTis are literally designed to make you want to buy new PC.

Perhaps the native version is actually a cunning ploy to increase multi/poly hardware sales

[mentooool]

___________VCA TYPE
changing vca type has no effect on sound even when pushing drive; yes, the amp envelope times & curves change but what about level/saturation?

From my quick testing with the demo I got the impression that the drive is a standalone effect that doesn't change with VCA type as you describe. But I believe the VCA saturation is driven from the Level control. Try jacking the VCA Level up and compensating with the output volume. It's not a full-blown distortion at any setting, but it's doing something.

nice tip! i'll give it a try

[Funkybot's Evil Twin]

Version 1.0.2

• Added more right-click targets for copy/paste, including the Sequencer title in the menu bar and the Kaoss Physics and Arpeggiator section titles on their edit page.
• "Enter chord from MIDI" now turns off automatically when Chord is turned off.
• Previously, Oscillator 4's wavetable mesh was drawn incorrectly when X-Mod was enabled in Double mode. This has now been fixed.
• The Sync To Host button now clearly displays the on/off state.
• Previously, when Sync To Host was enabled on, changing Performances could re-enable local tempo. This has now been fixed.
• Previously, closing the UI while playing notes could cause a crash. This has now been fixed.
• Made minor cosmetic improvements.

[kritikon]

Finally had a quick play with it today. First impressions:

1. Fkn ridiculous demo implementation - downloaded it, loaded it up, then got told I needed to also download the factory data which was a silly rigmarole to locate the folder. The one it wanted was completely different to where the demo was and made no sense whatsoever, had to make my own folder after finding multiple times that the one it pointed to didn't work. WhyTF couldn't the demo include the factory data ? Took me 15 mins just to get the fkn thing working. Not a good start.

2. Initial load up was a vast screen that took up my whole monitor and couldn't find the dots to reduce it - again took 5mins to be able to actually use it. Fkn stupid.

3. Actually sounds pretty good. The new filters are mostly great. Lovely SEM filter, and the MS filters seem a bit more realistic (i.e. unstable and screamy in a musical way).

4. Quite like the GUI - simple but it works IMO.

5. Didn't use up much CPU even on my 3yr old laptop. Was expecting much more from the previous topic shenanigans. Well done Korg for not f**king it up like the Wavestate/Opsix release.

6. Hmmmm I think I might veer to the hw rather than Native. I prefer Wavestate Native for programming rather than hw. Prefer Opsix hw over Native (although Native is f**ked so I can't use it anyway). Things like the chaos pad are silly in sw, make more sense in hw and the hw looks more knobby and intuitive than the sw. I wasn't expecting that- initially really diddn't entertain buying the hw.

7. Still not entirely sure I wan't it at all. Ultimately it's a VA. Not sure I see the point. Nice step up in Korg sw though...

[Hanz Meyzer]

I think Korg could benefit to do a semi-public open beta phase first with a lot of more people and a feedback system. It is very natural for such a complex project to suffer from bugs and developer blindness at the beginning... The client here is the customer, remember... Doing mistakes is normal and natural, even human nature. The problem is how companies often deal with that, pretending it is a perfect product, made of stone. If it would be openly communicated instead, like "hey it's the initial version, there might be bugs, but you can report it here, and we are happy to fix it", why not?

Having no feedback with the customers usually leads into frustration, not the bugs... I mean, there are even very positive examples of companies in this forum, how to do it the right way.