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

VST, AU, AAX, CLAP, etc. Plugin Virtual Instruments Discussion
Post Reply New Topic
RELATED
PRODUCTS
Multi/Poly Native

Post

Triton does take a long time to intiialise but it seems stable enough to me. I've never had even the slightest problem with ARP Odyssey, which is one of my all-time faves that I use a lot. My computers always have grunty graphics, I wonder if it's a GPU issue?
NOVAkILL : Legion GO, AMD Z1x, 16GB RAM, Win11 | Audient EVO 8 | Lumi Keys | Studio Pro 8
Korg Odyssey, bx-oberhausen, Proxima, PolyMax, GR8, JP6K, Union, Atomika,
Invader 2, Flow Motion, Olga, TRK 01, Thorn, Spire, VG Iron

Post

GPU is possible. TBH I know little to nothing about internal computer workings. My laptop has a bog standard ubiquitous NVidia thingy though, so unlikely? The whole laptop is pretty mainstream HP with no unusual components that I know of. It's definitely only a Korg thing - no other sw I've come across gives me the shits. As you say, things just work nowadays, or usually do. And it is a shame as I'm with you on the Korg Odyssey, I think it's a lovely sounding plugin and easy to work with.

Post

bmanic wrote: Sat Jan 18, 2025 3:40 am

Of course the exact same audio sources would be run. Duh.. :dog:

I've done these tests before and I'm not at all talking about a super subtle result. Almost anybody who has even a modicum of critical listening skills, which means pretty much everybody here on these forums who are arguing in synth threads over differences between Jupiter/Moog/whatever emulations, should be able to hear the differences.

Weather these differences matter or not is a completely different discussion. The whole point to my argument is that there ARE differences. That's it.
But how would you "prove" that in a thread or forum? Everyone would be listening to it on their own device using their own D/A

And to record the sound of a sound card you would have to record the analog outputs which means the inputs of whatever device you are recording to makes a difference and would color the sound

Not sure how you would prove anything other than your own confirmation bias

On top of that music isn't consumed in a treated room with perfect acoustics and high end D/A conversion and monitors

Most music is steamed over the Internet and played on pretty shitty equipment

So when you comoare the output of a hardware Wavestate to the plugin, once it's uploaded to Spotify, streamed to a phone and listened to on beats headphones with an iPhone no one can tell the difference

The same goes when it's played on the stock speakers of a 2014 Honda Civic, or Apple Air Pods, or through some crappy Home Theater system from Best Buy

Or maybe you press vinyl so someone can listen to it on their $100 Crosley Turntable and crappy 2 way bookshelf speakers they got off Amazon

Post

Even with the best equipment in the perfect environment, they won't know in the context of a mix. Maybe in a direct A/B comparison but the differences would be so slight, so subtle as to be meaningless in any real-world scenario.
NOVAkILL : Legion GO, AMD Z1x, 16GB RAM, Win11 | Audient EVO 8 | Lumi Keys | Studio Pro 8
Korg Odyssey, bx-oberhausen, Proxima, PolyMax, GR8, JP6K, Union, Atomika,
Invader 2, Flow Motion, Olga, TRK 01, Thorn, Spire, VG Iron

Post

BONES wrote: Sat Jan 18, 2025 5:27 am Even with the best equipment in the perfect environment, they won't know in the context of a mix. Maybe in a direct A/B comparison but the differences would be so slight, so subtle as to be meaningless in any real-world scenario.
Very true

Post

IvyBirds wrote: Sat Jan 18, 2025 1:49 am
PAK wrote: Fri Jan 17, 2025 12:09 pm OpSix has spiking issues on the 1st Gen i7 920's too......Once a very popular CPU, it’s 17 years old this year,
Seems like you very much are complaining about Korg not supporting ancient CPUs

That is what I was responding to
The issue, which was being discussed, isn’t isolated to one generation of anything. You appear unable to grasp this. As well as 8th gen, other generations have also been mentioned elsewhere, along with AMD. The response was adding context that the problem went as far back as first gen i7.

Why mention the Triton too? Because it’s another example, I’m aware of, where Korg state minimum system specs which don’t always align with what their code supports. Both issues have been around for a long time, Korg are aware of them, and nobody still impacted likely expects fixes at this point. As such they should change the minimum requirements to reflect what they’re prepared to support. Whilst it's always best to demo something anyway, it helps if users know where the cutoff point is for when something can be reasonably expected to work without issue.

Regarding the remarks you quoted.. Do you see the usage of “once”? That implies “no longer”. Directly referencing the age, immediately followed by a comment (about the Triton) and the fact it otherwise works ok still “being something” implies surprise it works at all on a config of that age. You’re either trolling or, worse, are incapable of deciphering nuance from the written language. Either way this is my last response to you on this thread, regardless of what you post.

Post

PAK wrote: Sat Jan 18, 2025 6:27 am
IvyBirds wrote: Sat Jan 18, 2025 1:49 am
PAK wrote: Fri Jan 17, 2025 12:09 pm OpSix has spiking issues on the 1st Gen i7 920's too......Once a very popular CPU, it’s 17 years old this year,
Seems like you very much are complaining about Korg not supporting ancient CPUs

That is what I was responding to
The issue, which was being discussed, isn’t isolated to one generation of anything. You appear unable to grasp this
But again I was specifically responding to your complaint about a 17 year old processor

You appear to be unable to grasp that

However when you are talking about the 8th Generation i7 that is still pretty ancient as they are now making the 14th generation

I have an 8th Generation i7 laptop. I got a long time ago in 2017.

I don't expect any modern plugin I buy today to run all that great on it even with 16GB of RAM

I still use it to run some older plugins as part of an Audio Gridder Network however

My most recent PC build that I did not to long ago has the Intel Core i9-14900KF. That one flies like the wind

FWIW building PCs is a hobby of mine and I am very familiar with every generation of the Intel and AMD chips

I might not know a lot of things but I do know the hell out of how to build a PC and what CPUs are available

Post

BONES wrote: Sat Jan 18, 2025 2:44 am
frag wrote: Fri Jan 17, 2025 9:49 pmAnd the main question: does M/P Native have multi-threading?
That's not a question I have ever asked myself about anything. In 2025 everything just works, doesn't it? That has certainly been my experience of the last few years. I think maybe when I had an 7th Gen Core i5 was the last time I had any issues with plugins and CPU power, and that was back in 2016-17. In fact, surely it makes more sense for each plugin in a project to have it's own thread, rather than have one plugin hog them all, doesn't it?
Well I disagree. On old machine I run server CPU, and it's still in high end list even after 13-14 years.
Why would I throw it away?
With MT on, it can run hundreds of voices of Diva in divine quality. Most demanding patches, less than 100 voices. Still a lot! Huge!
But then Arturia makes SynthX V which aims at Diva sound quality, but skips MT, which is quite amateurish. There's nothing wrong with i7 920, it's the poorly written software. Cherry Audio is worse, they release completely unoptimized synths.

Korg obviously wants you to buy HW version - with such a shitty interface and no multi-threading, plugin is most useful as librarian, not a standalone instrument.

Post

PAK wrote: Sat Jan 18, 2025 6:27 am
The issue, which was being discussed, isn’t isolated to one generation of anything. You appear unable to grasp this. As well as 8th gen, other generations have also been mentioned elsewhere, along with AMD. The response was adding context that the problem went as far back as first gen i7.

Why mention the Triton too? Because it’s another example, I’m aware of, where Korg state minimum system specs which don’t always align with what their code supports. Both issues have been around for a long time, Korg are aware of them, and nobody still impacted likely expects fixes at this point. As such they should change the minimum requirements to reflect what they’re prepared to support. Whilst it's always best to demo something anyway, it helps if users know where the cutoff point is for when something can be reasonably expected to work without issue.

Regarding the remarks you quoted.. Do you see the usage of “once”? That implies “no longer”. Directly referencing the age, immediately followed by a comment (about the Triton) and the fact it otherwise works ok still “being something” implies surprise it works at all on a config of that age. You’re either trolling or, worse, are incapable of deciphering nuance from the written language. Either way this is my last response to you on this thread, regardless of what you post.
I see old arguebirds is here again. You're wasting your time and breath mate - he'll find something to argue about endlessly. He's an expert you see, knows everything about every kind of synthesis ever made and everyone else is always wrong about something. TBH he's the only one I've ever blocked on KvR. Just makes it easier. All I can see is nice calm blank spaces now. I suggest you do the same if you value your sanity. There'll no doubt be some argument about how everyone one of us who has issues with these Korg plugins are doing something wrong. The fact there are so many people that can't play Korg plugins properly must go to show how many stupid people there are, not anything to do with poor coding. Can't possibly be Korg's fault even though in his own words Korg make shit synths. Circular arguments.

The nice thing with blocking means I know arguebirds is here, and I can make comments that I know will make him really want to argue until the cows come home...but I can't see it, don't get dragged into pointlessness. Naughty, I know, but WTF. :hihi:

Post

frag wrote: Sat Jan 18, 2025 9:38 am
Korg obviously wants you to buy HW version - with such a shitty interface and no multi-threading, plugin is most useful as librarian, not a standalone instrument.
TBH the interface is absolutely subjective. Loads of people detest the Opsix Native GUI, but I find it quite easy on the eye, restful and very easy to work with, though I normally wouldn't go for those kind of flat GUIs. The MultiPoly GUI looks ok to me at first glance. It looks very workable, same as the Wavestate Native GUI, which I initially didn't think I'd like but find it very intuitive for what can be very complex synthesis parameters.

As to multi-threading, I have no idea. Is that what f**ked up Opsix? Would make sense.

Re the hw versions - I assumed they'd want to protect hw and hang on to releasing sw versions for a lot longer. The fact they continue to release equivalent sw in ever shorter time spans suggest that Korg found hw and sw markets are so separated that one doesn't affect the other. I suspect if they released hw and sw at exactly the same time they'd still get normal and expected hw sales. I'm speculating, but it follows with Korg's sales tactics. As a keen hw synth user I know plugins don't even come into the equation when I consider buying hw. I'll consider buying MP Native even after discarding the idea of ever buying a hw version, so they could make MP Native as horrible as they could manage and it would never push me into buying a hw MP :shrug:. For my own use, hardware VA makes not much sense, whereas sw VA does. I don't have anything against digital hw synths per se - it all depends on what they do, how they program and how they sound.

Post

Demo when?

Post

frag wrote: Sat Jan 18, 2025 9:38 am
BONES wrote: Sat Jan 18, 2025 2:44 am
frag wrote: Fri Jan 17, 2025 9:49 pmAnd the main question: does M/P Native have multi-threading?
That's not a question I have ever asked myself about anything. In 2025 everything just works, doesn't it? That has certainly been my experience of the last few years. I think maybe when I had an 7th Gen Core i5 was the last time I had any issues with plugins and CPU power, and that was back in 2016-17. In fact, surely it makes more sense for each plugin in a project to have it's own thread, rather than have one plugin hog them all, doesn't it?
Well I disagree. On old machine I run server CPU, and it's still in high end list even after 13-14 years.
Why would I throw it away?
With MT on, it can run hundreds of voices of Diva in divine quality. Most demanding patches, less than 100 voices. Still a lot! Huge!
But then Arturia makes SynthX V which aims at Diva sound quality, but skips MT, which is quite amateurish. There's nothing wrong with i7 920, it's the poorly written software. Cherry Audio is worse, they release completely unoptimized synths.

Korg obviously wants you to buy HW version - with such a shitty interface and no multi-threading, plugin is most useful as librarian, not a standalone instrument.
Multi-threading within a plugin itself means the plugin spawns worker threads to do computation, instead of doing all computation within the audio thread it's executed on by the DAW/host. There are plenty of circumstances where this doesn't work very well, like with low latency. If I switch on multi-core processing in Dune or Diva with a 128 sample buffer size at 48khz, I can actually get lower total polyphony across multiple instances than if I had it turned off. It can also cause problems as the total system load increases, because the DAW can't be sure how many free cores it has to work with and what's actually being occupied. CLAP tries to solve this somewhat with a worker pool API for plugins, which I think will be a better long-term solution. I don't think internal self-managed multi-threaded processing for synth plugins, as-is with VST, is necessarily a mandatory or good feature today.

From the Dune 3 manual:
Furthermore, it's important to choose a good latency/audio buffer size. We recommend using
between 10-20ms, or 512 samples at a 44.1 / 48 kHz sample rate. On most systems, this should
result in a good balance between low latency, realtime feel and CPU performance. Note that
using less than 128 sample buffers will disable multi-threaded processing, as the thread
synchronization overhead will become too significant.

Post

Hanz Meyzer wrote: Sat Jan 18, 2025 10:08 am Demo when?
End of February seems to be the consensus.

Post

BONES wrote: Sat Jan 18, 2025 2:44 am
I don't think the GUI is ugly, as such, but f**k me! there are way too many screens and pages. It reminds me of one of their hardware workstations, where some parameters are buried 3 or 4 layers deep into the GUI. It is a complete workflow killer and it wouldn't matter how pretty it was or how good it sounded, I'd never have any interest in it.

As is often the case, it's the layers that f**k it up. Without them it would be reasonably straightforward but the way they've implemented the layers means you can't just ignore them, like you can with DUNE, for example. It's a total clusterfuck.
Well, let's see when they release it. Maybe in use it feels different. But yes the layers I'm not a fan of. I much prefer one layer or at most 2 layers but not more. What's the point of those layers?! I don't have only 4 tracks! I can layer whatever tracks and instruments if I want to.

I think Korg is trying hard to optimize their new synths. They are better now with the latest updates but they still run better and smoother on my M1 Mac than my Ryzen 5800x PC (which is not bad really).
Using: Cubase Pro 15, Reason 13, Tascam US-4x4HR, MODX6, DM12D, LaunchKey 49, Yamaha guitar(Pacifica 612v) and bass (BB234) and some virtual instruments and synths.

Post

IvyBirds wrote: Sat Jan 18, 2025 7:49 am
PAK wrote: Sat Jan 18, 2025 6:27 am
IvyBirds wrote: Sat Jan 18, 2025 1:49 am
PAK wrote: Fri Jan 17, 2025 12:09 pm OpSix has spiking issues on the 1st Gen i7 920's too......Once a very popular CPU, it’s 17 years old this year,
Seems like you very much are complaining about Korg not supporting ancient CPUs

That is what I was responding to
The issue, which was being discussed, isn’t isolated to one generation of anything. You appear unable to grasp this
But again I was specifically responding to your complaint about a 17 year old processor

What? Do you mean it won't work on my Atari Mega ST4 :(

Post Reply

Return to “Instruments”