the hardware fetish marketing worked so well on many people that it creates cognitive dissonance to realize that HW and SW are the same thing in the case of digital synths. its good to work through those feelings.

I have the opsix HW and the plugin and I am pretty happy. I dont know what I will use the HW for but I got a full year of everyday use out of it.

could you please record and upload level matched identical patches from hw opsix and sw render?
Very interested in such comparison.
Thanks

If this plugin will be successful, they may release virtual version. MMT is a great VA engine even today, indeed. They actually still sell MMT hardware called Microkorg, the best selling synth in the history.

Hey, Dan. I'm having glitches when the plugin is playing back MIDI. I'm hearing weird artifacts almost like wave sequencing even when I'm just using a single multi-sample with no wave sequences. Is this a known issue? Like I'll hear random percussive noises even in patches without them.

It's not a sound card issue, my computer isn't breaking a sweat and this is pre-effects because these glitches reverberate when the reverb is on for example.



I just did some testing and made a screen recording. It seems to only happen when there's multiple instances of Wavestate Native. It's as if the instances are grabbing sounds from the other instances.

Like here I have several active instances of Wavestate.
Instance A (the one we're listening to) is just two layers of "single multisample" a piano and brass stack (no choir sounds). No wave sequencing.
Instance B is a choir sound that does have wave sequencing. When both are receiving MIDI, I am hearing artifacts in A that are sounds from B. A you can hear in the video. You can hear glitchy vocal sounds.

We are ONLY hearing instance A. Instance B is receiving MIDI but I have it muted so you can hear that we're somehow getting sounds leaking from B to A.

Now, when I stop sending MIDI to instance B then the problem goes away (about 45 seconds into the video). And then the problem comes back when I start sending MIDI to instance B again (at 57 seconds).

I don't know what is going on but it's almost as if the instances are sharing memory space and it's being aliased somehow in a bad way.

You probably should also mention your DAW and OS versions, and which type of plugin format you're using.

It's not reproducible on Logic Pro (using AU plugins), with 5 instances of Wavestate all on different patches, as I just tested. Rather than mute 4 of them I routed them to a separate bus that had the mixer turned all the way down so they were definitely all still active - but the track I was listening to (Iceland Piano Pad) still sounded normal.

I suppose you need to be sure that the sound is coming from instance B and not instance A. if it appears on the output of instance B, then maybe eliminate any possibility of a midi feedback loop, back to any external keyboard and Midi thru-ing back in. I don’t really know any behaviours of your software or your setup other than that, it may partly depend on if you can find other people having similar experiences on different DAWs / setups. I see some others are not finding similar behaviours already. I think VST3 can output midi iirc.

Well, you could do that....Or you could consider that maybe, just maybe, there are other reasons for owning hardware beyond believing it is inherently superior to everything else, and Korg understand this, and thereby understand there is a market for both hardware and software versions of its products, and it's not (despite what some angry forum denizens will say to the contrary) a one vs. the other scenario, and never really was. This is all the more true, I'd say, for instruments that are literally digital to begin with.

Or, you can sell you synths and keep up the bitter, persecuted dude act...

You can also create routings by dragging a mod source onto a destination. When you have time, check out the Modulation section in the manual's Getting Started chapter!

Interesting. I've filed an issue. However, I've just confirmed that it *appears* to work fine on our system with a MacBook Pro M1Max running Logic Pro. So, maybe there's another relevant variable at work. I'll send a DM to get more info.

In my current track I have one track driving the HW opsix and one driving the VST, I cant tell the difference.

Wow! I'm sorry to hear that response. My take is utterly different.

There is some overlap between hardware-based and software-based musicians. There are simple, practical benefits to having a hardware instrument you can just pick up and play, just like there are simple, practical benefits of being able to have multiple instances of a plug-in inside a DAW. Many people, myself included, use hardware and software instruments together because of these different, complementary advantages. For these customers, we've made it easy (to the tune of $49) to pick up the plug-in to accompany their hardware, and get the best of both worlds.

However, there are also many musicians who are firmly in one camp or the other. A lot of the customers I talk to need the visceral, hands-on experience of tweaking knobs, and some of them strongly prefer to keep computers out of their music-making altogether. The hardware products are for these people.

Meanwhile, many of my pro musician friends - people who make their entire living from music - are now completely ITB, and have no interest in hardware synths at any price. The plug-ins are for them.

With the release of wavestate native alongside the hardware wavestate, the instrument can fit into both hardware- and software-based workflows. The instrument is the technology and the sound that it makes; a plug-in is just another form-factor for that instrument.

Offering wavestate as a plug-in, specifically, shows that we value the wavestate as an instrument for professional use.

Dan, is wavestate performing well for you natively in Logic in Monterey on your M1Max test system? Because I'm seeing very poor performance on M1Pro here, natively (Rosetta is ok).

It maxes out Logic's Live Mode core with no notes playing. CPU usage drops when voices are sounding, but on release of those voices, it pegs the core at 100% again.

Monterey 12.3, Logic 10.6.3 -> 10.7.3, MBP 16" M1Pro 16GB.

(Not asking for support as such, just getting anecdotal data really...)

DAW is FLStudio 20.9.1. Windows 11.
Plugin is VST 3.7.2 64-bit.

Sound is definitely coming from A. B is muted entirely in my mixer.

On our system, it's OK. So, let's figure out what's happening here!
A shot in the dark: in Preferences -> Audio -> Devices, what's your setting for Multithreading?

I'm attaching the FLStudio project file with the glitch.

Thanks for the detailed report! I'm sorry that you're having this weirdness, but let's figure it out. What computer hardware are you using (processor, mobo)?

Here's the specs of my machine. Also I'm using a MOTU interface but the problem persists even if I use my Primary Sound Driver.

Thanks. Interesting. I've had reports of other M1Max owners with similar performance behaviour to what I see (and some Windows users too, with the CPU increasing to max out the core when voices are not sounding).

Generally, I choose 8 HP Cores (rather than Automatic), multithreading is Playback & Live Tracks, Buffer is 64 samples, process buffer range is Large. (44.1KHz too.)

Changes to any of these don't make a difference to the behaviour I see, except if I up the audio buffer size to 512, where the core behaviour calms down a bit and becomes "usable" (single-core CPU when wavestat is idle is about 35%, rather than 100% at *any* smaller buffers)

Both on internal MBP audio, and my USB audio interface which is otherwise rock solid...

Quick demo:



At the start of the video, wavestate is idle, no voices sounding, in Live mode, with the core pegged at 100%. I play one note, and cpu drops to 20% or so. I release the note, and the cpu climbs back to 100%. I do this a couple of times.

Then I select a different (empty) track so wavestate is no longer in Live mode, and when idle, is split over two cores. Releasing a sounding note does result in a cpu increase, but not anything that would max a core and cause overloads/glitches, because of the core distribution. I don't know why it uses more CPU when no voices are sounding than when they are.

If I close the gui, and play a note it seems to use even more CPU and causes hard glitches in the audio.

Bugs aside, this plugin is awesome. It's the closest thing to a current-gen Korg workstation sound engine in the DAW.
Even without wave sequencing, the multi-sample library is next level.

The hardware 'experience' would fly more; if they didn't cut more corners than an Origami master on the quality don't you think? I understand that employees have a obligation to bridge the differences and promote - so nothing personal against Dan's commentary here to the slightest.

But I dare say people aren't blind. The reasons for quality commentary all over the internet; isn't just cognitive dissonance. People have a lot of experience now with their keybeds and build quality. From the Hydra/Desktop and the Peak/Summit chuck about; right over to these little boards and beyond. It's been a good many years of the same ol' reviews has it not?

As per my last post it's a question of logic. There was a time when the laptop was almost a cumbersome taboo. Now I can expect to be spending more time discussing apple and ableton than the synths next to it. That's just the direction, the stability and the willingness to use.

Since the keybed is the core expression. A controller that cuts no corners; thats no different than those thinking they shouldn't skip down a tier and lose the quality bed. Only this time they are saving big. How many expensive quality beds does one person need?

A DP or a clonewheel fine. I can feel anyone who says 'I must have the right bed for the job'. But outside of that software all the way.

Your predictions on how much the software will fly; vs a repeatedly hated and cost cut short octave bed that really doesn't do justice to the tech. It is after all why the immediate cry was, 'dude. where is the SE?'.

I've blabbed on enough. Being a fan of tactile HW doesn't mean you need to own that variant of everything. Just use the best bed and the software you actually like. Nothing wasted. Simple.

Totally agree Dan

With Kronos at the end of the road, hopefully we will see some of it's engines make it into plugin formats. All of them would make great plugins

Who has time for manuals???

However, the drag/drop doesn't appear to work for any of the "keyboard" mod sources. And that happened to be the very first mod routing I tried... Aftertouch to Filter Env Decay.

I am able to drag/drop routings from other sources like LFO's.


Now I can no longer reproduce this issue; something previous step(s) must have put the UI in a weird state prior.

Dan I have to admit my intial reaction is a little the same as OnceTwiceMelody. I currently own an Opsix and with the VST being available I'm beginning to wonder if I might not be better off using one of my MIDI controllers with the much nicer keybed, aftertouch etc and selling my Opsix. In the current market here in Australia I'd be $500 in front that I could stick into something else. As a customer I'm in two minds about whether the sliders on the UI are worth the $500 difference between the hardware Opsix and the software one. Especially if the sliders are addressable via CCs or Sysex as I'd be able to map them to knobs on one of my controllers.

Your strategy may end up cannabilising sales of the Opsix and the Wavestate. Especially if the $49 crossgrade price isn't permanently linked to having the hardware. ie I could buy the VST for $49, then sell my Opsix. Would the person who buys it be able to do the same thing? Or is it once per serial or identifying code on the hardware?

Potentially you've knocked the resale value of hardware Opsixs as well, pretty early in the life of the product.

Edit: In the interests of doing some research on the point about selling the hardware in favour of the VST - on our local second hand site, two Wavestates have been listed in the last 24hrs. I guess it could a coincidence, but I don't see them that often. I suspect I'm not the only one who's gone through the same thought process.

So wait, help me understand. Because the software version exists, secondhand prices are dropping? Why?

Why would resale value be affected when there's brand new hardware units available at the same price they've always been at?
If someone wants the hardware they're going to buy the hardware. If they want it new they're going to buy it new. If they want it used then they're probably always willing to pay 80% of the new price. I don't see how a software version affects this aside from there simply being fewer potential second hand buyers.

If the price of the hardware had dropped then yes I could see how that affects secondhand prices.

It would make sense for the vst to be free for people have bought the hardware.

Some would think that the hardware should have some capabilities that the software does not. For me since I dont own a hardware opsix or wavestate, so I dont care but I could see how others would.

I will possibly be buying the opsix vst though. Cant deny. Would be stoked to use it with aftertouch and 61 keys.

Korg OPSIX and DX7 Patches, I know this is possible with the OPsix Hardware, but how's the import process done correctly for DX7 patches for OPsix Vst?

Art the second hand prices will drop because people will question the value of the hardware unit, like I myself am doing. The keybed is terrible and is 37 keys only with no aftertouch. The sliders are great, but potentially could be mapped to a MIDI controller with a decent keybed with 61 keys and aftertouch, which I'd suggest a lot of people ITB or using DAWs will have. This was not an option before. You had to buy the hardware units to get the engine. So, is there $USD600 worth of UI (ie the difference between a hardware Opsix and the VST) and independence of a DAW for a new hardware Opsix or Wavestate? For some maybe, but it would be interesting to understand how many Opsixs and Wavestates are being used independent of DAW to begin with. I'd hazard that a majority do have them hooked up to a computer in a studio.

This then flows into the perceived value in the 2nd hand market, if people don't see the value in the new units. The last couple of Wavestates I saw on the 2nd hand page I mentioned where priced about $AUD100 higher than the ones listed in the last day or so. Again could be a coincidence, but it's an interesting data point.

Edit: Just looked at the manual - CC's 102-113 are the OP Level Sliders and Ratio knobs. So pretty straight forward to map that onto most MIDI controllers. You'd probably have to do two banks on the controller as most only have 8 knobs and there are 6 each of the OP level sliders and ratio knobs, but not a biggie.

I guess the biggest benefit of the hardware is that it will always work. VSTs have a lifespan.

No. They haven't. The hardware version is the software plus a wooden/metal/plastic case around it and some DAC(s). Thats it.

The linux pc underneath it all has a lifespan too. As does any electronic device of course.

I suspect a 64-bit VST3 has quite a lot of life left on it though. Especially if you're on the Windows side where most OS upgrades don't seem to affect such things. I could easily see someone getting a decade of use with it, even if Korg didn't provide any updates starting today to it.

One could even recreate a wavestate to some degree with a dedicated laptop or small form factor PC, a cheap audio interface, and a midi keyboard controller. Not the same by any means as having all the controls laid out is an asset, as is an all-in-one setup. But in theory that cobbled together combination could last indefinitely so long as one could keep sourcing a cheap PC, audio interface, and keyboard.

The UI on a large screen is soooo much easier to use than that tiny display on the actual wavestate hardware as well.

Now to find a midi controller that will do nicely for the Opsix and eventual Modwave VSTs