The No.1 Website for Pro Audio
K3000 from Kurzweil in 2017 How Likely ?
Old 4 weeks ago
  #631
Gear Guru
 
EvilDragon's Avatar
If it's release velocity, it won't happen because FX are not polyphonic, so they can only be modulated by monophonic/global modulators.
Old 4 weeks ago
  #632
Gear Maniac
 
YashN's Avatar
Quote:
Originally Posted by EvilDragon View Post
Yeah, really. No other vendor has such deep MIDI implementation as Kurzweil.
Missed the point: putting MIDI Capabilities above the mind-boggling power-house that is V.A.S.T. is a bit odd...

Quote:
all the hoops you have to jump through in certain cases (things get REALLY complicated if you want to tweak a single parameter across 10+ layers say, and you don't have mod slot or FUN to spare so you can't link it to a slider... you get the point. Why can't we have a simple offset editing mode for selected layers?).
Correct, some things could be improved, especially something like this. But then you expect them to know how their synths are going to be used in all cases.

I do need something like that as I used the Kurzweil to build my own hybrid synthesizer. Therefore, in mono Osc mode, I have duplicate layers I need to tweak simultaneously.

Just for editing, I use a work-around, that of working solely on one layer until I need them all and then I duplicate.

This kind of thing is good feedback to transmit to Kurzweil, explaining why they are needed, instead of narrowing down these machines to some subset of their capabilities and inabilities as if everything revolves around these.

Quote:
What I do like is Kurzweil's effects.
Oh really? Not the V.A.S.T. engine? Fairly sure you miss my point here again.

Quote:
Yeah. It only took them 20 something years to get partially to the point Yamaha was in 1989 with SY-77
You forgot or you ignore that Yamaha specifically prevented them from including the FM algorithms in their machines for the longest time?

Quote:
there's still some R&D happening over there that is not just for developing bloody home pianos for Asian market...
Try coding an engine like that. You think it's easy?

Last edited by YashN; 2 days ago at 03:42 PM..
Old 4 weeks ago
  #633
Gear Guru
 
EvilDragon's Avatar
Yamaha's FM patent expired already in 1995, from then it was a free-for all, but not for Kurzweil apparently (for example Korg started using it in Prophecy/Z1, later OASYS PCI and actual OASYS etc.).

Quote:
Originally Posted by YashN View Post
Missed the point: putting MIDI Capabilities above the mind-boggling power-house that is V.A.S.T. is a bit odd...
It's because I have even more mind-boggling powerhouses right on my computer. Namely Reaktor. On which I've actually learned synthesis (back when it was still called Generator). I'm not missing your point. I just think VAST is... a decently flexible synth engine with certain limitations that should have already been lifted considering the advancements in processing power over decades. Nothing to fawn over in 2020 IMHO. In 90s it was all the rage, though!

Quote:
Originally Posted by YashN View Post
Just for editing, I use a work-around, that of working solely on one layer until I need them all and then I duplicate.
Sure. But that doesn't help when you return after some time to edit that same patch.

Quote:
Originally Posted by YashN View Post
This kind of thing is good feedback to transmit to Kurzweil, explaining why they are needed
They have been made well aware of these things years and years ago, back in Sonikmatter times.

Quote:
Originally Posted by YashN View Post
Try coding an engine like that. You think it's easy?
"Boohoo it's hard" is not an excuse. Kurzweil has smart enough people onboard. But Hyundai is scared of risks so focus on home piano for Asian market it is.
Old 4 weeks ago
  #634
Gear Maniac
 

It's the sound.

'nuff said.
Old 4 weeks ago
  #635
Gear Guru
 
EvilDragon's Avatar
Plus aliasing.
Old 4 weeks ago
  #636
Lives for gear
 
syntonica's Avatar
Quote:
Originally Posted by Tim Kleinert View Post
It's the sound.

'nuff said.
I think that was true 30 years ago, but not so much now. I sold my K2vx when I got my PC3k and immediately regretted the decision. The PC sound just wasn't as sweet and I could hear artifacting in the various patches. To somebody in an audience listening, I doubt they could ever tell, but with a set of cans, I definitely could.

Having sold the PC3k6, the only thing I regret is the money I lost on it thanks to Kurzweil and their price drop.
Old 4 weeks ago
  #637
Lives for gear
Quote:
Originally Posted by EvilDragon View Post
If it's release velocity, it won't happen because FX are not polyphonic, so they can only be modulated by monophonic/global modulators.
How about envelopes - what are they classified as?
Old 4 weeks ago
  #638
Gear Maniac
 

Quote:
Originally Posted by EvilDragon View Post
Plus aliasing.
Aliasing is an unavoidable byproduct of an open-ended flexible digital sound generation engine, and it's part of the skill set of the competent sound designer (or at least it was...) to either avoid it completely, or otherwise design it out of the system.
Old 4 weeks ago
  #639
Gear Guru
 
EvilDragon's Avatar
Quote:
Originally Posted by felis View Post
How about envelopes - what are they classified as?
They're per voice. Well you could use ASR2 if you enable Globals, at least.

Quote:
Originally Posted by Tim Kleinert View Post
Aliasing is an unavoidable byproduct of an open-ended flexible digital sound generation engine, and it's part of the skill set of the competent sound designer (or at least it was...) to either avoid it completely, or otherwise design it out of the system.
Sure but there are ways to mitigate it at implementation stage with oversampling and/or bandlimiting (especially required for non-linear processing) which not many hardware synths are actually doing (except I suppose Nord Modulars and Solaris which are running at 96k, but even that is oftentimes not enough depending how rowdy you wanna go). I mean, I know you know this stuff. If the whole engine were running at 192k we'd get (using current chips that Kurzweil has) much lower polyphony but also much cleaner processing with less junk. Provided their chips can even support 192k processing. Would be nice to at least have an option.

So yeah let's avoid use any non-linear processing because it creates too much junk. Meanwhile waveshaper coming natively with FL Studio has an 8x oversampling option and sounds... well, "it's the sound".
Old 4 weeks ago
  #640
Gear Maniac
 

Quote:
Originally Posted by EvilDragon View Post
So yeah let's avoid use any non-linear processing because it creates too much junk. Meanwhile waveshaper coming natively with FL Studio has an 8x oversampling option and sounds... well, "it's the sound".
Oversampling always introduces latency due to the required up- and downsampling from and back to the base system sample rate. Not so much a problem in a ITB situation, where there's always a fair amount of latency due to the non-RTOS nature of the system anyway, and also the host (DAW) software being able to make up for bespoke latencies so that all signals ultimately arrive at the designated outputs at the (appropriately delayed) correct times.

Not so much possible in a scenario like VAST or any other (semi-)modular system, where downstream signals bypassing certain modules are potentially summed together again or processed together in some other way. If any specific module within such a framework asks for oversampling in any way, you're fvcked. Don't even think about feedback, lol.

Only solution is to insanely oversample the entire system. But if current Kurzes would do, let's say, the 8x oversampling you touted and thus lower the voice count from 128 to 16, you'd definitely be complaining about that instead.

If "sound" is about specs and oversampling factors, then Kurzweil isn't the right thing for you I guess.

Last edited by Tim Kleinert; 4 weeks ago at 12:33 AM.. Reason: typos n stuff.... it's late
Old 4 weeks ago
  #641
Lives for gear
 
psionic11's Avatar
I'd like to naively jump into this conversation.

Would you both agree that the Kronos engines have some very high quality non-aliasing components? At least this is how the various manuals portray their oscillators and filters, and I have to agree that the way aliasing is minimized is well done.

VAST is a wunderland I've yet to explore with my new / old K2500RS, and I'm looking forward to the possibilities. But I can already tell from my initial experiments that Korg is *at least* a generation beyond Kurzweil -- in both DSP tech as well as UI experience and its pragmatic implications as well. This holds not only for Kronos but also KingKorg and Volcas and the 'logues.

VAST is great and all, and I think it has so much potential. In my mind, the order of greatness for modern hardware DSP and tech goes Korg, then Kurzweil, then Roland, and a distant 4th for Yamaha. Reputation, UI, market, and functionality notwithstanding.

Old 4 weeks ago
  #642
Gear Guru
 
EvilDragon's Avatar
Yes, Kronos by default has much cleaner DSP from the get go (this includes waveshapers and hardsync, just much cleaner than on Kurz) . KingKorg starts aliasing pretty badly once you dime the resonance. You know which synth didn't suffer from any of this? AN1x, 1997.


If Kurzweil resurrected their VA1, made it more modular, threw a couple of LENAs at it, allowed audio rate modulation throughout the system and ran it at 192k ("oversampling it"), if it had like 16-32 voices total I would buy it sight unseen and unheard.
Old 4 weeks ago
  #643
Lives for gear
 
psionic11's Avatar
So then why do modern Kurz's still have aliasing issues? Have they not ramped up internal oversampling as CPU / DSP horsepower has ramped up?
Old 4 weeks ago
  #644
Gear Maniac
 
YashN's Avatar
Quote:
Originally Posted by EvilDragon View Post
Yamaha's FM patent expired already in 1995, from then it was a free-for all, but not for Kurzweil apparently (for example Korg started using it in Prophecy/Z1, later OASYS PCI and actual OASYS etc.).
How is Korg even comparable with Kurzweil here? Did you hear what I said: they were explicitly prevented by Yamaha from including FM in their synths for years.

OTOH, Yamaha actually OWNED a controlling interest in Korg at one point, meaning they could share whatever they wanted with them.

Quote:
It's because I have even more mind-boggling powerhouses right on my computer. Namely Reaktor.
I've used Generator and Reaktor too. When I'm working on my Kurz, I don't care about software at all.

Quote:
I just think VAST is... a decently flexible synth engine with certain limitations
Good, you win understatement of the year 2020.

Quote:
that should have already been lifted considering the advancements in processing power over decades. Nothing to fawn over in 2020 IMHO. In 90s it was all the rage, though!
When was Cascade Mode added? When was Dynamic Vast added? When was high-speed Flash streaming of samples added? When was FM Added?


Quote:
Sure. But that doesn't help when you return after some time to edit that same patch.
And? Do you think suddenly it makes my Kurz worthless?


Quote:
They have been made well aware of these things years and years ago, back in Sonikmatter times.
Where's Sonikmatter now? Should we care?

Quote:
"Boohoo it's hard" is not an excuse. Kurzweil has smart enough people onboard.
Have you ever coded and maintained large IT systems? Priorities can get shifted for development as resources are finite and timelines cannot be extended indefinitely for a new product launch.

If they shifted things further away from consideration that very few people want in favour of major re-hauls or additions, that fits within normalcy.

As I said, naysayer...
Old 4 weeks ago
  #645
Gear Maniac
 
YashN's Avatar
Quote:
Originally Posted by EvilDragon View Post
So yeah let's avoid use any non-linear processing because it creates too much junk. Meanwhile waveshaper coming natively with FL Studio has an 8x oversampling option and sounds... well, "it's the sound".
FL Studio (or any SW) => Computer SMPS. Do you even know what kind of junk this causes through the audio chain?
Old 4 weeks ago
  #646
Gear Maniac
 
YashN's Avatar
Quote:
Originally Posted by psionic11 View Post
Would you both agree that the Kronos engines have some very high quality non-aliasing components? At least this is how the various manuals portray their oscillators and filters, and I have to agree that the way aliasing is minimized is well done.
Should be compared to the Kurz machines which include the V.A. Oscs. Do they also have optimsied V.A. Filters, not sure but someone will chime in.

Quote:
Originally Posted by psionic11 View Post
VAST is a wunderland I've yet to explore with my new / old K2500RS, and I'm looking forward to the possibilities. But I can already tell from my initial experiments that Korg is *at least* a generation beyond Kurzweil -- in both DSP tech as well as UI experience and its pragmatic implications as well
You should characterise which version of V.A.S.T. you're comparing it to. V.A.S.T. today is much vaster than V.A.S.T. in the 90s...

Quote:
VAST is great and all, and I think it has so much potential. In my mind, the order of greatness for modern hardware DSP and tech goes Korg, then Kurzweil
Does Kronos allow me to build any synth architecture I can come up with in my mind today?

A modern V.A.S.T. synth does.
Old 4 weeks ago
  #647
Gear Maniac
 
YashN's Avatar
Quote:
Originally Posted by psionic11 View Post
So then why do modern Kurz's still have aliasing issues? Have they not ramped up internal oversampling as CPU / DSP horsepower has ramped up?
Do the V.A. oscs alias? Not sure, but the old DSP osc can for sure, upper registers mostly. Hasn't prevented many from enjoying their synths all these years, just as digital aliasing in many Roland, and other big name synths didn't over the years.

Yes, we can do better nowadays, we have the technology.

Here's a better question if you're an end-user:

How can I build an architecture within the synth to counter this?

Think method that Korg uses in a DW-8000 which is a hybrid.

Since the V.A.S.T. allows use of custom multi-samples, can't you do that too?

How is that an issue then?

If the V.A. Oscs still have some aliasing, then maybe they could improve this in future versions.

In the meantime, for end-users: Can this be mitigated somehow?

Does the use of analogue filters post Osc help?

etc... etc...

Let the naysayers focus on their narrow nitpicked feature (object management, some aliasing in some cases), build stuff, listen, tweak, compose, have FUN instead.

Last edited by YashN; 2 days ago at 05:02 PM..
Old 4 weeks ago
  #648
Gear Maniac
 

Quote:
Originally Posted by EvilDragon View Post
Yamaha's FM patent expired already in 1995, from then it was a free-for all, but not for Kurzweil apparently (for example Korg started using it in Prophecy/Z1, later OASYS PCI and actual OASYS etc.).
... for example Korg started using it in DS-8/707 ...
Old 4 weeks ago
  #649
Lives for gear
 
psionic11's Avatar
Quote:
Originally Posted by YashN View Post
Should be compared to the Kurz machines which include the V.A. Oscs. Do they also have optimsied V.A. Filters, not sure but someone will chime in.

You should characterise which version of V.A.S.T. you're comparing it to. V.A.S.T. today is much vaster than V.A.S.T. in the 90s...
While the K25's only have non-aliasing versions of sine, saw, and square, the PC4 has 11 versions.

I only had my hands on a Forte for about a month. So just enough time to get familiar with the sounds, presets, and navigation of VAST. It wasn't as intimidating as it's reputation would have you believe, but of course at the time there were a lot of mystery black boxes, not knowing having enough time to explore the algorithms and FUNs.

Quote:
Does Kronos allow me to build any synth architecture I can come up with in my mind today?

A modern V.A.S.T. synth does.
You've mentioned that with VAST you can build a basic wavesequencer. The Kronos wavesequencer can do sample round robins, wavetable, granular, and transwave synthesis. You can choose to use internal ROM sounds, or you can use your own processed sounds from its very capable sampler.

Many have also mentioned how they've created a Karplus Strong synth using VAST. Well, the STR1 string physical modeling synth is fully-spec'd. You've got individual strings and pickups, various filter types, pluck sources, harmonics and damping control. You can even use samples to excite the virtual strings. And the results range from natural sounding strings to bells to eerie soundscapes. So there's two engines you can use as starting points to build various synth types.

But it doesn't stop there. You'd probably be pleasantly surprised by the vast number of synth architectures you can build inside a Kronos, which is *much* more modular than commonly known. Its internal routing is very flexible, so synth engines with audio IN can process the output from other synths (VAST cascade mode).

The are 3 synths that accept audio IN.

The semi-modular (polyphonic) MS20 has a functional ESP section, which grants external sounds access to its characteristic dual HPF/LPF, or any other modular patch point, which then can be routed out again for further processing (a la cascade).

The AL1 is a very able synth designed from the ground up with the tools to emulate analogs, which is partly why I said the Kronos is a generation beyond even dynamic VAST. The DSP OSCs in VAST offer a good deal of capabilities, but with AL1 it's a snap to sync an ENV-shaped PWM with a velocity-driven detune saw that is FM'd by an octave and blended with a rectified Ring Mod version of those, with the sub flipped in phase to draw out different harmonics. And using keytracking and aftertouch to dial in those LFO rates assigned to OSC waveshape and detune. The multi-filter is a granularly controlled dual SVF, where LP/HP/BP/Dry amounts and polarities can be crossfaded with its other "poles". Then it can be linked/unlinked in serial or parallel with the second filter, and per-voice drive and low boost to further shape response to your desired character. All achieved rather intuitively thanks to the UI and the mod matrix. My head boggles how to achieve that with VAST DSP blocks and layers.

The MOD7 engine is the most modular synth in the toolkit. Its more than a simple FM engine, with free wiring, multiple sub mixers, multiple osc waveforms and audio IN, multi filters, precise control of phase, ring mod, and FM intensity ("brightness and timbre"). You can route an operator into a filter, out to another waveshaper operator to a submixer receiving a custom 3-op block. Each engine can be routed to FX, then routed back into the 2nd engine. MOD7 can be set up to emulate a DX7, any TX series, and even the SY/TG series...and can probably do almost all of what Kodamo Essence can do (minus the freeform drawing of oscillator waveforms). Its waveshaper engine has 101 wavetables (!) with programmable drive, offset, and key slope, and can be freely intermingled with the phase modulation and ring mod portions of the engines. A bit more evolved than a VAST Saw + Wshp block. With audio IN, you can even FM / waveshape your custom wavesequences.

I've used the MOD7 engine and combi mode to create an additive synthesizer and a poor man's version of a Synclavier.

The modulation matrix of the Kronos is top of the class. Its AMS mixers can do most of what FUNs can do, but more importantly, almost every other synth parameter can be a destination. Therefore a single LFO can simultaneously control dozens of parameters, each with its own amount and polarity, and if coupled via an AMS mixer, variously triggered or modulated by velocity, aftertouch, switch, sequencer, gate threshold, etc. This includes multiple live controllers like ribbon, pedals, and vector joystick as well as FX parameters.

As you can see, the Kronos is very modular indeed, so creating your own synth architecture is readily done. And of course, like in VAST, you can assign the sliders, buttons, and knobs to whichever parameters you want, in essence creating your own physical panel for a roll-your-own synth. Also like VAST, this is brought to next level using the multi/combi mode to have the "layers" interact with each other (Reaktor-like). VAST riffs are Korg RPPRs, there are multiple internal step sequencers with smoothing for animating parameters. VAST "live mode" is possible not only with audio IN and internal re-routing, but also via external processing via SPDIF and USB... the latter most powerful when coupled with AUM and the IOS world of synths and FX.

This machine is more than a gigging board with smooth sound transitions and a nice Set List function. It's a sound designer's deluxe playground.

Horses for courses. I'm still looking forward to getting to know VAST much better. I simply mention all the above for context and perspective.
Old 4 weeks ago
  #650
Gear Maniac
 

Not meaning to be a nitpicking ****** but the FM patent was by John Chowning at Stanford and then exclusively licensed to Yamaha in 1975. NED re-licensed it from Yamaha for the Synclavier. AFAIK the K2000 could already do it, via the now well-known SAW+ SHAPER chain. My only beef with this original Kurzweil implementation of FM is that SHAPER seems to be a low order Maclaurin approximation for the sine function rather than a sine LUT, which makes for that trademark Kurzweil "grungy" FM sound. At least that's my personal assumption based on the fact that the 1-block SINE oscillator is impure (fewer approximation terms) while the 2-block SINE is spot-on. I've even tried to build a purer sounding sine SHAPER by rolling my own Maclaurin series approximation via cascaded xGAIN/MIX mults/adders, but I got fed up fine-tuning the coefficients with the dB scale. Probably could be done though with enough patience and fiddling.

One often overlooked area where the Kurzweils shine is sample playback. I tested it by making a 2-sample loop of plus/minus unity gain and transposing it down into the hearing range. The result was a perfect sine wave. Most interpolators would produce some extra odd harmonics in this scenario. Don't know how they do it (no, it's not an independent digital clock per voice, a ludicrous myth that I've seen floating around here) but it's superb.
Old 4 weeks ago
  #651
Gear Guru
 
EvilDragon's Avatar
Quote:
Originally Posted by YashN View Post
OTOH, Yamaha actually OWNED a controlling interest in Korg at one point, meaning they could share whatever they wanted with them.
Yes, Yamaha owned Korg until they had runaway success with M1 allowing them to buy themselves out of their clutches. Which means when Yamaha's patent on FM expired in 1995 Korg or anybody else could do FM no problem without Yamaha's lawyers breathing at their necks. Same was true of Kurzweil - since 1995 Yamaha had no legal grounds to hold back Kurzweil or anyone else for that matter regarding FM. I thought that was clear enough from the way I stated it.

Quote:
Originally Posted by YashN View Post
FL Studio (or any SW) => Computer SMPS. Do you even know what kind of junk this causes through the audio chain?
I do and I don't hear it (thank you, Furman power conditioners). On the other hand, I do hear Kurzweil's aliasing when doing the same waveshaping that sounds WAY CLEANER in software doing exactly the same thing but better.

Quote:
Originally Posted by Tim Kleinert View Post
One often overlooked area where the Kurzweils shine is sample playback. I tested it by making a 2-sample loop of plus/minus unity gain and transposing it down into the hearing range. The result was a perfect sine wave. Most interpolators would produce some extra odd harmonics in this scenario. Don't know how they do it (no, it's not an independent digital clock per voice, a ludicrous myth that I've seen floating around here) but it's superb.
That is true, Kurzweil's sample playback is pretty great. MARA chip specs (from service manual) only mention: "64 sample playback channels, each with independent playback rate". Obviously, you gotta have independent playback rate if you want to repitch your voices or add pitch modulations... but the way in which it was done is anyone's guess. Wouldn't it be some sort of sinc interpolation in that case?

Quote:
Originally Posted by psionic11 View Post
My head boggles how to achieve that with VAST DSP blocks and layers.
This, in a nutshell. I'd really like to know how many layers would something like what psionic explained here about AL1 or MOD-7 take with VAST. And then you don't even have all the waveshaping tables that MOD-7 offers, nor the same immediacy of editing all those parameters...

And lest it not be forgotten, MOD-7 was first introduced in OASYS version 1.3, in year 2007. It is practically unchanged in Kronos.


EDIT: Actually I've found a pretty old post by Dan from Korg's USA R&D regarding MOD-7 approximation in VAST form: http://www.korgforums.com/forum/phpB...=399121#399121

Last edited by EvilDragon; 4 weeks ago at 11:55 PM..
Old 4 weeks ago
  #652
Lives for gear
 
psionic11's Avatar
Quote:
Originally Posted by EvilDragon View Post
EDIT: Actually I've found a pretty old post by Dan from Korg's USA R&D regarding MOD-7 approximation in VAST form: http://www.korgforums.com/forum/phpB...=399121#399121
Good post, it also answered a few of my questions on the differences between the K25/K26 and the PC series.

Also, this quote is straight from the Kronos manual on the MOD7 chapter. Dan also wrote the manual, and was part of the OASYS PCI team.

Quote:
Originally Posted by MOD7 Overview
“Vast” sample‐mangling capabilities, with
incredible flexibility and power. Use samples as FM
modulators, and create incredibly rich processing
environments combining multiple stages of
filtering, waveshaping, and ring modulation,
patched together however you’d like.
Sounds like Dan was pretty familiar with VAST, judging from his post. I wouldn't be surprised if they had a Kurzweil keyboard on hand while laying the DSP groundwork for the future Oasys and Kronos.
Old 4 weeks ago
  #653
Gear Maniac
 
YashN's Avatar
Quote:
Originally Posted by psionic11 View Post
Also, this quote is straight from the Kronos manual on the MOD7 chapter. Dan also wrote the manual, and was part of the OASYS PCI team.

Sounds like Dan was pretty familiar with VAST, judging from his post. I wouldn't be surprised if they had a Kurzweil keyboard on hand while laying the DSP groundwork for the future Oasys and Kronos.
Looks like he was indeed, just read that part of the manual a few minutes ago and he has quite in-depth knowledge of the Cascaded + Dynamic V.A.S.T. engines.

He also gives a little diss (only simple DSPs processing or something like that).

But the end of the text says volumes.

Did Kurz know that Korg would release that Mod-7 engine? No. But the new V.A.S.T. is sufficiently flexible to allow you to rebuild something like Mod-7 within half of it (15 layers on a max of 32 layers possible).

What kind of combinatorial explosion do you get with the rest of the layers' possibilities?

I'd have more to say on posts above, but no time right now. Will revisit at one point.

A lot of comparisons should be made with Forte/PC4, and I see people really confused about 'Kurzweil's original FM implementation'. There wasn't one: there was going to be one, but Yamaha vehemently refused to let them release the Kurz with the full FM implementation.

So, some clever people understood how to do simpler forms of FM using existing V.A.S.T. DSP blocks - see Paul Dillen's DLN Sounds.

Calling this 'Kurzweil's original FM implementation' is disingenuous at best and comparing to that is equally disingenuous.

If you need to compare, then have a look at the Forte / PC4 FM implementation, and how that can be used with modern V.A.S.T. or not - some more knowledgeable people can chime in here.

BTW, also in the Kronos manual: 'V.A. does have aliasing in certain extreme conditions'...

Yeah, much like everyone else (although Clavia might be doing a stellar job in this department - there's a thread around where a user replaced his need for analogue Polyphonics with a Nord Lead 4. it does seem to sound good).

Now if someone would start whining about the Kronos' aliasing as if it's the end of the world and without mentioning all the good things you can do with it (Mod-7, AL-1, MS-20 + Polysix + PM), that would be a bit odd, wouldn't it?
Old 4 weeks ago
  #654
Gear Maniac
 
YashN's Avatar
Quote:
Originally Posted by psionic11 View Post
I wouldn't be surprised if they had a Kurzweil keyboard on hand while laying the DSP groundwork for the future Oasys and Kronos.
Should be required studying for anyone who manufactures KBs.
Old 4 weeks ago
  #655
Gear Maniac
 
YashN's Avatar
Going to be brief here for now:

Quote:
Originally Posted by psionic11 View Post
You've mentioned that with VAST you can build a basic wavesequencer.
Where exactly did you get this info from that wavesequencing you can build in V.A.S.T. is 'basic'?

Quote:
Many have also mentioned how they've created a Karplus Strong synth using VAST.
How many people exactly? And when?

Quote:
Well, the STR1 string physical modeling synth is fully-spec'd. You've got individual strings and pickups, various filter types, pluck sources, harmonics and damping control. You can even use samples to excite the virtual strings. And the results range from natural sounding strings to bells to eerie soundscapes. So there's two engines you can use as starting points to build various synth types.
Not fully specced and never even mentioned by the manufacturer and devs themselves on Kurz, but I did something like that 20 years ago.

This is testament to the flexiblity of the architecture sometimes (unforeseen abilities). In this case, LiveMode with KDFX...


Quote:
My head boggles how to achieve that with VAST DSP blocks and layers.
Wait till you see how much you can do with it. But, since you already know Kronos well, it wouldn't surprise me if at one point you want Cascade + Dynamic + the other types of synthesis.

Quote:
I've used the MOD7 engine and combi mode to create an additive synthesizer and a poor man's version of a Synclavier.
Both doable with V.A.S.T.

Quote:
The modulation matrix of the Kronos is top of the class. Its AMS mixers can do most of what FUNs can do, but more importantly, almost every other synth parameter can be a destination. Therefore a single LFO can simultaneously control dozens of parameters, each with its own amount and polarity, and if coupled via an AMS mixer, variously triggered or modulated by velocity, aftertouch, switch, sequencer, gate threshold, etc. This includes multiple live controllers like ribbon, pedals, and vector joystick as well as FX parameters.
Not sure how much AMS is like FUNs, but I don't see anything mentioned above that can't be done on V.A.S.T. Everything modulating everything is called a Mod Matrix (a rather complete one at that). FUNs allow you to take two different inputs and effect a Mathematical Operation (from a list of several) on them before routing them. And in reality, you can chain FUNs like that for even more complexity.

There are also KDFX FUNs...
Old 4 weeks ago
  #656
Gear Maniac
 
YashN's Avatar
Quote:
Originally Posted by EvilDragon View Post
I do and I don't hear it (thank you, Furman power conditioners). On the other hand, I do hear Kurzweil's aliasing when doing the same waveshaping that sounds WAY CLEANER in software doing exactly the same thing but better.
The conditioner helps but is still not enough.

Quote:
EDIT: Actually I've found a pretty old post by Dan from Korg's USA R&D regarding MOD-7 approximation in VAST form: http://www.korgforums.com/forum/phpB...=399121#399121
Pretty interesting info. Thanks for sharing.

We can also see the old non-jaded EvilDragon several times in that same thread, can't we?

Used to be more enthusiastic.
Old 4 weeks ago
  #657
Gear Guru
 
EvilDragon's Avatar
Yes that was 9 years ago before the limitations started to annoy me more than the good things it has. In fact, since that was 2011, I had some experience with friend's PC3x, and in 2013 I got one of the first PC3K8s in Europe.

I became jaded because the good things Kurz has are actually even better in various software plugins, and since I don't discriminate between hardware and software, it's fair game to compare one with another and just say that Kurzweil is lagging behind as far as quality of DSP algorithms is concerned. Tech debt caught up with them (in fact when modulation speed is concerned, it already caught up with them in second half of 90s when certain VAs got quite respectable modulation speeds without ridiculousness of aliasing LFOs that can't properly reach 10 Hz etc).

Quote:
Originally Posted by YashN View Post
Calling this 'Kurzweil's original FM implementation' is disingenuous at best and comparing to that is equally disingenuous.
How so? It was for the longest time the only way to do FM on Kurzweil. Of course that should be a point of comparison and not something that should be simply forgotten.

Quote:
Originally Posted by YashN View Post
But the new V.A.S.T. is sufficiently flexible to allow you to rebuild something like Mod-7 within half of it (15 layers on a max of 32 layers possible).
Sure, and lose most of polyphony in the process, while also NOT being as fast or intuitive to edit as MOD-7. Not really a win for Kurzweil here. Meanwhile MOD-7 retains quite enough polyphony for the DSP muscle it lifts. That's the point here.

Quote:
Originally Posted by YashN View Post
Where exactly did you get this info from that wavesequencing you can build in V.A.S.T. is 'basic'?
It IS basic compared to what you can do with wavesequencing in OASYS/Kronos. For starters how do you even pick out exactly which waveform you want to use for which step of wavesequence?

You may hack it somehow but it is NOT intuitive, fast and easy as on OASYS/Kronos.

Last edited by EvilDragon; 4 weeks ago at 10:08 AM..
Old 4 weeks ago
  #658
Lives for gear
 
psionic11's Avatar
Quote:
Originally Posted by YashN View Post

Did Kurz know that Korg would release that Mod-7 engine? No. But the new V.A.S.T. is sufficiently flexible to allow you to rebuild something like Mod-7 within half of it (15 layers on a max of 32 layers possible).
It is indeed a testament to VAST that it can partially re-create MOD7. The last algorithm on the TG77 also allows something like free wiring, so in some sense it can also partially re-create MOD7. But both architectures miss many crucial elements -- multiple mod sources and destinations, envelope curve amounts, control of operator phase, polyphonic unison.

If you take various synths and try to re-create the same sound on them, even using simple 2-op FM blocks, you'll find that they behave differently as you increase the modulation amount on the operators. That is, FM synths (or primarily non-FM synths like VAs that have some FM capability) can sound different, depending on the DSP code used to create FM. This manifests itself in differing reactions to increasing feedback, and in the quality of the "harmonics" various amounts of modulation will generate. MOD7 has a couple macro parameters that somewhat address this (brightness and timbre), and tweaking them ekes out various "harmonic profiles" that might be useful when emulating other FM synths.

And of course a large part of MOD7 is the waveshaper engine. All the variation found here is in the subtle use of drive, offset, and modulation of those shaper tables. The sonic results live somewhere between wavetable and FM synthesis. I don't know of any other synth engine that offers the sonic range of this particular form of synthesis, and that includes Korg's own previous incarnations, upon which the Kronos Waveshaper FX is based.

It is respectable that partial re-creation can be done with VAST. It is equally impressive that there are many features unique to VAST that cannot be re-created on other systems. But VAST, like the Kronos, has certain limitations that are, in some cases, annoying at best, or crippling at worst.

Pointing them out is, at the least, merely an objective statement of fact, and need not be dismissed or characterized as nitpicking or being disingenuous. Great praise for something should not be so overwhelmingly absolute that it masks out or ignores useful constructive criticism, right?

Workflow and UI has improved since the birth of computer programming due to those criticisms and developer response to address them.

Quote:

So, some clever people understood how to do simpler forms of FM using existing V.A.S.T. DSP blocks - see Paul Dillen's DLN Sounds.

Calling this 'Kurzweil's original FM implementation' is disingenuous at best and comparing to that is equally disingenuous.
On the contrary, I think it's a compliment to the flexiblity of VAST that any type of useful FM synthesis could be created.

Quote:
Now if someone would start whining about the Kronos' aliasing as if it's the end of the world and without mentioning all the good things you can do with it (Mod-7, AL-1, MS-20 + Polysix + PM), that would be a bit odd, wouldn't it?
It is frustrating that the Kronos gets a bad wrap by many due to its UI. There are other serious limitations and entirely missing features as well -- piano roll, a list of what modulation controllers are in use and where, tiny font, etc ad nauseum. For many, it's a deal-killer, so they move on to other synths.

But I think the whole point of this back and forth between power users is of a different nature. VAST is an ingenious system of implementing digital modular. Kronos is a mature platform that serves many markets. Both are great, neither is perfect. When mostly satisfied users of a synth frustratingly point out certain weaknesses, I think it's fair game. In today's immediate feedback cycle between developers and users, this is a useful method to iterate a product to deliver something to make the user base happier (applies to synths, smartphones, video games, etc).

Of course, not all companies are equally agile, and there are various issues of marketing priorities, developer resource management, and overall project management and long term strategy. Only Korg and Kurzweil know where they fall here.
Old 4 weeks ago
  #659
Gear Guru
 
EvilDragon's Avatar
Quote:
Originally Posted by psionic11 View Post
It is respectable that partial re-creation can be done with VAST. It is equally impressive that there are many features unique to VAST that cannot be re-created on other systems. But VAST, like the Kronos, has certain limitations that are, in some cases, annoying at best, or crippling at worst.

Pointing them out is, at the least, merely an objective statement of fact, and need not be dismissed or characterized as nitpicking or being disingenuous. Great praise for something should not be so overwhelmingly absolute that it masks out or ignores useful constructive criticism, right?
I think you've hit the nail on the head here - this statement rings very true.

There is no perfect board. And indeed, Kronos also has certain things that are quite infuriating. One that gets me is the transpose - you have to assign the two buttons above the joystick to up/down per patch, when those two buttons could be better used for something else. On Kurzweil, you get semitone and octave transpose soft buttons right below the display, both in Program and in Multi modes... It's always there. Easy. Dunno why Korg couldn't do just that.
Old 4 weeks ago
  #660
Lives for gear
 
psionic11's Avatar
Quote:
Originally Posted by EvilDragon View Post
I think you've hit the nail on the head here - this statement rings very true.

There is no perfect board. And indeed, Kronos also has certain things that are quite infuriating. One that gets me is the transpose - you have to assign the two buttons above the joystick to up/down per patch, when those two buttons could be better used for something else. On Kurzweil, you get semitone and octave transpose soft buttons right below the display, both in Program and in Multi modes... It's always there. Easy. Dunno why Korg couldn't do just that.
Thanks. What's scarier is the statement rings true as well in the wider world of culture and politics, and science, or its denial. But we're living in a post-truth world it seems.

I do admit that the first thing I do when starting an INIT patch is to enter -12 on the transpose. I prefer the lower registers.

Still, it takes all of 5 seconds to tab to the Controllers page and assign SW1 Octave Down and SW2 Octave Up.

I prefer this default method if I want dedicated octave transpose buttons. I think it was a fair enough design decision to leave those two switches open to user assignment, rather than hard coding them to transpose, and then having to go back later and un-assign them if you want other functionality instead. I'm more often using those switches for functions other than octave switch. YMMV

EDIT: I missed your point about wanting an *additional* 2 buttons dedicated to octave switch. Yes, it would have been nice if Korg simply put 4 switches above the joystick, 2 dedicated to octaves. Funny, looking at how the 2 switches are not centered but off to the left side of the joystick, it almost begs for balance by putting in those 2 additional switches.

You still have the option of assigning Tone Adjust controllers to osc transpose.

Or, you could use the new GLOBAL / Function Assign capability which opens up to external CC control of any of the 3 dozen footswitch functions, including Octave Up / Down.
📝 Reply

Similar Threads

Thread / Thread Starter Replies / Views Last Post
replies: 42 views: 31894
Avatar for derneustart
derneustart 24th April 2013
replies: 240 views: 55259
Avatar for YashN
YashN 34 minutes ago
replies: 95 views: 9534
Avatar for abruzzi
abruzzi 7th September 2020
replies: 2384 views: 154137
Avatar for Fleer
Fleer 3 days ago
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
🖨️ Show Printable Version
✉️ Email this Page
🔍 Search thread
🎙️ View mentioned gear
Forum Jump
Forum Jump