9 engines, but 5 types of synthesis

[peter_schwartz]

@ X-Trade, AL-1 is indeed an analog synth model just as the MS-20 and Polysix are. AL-1 can and does possess characteristics of analog sounds because it's an analog synth model. And when talking about "analog" we're talking about subtractive synthesis, the tried-and-true model of oscillators feeding into a filter, feeding into an amplifier. And they're all models, as the sound generation takes place entirely in software.

@ EvilDragon...

FM, Subtractive Synthesis, Organ Modeling, Physical Modeling, Sampling, and then there's whatever they're doing with EP-1 and the Piano... I count six types.

You can go even deeper with the analysis, considering that AL-1 includes FM capability and MOD-7, HD-1, and STR-1 all include analog-style filtering. So those are hybrid types, in a sense.

But that said, I'm curious to understand what the point is in comparing the numbers of different sound engines with numbers of synthesis types? Are you trying to suggest that Kronos is somehow "less than" because it doesn't include all known synthesis types? (What workstation does?) Or...

As much as "nine" sounds big, it's really just "five" in practice.

...is it that you think the marketing approach is overblown and deceptive?

Based on your premise, a person could reduce the number of any spec for any specious reason. The spec of a car advertised as "seats 6" could be smacked down because it'll only seat 4 if you bring along two week's worth of camping gear.

Nine engines are nine engines.

[EvilDragon]

...is it that you think the marketing approach is overblown and deceptive?

Something like this. It is slightly overblown, yes, from the perspective of actual number of synthesis "cores" built in (so was OASYS in this regard, well, Kronos draws from it anyways). Deceptive, no, it's 9 different paradigms of work, definitely, but some of those engine share the same characteristics tied to certain very distinctive synthesis types. I am separating "synthesis" from "engine" or "paradigm of work" here.

AL-1 has FM, well, Prophet-5 had cross-modulation which is FM, too, and it's pure analog, and not a hybrid (aside the VCO-DCO disputes etc.). I wouldn't call AL-1 hybrid. It's purely subtractive synthesis.

I could say that MOD-7 is enhanced FM because it has added filtering (yet again a subtractive synthesis element) and a vastly different input waveform selection, borrowing from samples and audio input (and I find this quite a big deal, very good to have for experimentation, but it still remains, at its core, FM synthesis).

And about SGX-1/MDS-1, I'm pretty sure they're just doing stuff you can do in Kontakt as well, with some scripting (you can make seamless velocity layers there, too, but using samples as source material as well). So it still counts as your ordinary samples+filtering (=subtractive synthesis), on steroids, though.

[ravenmek]

No matter if 6, 7 (still my count!) or 9 different kinds of synthesis:

The question was: are other kinds of synthesis, and updates, needed fast?

I just don't see anything important at the horizon.
But I can learn, if someone shows me something really important...

Someone said it in another trhead (x-trade? I don't remember ) but i would love some additive engine, and it would be awesome to have more physical modelling engines (wind, bowed strings?)

[danatkorg]

Now, I hope this thread of mine won't raise a flame war. But let's be straight on this one: Kronos does indeed have 9 "engines", however in total it has 5 distinct synthesis types inside:

Sampler/sample mangling: HD-1, SGX-1, MDS-1, I'm including wave-sequencing here as well
FM: MOD-7
Tonewheel modeling: CX-3
Virtual analog: AL-1, MS-20EX, PolySixEX
Physical model of a string: STR-1

Here's a brief description of the EP-1's MDS:

The EP-1 sounds begin as samples. These samples are processed as a group to create the MDS data, using proprietary techniques. The resulting data can then be played by the EP-1.

And a second, longer one:

MDS is a brand-new technology from Korg. To create the EP-1 models, we first sample the instrument at many different dynamic levels and pitches. Next, we separate the main pitched sound from the hammer and key release noises, so that you have separate control over each. Finally, we use proprietary techniques to transform all of these elements into an MDS sound.

The result delivers natural control over both the velocity and time "dimensions." Velocity becomes a continuous transition from soft to loud and sweet to strong, without the telltale velocity-switching of traditional sample playback. Time becomes flexible, so that you can control the decay and release in a natural way, without resorting to volume or filter envelopes.

* * *

I can't go into technical details (actually, my drafts of the explanations above were significantly more technical -- but the lawyers have the final say!). However, under the hood the EP-1's MDS is quite separate from the others, and scripting + sample playback could not get you there.

Re the MOD-7: calling it "FM" is selling it a little short. It does traditional FM/VPM, of course, but it also does waveshaping (think O1/W, drive/clip, pickup, etc.) and has "vast" possibilities for sample mangling (process a sample through 6 VPM carriers/waveshapers/ring modulators - all doing all three if you wish - plus two resonant multimode filters, in a patchbay environment which provides for easy splitting and mixing of signals). My guess is that this one will be the biggest surprise once synthesists start getting hold of the KRONOS. I hope that you all enjoy it!

[peter_schwartz]

Something like this. It is slightly overblown, yes, from the perspective of actual number of synthesis "cores" built in (so was OASYS in this regard, well, Kronos draws from it anyways). Deceptive, no, it's 9 different paradigms of work, definitely, but some of those engine share the same characteristics tied to certain very distinctive synthesis types. I am separating "synthesis" from "engine" or "paradigm of work" here.

It's true that some of the engines share the same characteristics, like analog-style filtering. But personally I don't see any hype in the idea of advertising 9 engines, because there are in fact nine. And it's not like the type of modeling going on in each engine is a secret.

AL-1 has FM, well, Prophet-5 had cross-modulation which is FM, too, and it's pure analog, and not a hybrid (aside the VCO-DCO disputes etc.). I wouldn't call AL-1 hybrid. It's purely subtractive synthesis.

P5 cross-mod isn't technically the same kind of frequency modulation as found in AL-1 and doesn't produce the same sonic results. And you're right, I wouldn't consider AL-1 a hybrid synth model, but if we're splitting hairs here...

I could say that MOD-7 is enhanced FM because it has added filtering...

...and talking about "pure analog synthesis" vs. something else, or "pure FM" as compared to something else, well... I'm not a purist (I used to be, but eventually I found that outlook to be creatively restricting, and an exercise in futility). So IMO, any synth stands solely on its own merits, regardless of the purity of its design. But more to the point of your thread --- whether or not Korg is somehow "at fault" for calling nine sound engines "nine sound engines". I just don't see it.

[EvilDragon]

whether or not Korg is somehow "at fault" for calling nine sound engines "nine sound engines". I just don't see it.

Calling it nine engines, or nine different working paradigms is ok. But 9 completely different types of synthesis, I'm not so sure.

[peter_schwartz]

Re the MOD-7: calling it "FM" is selling it a little short.

My bad for helping to sell it short

[ozy]

it's still VA in essence,

oh, finally a new theological dispute!

essences, substances, entities!

I'd missed that for the last three days,

since Akos threw in a "Pascal's bet" about the Oasys' future upgrades.

Flame war? Nahhh. This forum is worth a Clash of Civilisations

[peter_schwartz]

Calling it nine engines, or nine different working paradigms is ok. But 9 completely different types of synthesis, I'm not so sure.

To be fair, that's not what it says on their website. It says:

"9 Sound engines, each offering a unique sound-creation technology"

The architecture of the three different VA's is different out of necessity, I would imagine, in order to create different sonic characters (not mention the UI's). So in that sense each of them is "unique".

Is that too much of a stretch? I don't believe so. They're specifically talking about the sound-creation technology, and my interpretation of that is "a modeling architecture that produces a distinctive sonic result from other modeling architectures".

Anyway, I think that if anyone tries hard enough they can find fault with anything. On the flip side, there are those who turn a blind eye to even the most egregious and obvious deceptions. And sometimes, a cigar is just a cigar.

[Kevin Nolan]

Yep, it's 9 synthesis types. It does not say engines on any offical documentation.

Even if Analog modelling is done 3 times, it still makes no difference. They share nothing in common. They are all completely separate types of synthesis which offer completly different functions and sound.

Just my opinion, and clearly KORG's too.

Cheers
James

Indeed - the underlying methodology to the Virtual Analogue synthesis in AL-1 is very different the MS20 and P6. While the latter use component circuit modelling, the AL-1 doesn't. So the way that Oscillators and filters are implemented are very different in AL-1 and the other two. And you can hear it.

I'll stand corrected on this but even the filters in the HD-1 engine are yet again different to the EXi's - because the HD-1 uses digital filters while the AL-1 uses modelled filters (and the MS20 and P6 model actual circuits that behave as filters).

This should not be sneezed at. For example, as with a JD800 which offers very warm PCM bases oscillators (including its sawtooth wave) and very warm filters, it is often worth creating analogue synth sounds on the HD-1 engine because it offers a wide range of vintage sawtooth waves samples and then very warm digital filters. In this regard I've made the OASYS HD-1 engine sound more like a JD800 than say, the AL-1 synth engine (even with sawtooth based sounds).

Irrespective of the actual count of synth engines or types - you are getting a plethora of modern implementations of oscillator sources and indeed digital/modelled filter implementations. And because of features like ultra smooth filters across all engines, wide range of filter types, seamless and quantise-less real-time control of the filters and low aliasing oscillators, you get absolutely first rate options that are wide and deep. So the Kronos, with probably superior DAC's than OASYS and then that very exciting Vintage EXi as new oscillator material, will keep you busy for years. Again I know I'm like a broken record on this but you'll just have to bare it till you get your hands on a Kronos to confirm this - the Kronos is not like other modern synths. It's superior - it's deeper (a LOT deeper) and it will respond to the most demanding synthesist, performer or sound designer among you. You'll be spoilt for real, significant, noticeable (and wonderful) choice - among even just the VA implementations !!!

Kevin.


PS - The AL-1 has the capacity to alter the phase between two oscillators, and to modulate the phase from any one of dozens of sources such as LFO, so it can _really_ sound like a MicroKorg - and sound incredibly vintage (great little feature explained to me by Daz some time back).

[peter_schwartz]

I'll stand corrected on this but even the filters in the HD-1 engine are yet again different to the EXi's - because the HD-1 uses digital filters while the AL-1 uses modelled filters (and the MS20 and P6 model actual circuits that behave as filters).

They're all digital filters.

Regardless of whether one is an emulation of a physical circuit and another is an entirely non-emulative design, they're all digital.

[EvilDragon]

Kevin, not to dismiss anything you've said, it is rather obvious that there's lots to do with Kronos to keep you busy for your whole life if needed.


Too bad the engines aren't 100% modular, so you can take filters from one to another, etc. Well, that kind of flexibility is only offered by Kurzweil, not to mention DSP patching on layer level, along with really wide array of DSP functions. It's a different mentality to Korg, most definitely, and both have their pros and cons.


That's why a serious synthesist should have both! I so far have 50%, with my Kurzweil PC3K8. Will be getting K61 when possible, definitely. I just wanted to debate over this subject a bit

[Kevin Nolan]

Kevin, not to dismiss anything you've said, it is rather obvious that there's lots to do with Kronos to keep you busy for your whole life if needed.


Too bad the engines aren't 100% modular, so you can take filters from one to another, etc. Well, that kind of flexibility is only offered by Kurzweil, not to mention DSP patching on layer level, along with really wide array of DSP functions. It's a different mentality to Korg, most definitely, and both have their pros and cons.


That's why a serious synthesist should have both! I so far have 50%, with my Kurzweil PC3K8. Will be getting K61 when possible, definitely. I just wanted to debate over this subject a bit

Listen - I do not want to get into a debate on this issue so I'll just say this once and not respond again - there is absolutely no comparison between Kurzweil's latest offering and the Kronos. The Kronos is in a different league. This not a bias on my part, its an important statement.

I own a fully kitted out K2500XS, and I have to say, trying the PC3X was like playing a toy by comparison even to the K2500. Indeed, the PC3X keyboard action is about as bad as it gets - it's woeful. Absolutely junk. And the quality of the OS on Kurzweil is nothing short of a disgrace - and I mean an utter disgrace - its a con job.

As just one example (and I could list many) - it is impossible to alter, with ease, even, say, the filter cutoff of a single program on the Kurzweil. If, say a piano has 8 layers, then you have to edit the filter, separately, for each layer! Just how rediculous is that? And - it's not even obvious how to do that - it's utterly cryptic with value settings that mean nothing to me. It's the same for every single parameter of the synth engine. It's impenetrable.

Kurzweil now boast 32 layers in VAST - well even the 8 layers (or whatever it is on the K2500) is a nightmare - and absolute nightmare - to navigate and is why that synth engine has remained virtually completely unexploited by 3rd party sound designers to this day, despite the fact that the same OS has been around for virtually 20 years. Nobody can figure the bloody thing out!

I'm sorry - but Kurzweil have lost the plot. Yes VAST sounds excellent - but it is outrageously complicated and has been abandoned by all but a few die-hard Kurzweil enthusiasts. This is why Kurzweil are in dire straits and do not even have the money for R&D. What was a great company during the K2000 days is now a complete shambles.

So please do not compare Kurzweil to Kronos. You really ought to go and spend quality time on both (OK OASYS for now), and I can assure you that on every level, Kronos (OASYS) will deliver in spades what the Kurzweil simply cannot deliver in practice.

But I'll return to my original point - the keyboard action on the K2500XS is among the finest keyboard actions ever developed, while the current crop of Kurzweil keybeds are worse than the worst Fatar keybeds released when that company first came out. They are an utter disgrace and insult to keyboard players.

In any case; all instrument have their limitations. If you cannot design the sounds you want on Kronos then it is you at fault - not the instrument - it is gargantuan in scope and I can assure you that they will be lining up to purchase it because it delivers, as said above, deep, sophisticated and cutting-edge synthesis with the scope to make NI and Arturia stand to notice - but with close to zero latency at very high note counts. It is frankly ridiculous to be talking about Kronos limitations when in reality this instrument will deliver more options than any other single instrument in history, bar none, and do that with the elegance, sophistication and spontaneity expected of current music professionals not interested in trying to decorate the palace through the letterbox approach still offered by Kurzweil today.


Kevin.

[EvilDragon]

I'll disagree on most post you're making about Kurzweil, because it indeed is a different mentality, but I have no problem with that mentality, and not just that, myself, owning a PC3K8, have reached a completely different opinion through my time with the board.

Indeed, the PC3X keyboard action is about as bad as it gets - it's woeful. Absolutely junk.

Disagree there, right on the start. I really like the keybed on my PC3K8 (Fatar TP40L)- whereas I've heard TONS of people having issues with old Fatar keybed in K2500 series. That's fixed now, it's really smooth feeling.

And the quality of the OS on Kurzweil is nothing short of a disgrace - and I mean an utter disgrace - its a con job.

Another disagreement. OS is stable, no crashes for me at all, everything works just fine, nothing weird happening.

If, say a piano has 8 layers, then you have to edit the filter, separately, for each layer! Just how rediculous is that? And - it's not even obvious how to do that - it's utterly cryptic with value settings that mean nothing to me. It's the same for every single parameter of the synth engine. It's impenetrable.

You say ridiculous, I say more flexible. And indeed, I hate Korg for not using the usual real-life units of parameters everywhere (dB, Hz for starters). This changed my life when I switched from my old Korg to PC3K8. I love knowing that my filter is exactly at, say, 2750 Hz, and that I have 18 dB of resonance on it. That's just how my mind works, and I'd be really glad if Korg too would use real-life values for parameters, or have it as a global option. They just work, and I'm not the only one.

...is why that synth engine has remained virtually completely unexploited by 3rd party sound designers to this day, despite the fact that the same OS has been around for virtually 20 years. Nobody can figure the bloody thing out!

It is also not true that there aren't any 3rd party sets developed. There's quite a number of them. The VAST system is not at all difficult to crack, it's just not as visual as it COULD be, and that poses a problem to majority. Not to me, though (and I'm not the only one). I am absolutely satisfied with that board! I'm really spending some quality time with it, as you say, and am reaching completely different conclusions than you.

This is why Kurzweil are in dire straits and do not even have the money for R&D. What was a great company during the K2000 days is now a complete shambles.

You obviously don't know the story about hostile takeover, forced bankrupcy and massive layoffs that Kurzweil suffered without any merit? :/

So please do not compare Kurzweil to Kronos.

I'm not comparing between them, but what I'm saying is that the two boards will go great hand in hand. Because there's stuff that Kurzweil still does better than Korg. And vice versa. Now if you notice my other posts in this forum, I generally have a liking for what Kronos is, and that is also the reason why I'll get it sooner or later. But I also have lots of love for Kurzweil, too, and that's why I have it now.

[billbaker]

Oh, and while we're at it ask for those missing controllers back:
Theremin Controller / D-Beam
KAOS Pad
Drum Pads
Extended Ribbon Controller
Breath Control
Missing engines... hmm.

Good thought... the next gen [ZEUS son of KRONOS -- or would that be JOVE "It puts the lightning in your hands"] with 7 or 8 OTHER synth engines.

Make a list of unique and ultimately failed synths that had moments of brilliance but which were overcome by limitations to the technology of the day (i.e., memory, user interface, fashion, physical unweildiness...)

MOOG model subtractive synthesis
PPG & or Korg Wavestation Wave Table Synthesis
D-50 Linear Analog sample mangling
KAWAI additive synthesis
Formant Filtering module / Vocoder

THEN --- Include an engine for every synth company that's been riding the same horse for ten or more years... (KRONOS already ate Yamaha's history, so that leaves...) Kurzweil's 20+ year old K2000-on V.A.S.T., Roland's Juno-trhu-XV ROM-pler which has also seen multiple monkey-gland transplants [I'm YOUNG again -- buy ME!!], Novation, Virus etc.

It occurs to me that the inclusion of what appears to be TRUE clone of the DX FM engine in KRONOS may be at least partly due to the license exclusivity of the FM patents expiring, if not passing into public domain.

Which makes me wonder if there might be some Zaibatsu tit-for-tat coming where KORG engines will be available as plug in modules for the next-gen Motif.

Hmmm


BB