Jupiter 80 VA engine - new or same as GAIA?

The percentage on the last option says it all.

I'm just trying to understand the terminology here, it seems there has been a shift in what VA means. VA, as in virtual analog. I don't believe the JP-80 can possibly be a VA in the old sense of the word (virtually modeled oscillator generation). I think it is more akin to the piano mode or the EP mode of the Fantom, where certain aspects of certain synths are modeled..... not the oscilators, which have got to be pcm based.

Now, if it hurts you to think that the JP-80 is not a VA synth in the "classic" sense of the word, then you probably should stop here. I think the JP-80 is an amazing instrument & Roland is on to something good here, & they may have very well changed VA as we (I) knew it.

But logically speaking, if Roland did create a virtual analoq synth that can get you anywhere close to 256 voice polyphony, then the cost of Nord, Waldorf, Access, Moog, etc... ought to shoot down pretty damn quick. I'm talking $400-$600 for a Virus Ti shoot down real quick.

It has always been my understanding that the computing power required to generate the oscillators restrict note polyphony. But that isn't the case of the Jupiter 80?

How special.

How much would it cost you to get a V-Synth Gt right now?

& we are supposed to believe that Roland put the VSynth (28 voice polyphony) and the Fantom G8 together, minus the pads, minus the time-trip pad, minus the sequencer..... for the same price as the V-Synth GT?

I'm not buying it.

Roland came up with a more cost efficient VA technology..... I can buy that.

Sounds as good as the "classic" form.... sure, why not. Better..... I can buy that as well.

But it being the same technology & all that the Jupiter-80 does, it would have to cost as much as the Oasys.

You can search for the post but Vince confirmed the basic osc waveforms are modelled. That's everything but where it says PCM.

It's no minimoog and its no Virus TI2.

Ok guys,
we now have a scholastic dispute about whether the jp80 is real VA or not. It reminds me on the famous scholastic discussion between middle age monks who tried to discern, how many teeth a donkey has. And they discussed about that for days and weeks until a boy gave them the advice to simply open the mouth of the donkey and count the teeth!
:))
I know that VA is not the same as donkeys teeth, but isn't there a way to find it out whether the oscillator is a real VA one or it just plays a sampled wave? Maybe the jp80 users could jump in and make a test. Whatever the result will be, it won't change how the jp80 sounds.

what Devnor says..

basic oscillators are modelled.. some other waveforms are pcm.. oh, right, just like every other VA..

mojkarma wrote:

Ok guys,
we now have a scholastic dispute about whether the jp80 is real VA or not. It reminds me on the famous scholastic discussion between middle age monks who tried to discern, how many teeth a donkey has. And they discussed about that for days and weeks until a boy gave them the advice to simply open the mouth of the donkey and count the teeth!
:))
I know that VA is not the same as donkeys teeth, but isn't there a way to find it out whether the oscillator is a real VA one or it just plays a sampled wave? Maybe the jp80 users could jump in and make a test. Whatever the result will be, it won't change how the jp80 sounds.

Exactly, it won't change how the JP-80 sounds.

But there is no way they can be modeling the oscillators plus all the extras at this price point, not with the polyphony you can get out of this thing.

IMO, it's got to be something similar to what they did with the Virtual CloneWheels. There isn't any actual tone wheels in there, it's modeled, and a damn fine model at that.

I also read what Vince said, he didn't say modeled oscillators, he said modeled sunth, like they modeled an actual JP-8000, with a few extra features.

Mystic38 wrote:

what Devnor says..

basic oscillators are modelled.. some other waveforms are pcm.. oh, right, just like every other VA..

That's fine, let's go with that.

Then why does the JP-80 not cost twice as much as a Virus Ti? Which only has limited poly? The V-Synth GT only has 28 & they won't even guarantee that.

thunderkyss wrote:

That's fine, let's go with that.

Then why does the JP-80 not cost twice as much as a Virus Ti? Which only has 64 voice poly? The V-Synth GT only has 28 & they won't even guarantee that.

You don't see any difference between architecture of Virus vs. Jupiter?

Guess mojokarma is looking for an another argument...in your bitter little world you couldn't fathom something with with VA & PCM?

Wow.... this argument seems extreme to me. Simply because the price of the JP80 is, well, $3500 USD, this then means the Jupiter 80 uses PCM samples for oscillators? Thats quite a leap there, my friends!

VA = Virtual Analog. The Virtual in that phrase means ANYTHING is possible. Its a virtual re-creation of an analog oscillator. There must be over 100 different algorithms in existence for doing this, with varying results, from crap to near perfection....from basic waves to complex hogs. The programmers are free to employ whatever technologies and processes they feel blends the mix of sound quality, responsiveness, overhead, and process load. There is no such thing as a "True VA". That is defined by each developer independently, and the results may vary indeed. However, there is clearly a difference between a Synthesizer and Sampler (or sample playback device). And that is where we should be drawing the line.

Denvor is correct....there is a huge difference between the architecture of the Virus TI and the JP80. JP80 is FAR more basic....but what it does, it does remarkable GOOD. A better comparison would be between the GAIA and the JP80. GAIA is well known to use PCM samples for its oscillators, making it a dedicated ROMpler all its own....and while I love that little synth (more for its tactile control surface than its sonics), you can clearly hear how it differs from the JP80, and just about any genuine VA synth. There is also a huge difference between Roland and Access. Roland has resources that Access can only dream about. What Roland can bring in at a certain price point, and its distribution channels....makes this comparison very suspect in the first place.

The main thing about a VA is the dynamic real time continuous modeled waveform which is produced in software. It doesn't have to take ridiculous horsepower to create this. But JP80 does have plenty of horsepower. Computing power gets cheaper by the moment....and the JP80 being the far more recent hardware release has a very limited palette of basic VA waveforms. Compared to a Virus TI.....where you have SO much more depth and effects caliber, you also have 16 fully independent VA synths with effects for fully multi-timbral functionality. THIS requires horse power! The JP80 offers far less than a Virus TI where horsepower counts......so its price is not out of line for its function by any means.

At the end of the day.....you can hear and feel the difference by a long mile between the JP80 and GAIA, or Fantom G, or just about any other PCM based keyboard. Again...I ask you think about Korg Radias or MOSS, or Yamaha PLG. There are lots of examples of very powerful dedicated bona-fide hardware VA synths, costing only a few hundred dollars. Radias EXB has 4 part multi-timbrality for $300. I would argue the architecture and complexity of Radias far exceeds that of JP80. Is it so hard to believe that JP80, at $3500, is equivalent to about a dozen Radias cards? I think not. JP80 is a VA. Fantom G is not. Its really that simple.

Devnor wrote:

thunderkyss wrote:

That's fine, let's go with that.

Then why does the JP-80 not cost twice as much as a Virus Ti? Which only has 64 voice poly? The V-Synth GT only has 28 & they won't even guarantee that.

You don't see any difference between architecture of Virus vs. Jupiter?

Guess mojokarma is looking for an another argument...in your bitter little world you couldn't fathom something with with VA & PCM?

The Virus has both....

The V-Synth uses both.

Just to be clear my second statement was responding to mojokarma.

Virus is DSP based on sharc processors which are expensive chips vs. whatever Roland is using. Lots of knobs, buttons and ongoing software development too. You get what you pay for with Access. Fractal Audio does something similar with their line of FX processing.

Devnor wrote:

Just to be clear my second statement was responding to mojokarma.

Virus is DSP based on sharc processors which are expensive chips vs. whatever Roland is using. Lots of knobs, buttons and ongoing software development too. You get what you pay for with Access. Fractal Audio does something similar with their line of FX processing.

100% +1. and remember, those sharc's are a few years old now. Both the cost and the technology is dated. Were Access to release the same hardware today in 2011, it would cost less. When Access releases its next gen Virus hardware.....it will either have less expensive more efficient DSP...or it will have more expensive newer and bigger DSP capacity, and still cost as much as a small house. :)

I just think Roland has found something better than the traditional VA oscillator modelling. SuperNatural is a proprietary technology that builds on Roland's existing sample/synthesis, APR synthesis, & maybe some COSM technology as well, to model the entire synth, not just the oscillators.

This is beyond the traditional model, & less expensive to boot.

Kudos to Roland.

Devnor wrote:

thunderkyss wrote:

Guess mojokarma is looking for an another argument...in your bitter little world you couldn't fathom something with with VA & PCM?

I have no idea what you are talking about. Thanks, my world is fine.
For the case you didn't realize it, the recent discussion is about whether the analog waveforms are modeled or just pcm oscillators.
Now, you can keep discussing whether the technology is these days cheap enough to give you 256 polyphony on VA oscillators, you may discuss what this or that person says (the middle age monks analogy) or you can try it out and make some tests.
On the first page of this topic Artemiy has given two examples how to find it out. That's what I was referring to. Especially the second advice should reveal whether the analog wave is modeled or simply sampled.

And finally I said that all that matters is how it sounds.
I have no clue where you see a problem with my post. You should discuss ad rem and not ad hominem (if you know what it means).