JUPITER-80

Jupiter 80 VA engine - new or same as GAIA?

Re: Jupiter 80 VA engine - new or same as GAIA?

15 Apr 2011, 11:44
unless of course we want there to be virtual virtual analog synths!.. :D

[quote="Mystic38"]Getting elitist over semantic definitions of how a waveform is created before it gets into memory is just plain silly....quote]

Re: Jupiter 80 VA engine - new or same as GAIA?

15 Apr 2011, 12:02
Mystic, I am not debating about the sound here. But when we talk about technology, there is established terminology that you can't avoid using if you want people to understand you.

An "oscillator", by definition, is a device that is constantly running, i.e. "oscillating".

Re: Jupiter 80 VA engine - new or same as GAIA?

15 Apr 2011, 12:43
TRUST the specs. They can't lie. If the JP-80 had a separate VA or Real analog synth onboard that would be a strong selling point.

Employees wil tell you the synth was made in heaven to sell it, but only the wise consumer can read into what they're actually buying for the sake of FACT.

It is what it is, if a unit plays back PCM samples only, so be it. No need to lie about it. People care only about the sound.

Re: Jupiter 80 VA engine - new or same as GAIA?

15 Apr 2011, 12:58
Artemiy wrote:
In any case, here is how you can tell if a sound generator is sample-based:
1. Detune two waveforms of the same type: you will hear that their phases restart with each keypress, because all samples start playing back from the beginning; on a real VA the oscillator always oscillates so each keypress will make a slightly different sound
2. Hold a note, and do a pitch-bend down an octave or more: you will hear that the sound looses brightness, because the sample's frequencies are all lowered together; a real VA oscillator is always generating new harmonics
That's what exactly I was thinking regarding to the differences between a "real" VA or even a real analog and a sample based one.
Unless Roland uses some special (supernatural?) technique, there cold be audible artifacts when you pitchbend far away from an original tone, because the harmonics will also change respectively. Even with portamento used for some big steps between notes could trigger the same undesired effect.
I personally don't care so much about analog sounds, but I think Roland should talk the truth here and not try to excuse themselves because the VA term is "vague".
After all, any serious musician will decide by his/her ears. Those how need a real VA will probably buy something else.

Re: Jupiter 80 VA engine - new or same as GAIA?

15 Apr 2011, 13:25
Please everyone excuse my lack of knowledge, but is the V-Synth GT a VA? If so, what is the reason why not putting it in the J-80? That would have been the best synth ever no? They wanted to preserve the V-Synth to keep selling it maybe?

Re: Jupiter 80 VA engine - new or same as GAIA?

15 Apr 2011, 16:07
Art,
I wasnt debating with you, more the concept of making judgements on synthesizers based upon the perceived method of producing sound...

Originally the term virtual analog was created to describe synths that could sound virtually like an analog synth, "established" terminology came afterwards...

But, if you wish to bring semantics
1. Wiki clearly defines a VA synth to be restricted to a DSP based system only.. which is clearly bs... a processor capable of processing digital signals in time and frequency domain is not necessarily a DSP.
2. If i have a VA (and I have two..i think!), if i use a waveform, the synth is no longer a VA?.. mm..nope :)
3. An oscillator model running on a DSP is not an oscillator...its just a model.
4. A waveform resident in memory, or directly in the DSP, is simply another model.
5. An oscillator model (or waveform) driving an external analog circuit is not a true oscillator...nor is the external circuit, nor the two together.. there is no free running and no self excitation...self excitation is a key element to an oscillator
6. Clearly in both cases a repetitive desired waveform can be produced with a combination of model input>digital filter1>sum node input 1>DAC="osc output" and feedback loop of "osc output">ADC>Digital filter2>sum node input2... in this case the term filter is used in a software sense as a computational block.

Now, i particularly liked your two tests.. but that leaves the basis of your discussion on oscillators as essentially one regarding the relative QUALITY of modelling done.... not whether there are or not actually real analog free running self exciting oscillator sources in the synth or not.... So I submit that if a DSP (or Atom) based oscillator model (waveform based or not) is adapted to pass your two tests then it will presumably sound better, less virtually and more analog, even though there are still no actual oscillators...

As Mokjama said.. the terminology for VA is increasingly vague, and that is simply because it is being used to describe the details of a digital system representation of an analog system, when in fact the original and true intent is the description of the sound.

Artemiy wrote:
Mystic, I am not debating about the sound here. But when we talk about technology, there is established terminology that you can't avoid using if you want people to understand you.

An "oscillator", by definition, is a device that is constantly running, i.e. "oscillating".

Re: Jupiter 80 VA engine - new or same as GAIA?

16 Apr 2011, 04:40
kday wrote:
Why everyone thinks the JP-80 has a separate virtual analog synth engine in it when it doesn't?

This misinformation isn't even printed on the specs of the keyboard. I would be the first to shout for joy but unfortunately folks, the JP-80 only plays back pcm samples of analog waveforms with SuperNatural Behavior Modelling.

SuperNatural Behavior Modelling is modelling of PCM wave samples and not a separate modelling synth engine capable of creating or modelling full instrument sounds within itself without the pcm waveforms. A real VA or physical modelling synth don't need wave samples like the JP-80 to model or play an instrument sound within itself. If I'm wrong could someone disprove any of this in the specs?

This is just to clear this up here, people are thinking the JP-80 has separate VA or physical modelling synth engines in it, but come to find out they been misled.

Now, not having a VA synth engine onboard in no way undermines the quality of thickness of the analog sounding synth tones the unit can produce. It still will make some of the most gorgeous sounds yet, but don't deceive yourself or others into believing the JP-80 is something it is not.

I've been in touch with a product specialist in Roland who confirmend to me by email last week that there's a separate virtual analogue synthesizer, similar to the V-Synth GT and GAIA, but developed separately for the Jupiter 80 and separate to the supernatural engine.


Kevin.

Re: Jupiter 80 VA engine - new or same as GAIA?

16 Apr 2011, 05:11
With all due respect to you, and to Vince and Dan, I do have my ears,

I respect your ears mate :-). Like I said, I took it as a fact when you said in another thread the synth part was PCM based BUT I thought it made no sense for Roland to do it this way on this particular synth. Maybe we need some more clarification from Roland about this but till then I consider the Jupiter to also have a VA engine.

Re: Jupiter 80 VA engine - new or same as GAIA?

16 Apr 2011, 05:17
Artemiy wrote:
V-CeeOh wrote:
what really makes more sense, even from the sreenshots of the display - the Jupiter is also a VA
Hmmm... well, it's like judging whether the engine is petrol or diesel by the car exterior design.
Yes, I know what you mean but I also trust my eyes ;-).
Why would the Jupiter have a Tone Edit screen where you can select "VA" waves and PCM waves on diferent parameter settings if it was all made out of PCM waves? We could argue that Roland engineers would be playing tricky but I don't think they would go so far.

Re: Jupiter 80 VA engine - new or same as GAIA?

16 Apr 2011, 05:25
I asked this question if the waveforms are all PCM and if the engine is the gaia and Vince replied in this thread:

http://forums.rolandclan.com/viewtopic.php?f=44&t=37042

Basically he says it's a new VA engine with modeled oscillators, and an option to use a PCM sample instead if and osc waveform.

Re: Jupiter 80 VA engine - new or same as GAIA?

16 Apr 2011, 06:18
Actually - I must correct my last post for accuracy. The Roland (employee and not distributor) I spoke to said that the oscillators are virtual analogue. He did not say that the whole engine is Virtual Analogue. It realise there's a lot of confusion over what virtual analogue means - but from what I can determine in general there are two broad categories of virtual analogue synthesizer:

1) Component modelled VA's, as with Korg, where they do not generate direct digital oscillators and filters but instead model transistors, op-amps and so on from which oscillators and filters emerge (and can sound different to straight digital oscillators and filters and deliver more accurate legacy / classis sounds)

2) VA's that digitally generate, in discrete blocks like traditional analogue synths, digital versions of traditional waveforms such as sawtooth that are then routed into digital (IIR / FIR) filters


So I suspect this is the architecture of Jupiter 80 (with two discrete synthesizer engines):

a) a virtual analogue of the second category above - where modelled analogue wave forms and / or 350 PCM waveforms can be fed into digital filters, LFO's and envelopes all configured to represent a virtual analogue synthesizer. Think of this as a combination of t JP8000 and JD800. In fact the PCM samples of the JD800 are so clever that each of the 108 of them is as valuable as its sawtooth (PCM) which gives the JD800 so much better sound design capabilities than many of its contemporaries such as the M1 and T3. Indeed - the 4 faders and many of the black square buttons on the JP80 are of the JD800 style - so Roland are certainly nodding back to the JD800; while the supersaw wave is a nod back to the JP8000. So I think this engine is going to be stunning - and deliver awesome VA synthesis up to 250 voices. Certainly the video of the hotel room where they were playing Howard Jones lead lines really impressed me - it sounded uncannily like a JP8 to my ears and much sharper than the V-Synth or GAIA (accepted it's mp3 quality so take it or leave it on this point)

Actually - a giveaway in the JP80 graphics is the ability to detune the supersaw wave. You can only do that if indeed the JP80 has a supersaw wave in the first place - you could not do this with PCM samples. So this alone would strongly suggest VA oscillators (but as said I have indeed had this confirmed).

Then there is the supernatural engine. I suspect that this delivers a range of articulation offered over a range of say (purely guessing) 20 - 30 instruments, where you view each supernatural instrument like a traditional Roland add-on card, accessible via a large coloured button. I suspect that as you layer and use supernatural sounds, they take away from the 250 voices of the more traditional VA engine.

And since Roland have said all of this has been developed from the ground up, I'd say that the VA oscillators, filters, envelopes and so on are all new.

This is my best guess of the Roland JP80 architecture.

Kevin

Re: Jupiter 80 VA engine - new or same as GAIA?

16 Apr 2011, 06:50
knolan wrote:
Actually - a giveaway in the JP80 graphics is the ability to detune the supersaw wave. You can only do that if indeed the JP80 has a supersaw wave in the first place - you could not do this with PCM samples. So this alone would strongly suggest VA oscillators (but as said I have indeed had this confirmed).
According to what I remember, the supersaw has no detune/mix controls so here is another ball in the "all sample based" park.

Once again - get a JP-80, select one of the supposedly VA waveforms, do a pitch bend down, and you will hear for yourself. I leave my judgement until I have heard it live myself. Until then, I shut up.

Re: Jupiter 80 VA engine - new or same as GAIA?

16 Apr 2011, 07:00
Artemiy -

Check a screen shot of (or download the brochure and magnify) the oscillator section - you'll see the middle of 5 faders in the oscillator section is labelled "Supersaw Detune". But as said - (for the third time!) - I've received an email from a Roland employee verifying that the oscillators contain VA modelled ones.


Kevin.

Re: Jupiter 80 VA engine - new or same as GAIA?

16 Apr 2011, 08:10
Oh yes, you're right! I shouldn't have judged by the prototype's features... As I said, I am really looking forward to what this is. Even if it's similar to SH-201 or sample-based SH-01, it's a great engine.

Re: Jupiter 80 VA engine - new or same as GAIA?

17 Apr 2011, 06:58
So we now know, that the JP-80 is an all digital non VA and non Analog synthesizer, non full physical modelling synth.

Keyboard magazine is doing a poll on their website right now. These are the questions and answers.

Image
What do you think of Roland's new Jupiter-80 synth?

[*] It's one of the coolest synths yet
[*] I'm mad it's not an analog re-issue or a VA
[*] It's very cool but needs more knobs on it
[*] Panel design seems cool for live performance
[*] I don't get it about the panel design, actually
[*] It's a mashup of existing Roland technologies
[*] The whole is more than the sum of the parts
[*] My jury's out until I get to play one

Vote
---------------------------------------------------------
Results

It's one of the coolest synths yet - 16.22%
[16.22%]
I'm mad it's not an analog re-issue or a VA - 11.89%
[11.89%]
It's very cool but needs more knobs on it - 5.41%
[5.41%]
Panel design seems cool for live performance - 1.08%
[1.08%]
I don't get it about the panel design, actually - 3.24%
[3.24%]
It's a mashup of existing Roland technologies - 15.68%
[15.68%]
The whole is more than the sum of the parts - 2.16%
[2.16%]
My jury's out until I get to play one - 44.32%
[44.32%]
185 responses

Refresh

← Back to JUPITER-80