Duh! I mean: no editing of acoustic tones?!?

Forum for JUPITER-80
Devnor
Posts: 696
Joined: 20:22, 27 September 2010

Re: Duh! I mean: no editing of acoustic tones?!?

Post by Devnor »

Major difference between VL70 and the Jupiter is the VL requires a breath controller to shape the sounds vs. Roland doing it by measuring the time intervals between note events to trigger an alternate sound ie behavior modelling.
realtrance
Posts: 128
Joined: 16:39, 8 April 2005

Re: Duh! I mean: no editing of acoustic tones?!?

Post by realtrance »

The more I think about it, the more I can see how this is really a preferable approach. Why should a musician waste a lot of time trying to reinvent the wheel in terms of gestural/performance control when a perfectly viable solution has already been engineered in?

Understanding SN a little more, it's like having a synth full of the ARX cards of your dreams, really, integrated easily with a VA synth that's easy to mix in with SN sounds in a Live Set.

If you're really into experimenting with the details of electronic theory, there's plenty out there for that already: Reaktor, Nord Modular, even iMS-20 whether on iPad or Kronos.

If you're really into making expressive music above all, this is the synth to beat.
ozy
Posts: 169
Joined: 13:32, 1 December 2010

Re: Duh! I mean: no editing of acoustic tones?!?

Post by ozy »

Devnor wrote:Major difference between VL70 and the Jupiter is the VL requires a breath controller to shape the sounds vs. Roland doing it by measuring the time intervals between note events to trigger an alternate sound ie behavior modelling.
did somebody here EVER play a VL-1? Once?

I use the breath controller only for major studio solos.

It's normally driven by velocity+aftertouch (so called "touch envelope")

Bof, whatever.

I got it.

Fanboy forum.

I was looking for serious info, but... whatever.

bye
mojkarma
Posts: 618
Joined: 23:59, 8 August 2009
Location: Varaždin, HR

Re: Duh! I mean: no editing of acoustic tones?!?

Post by mojkarma »

Honestly, you guys are simplifying the VL vs. Supernatural thing a little to much.
First, we are talking here about completely different synthesis methods. Saying that VL equals to some typewriter comparison is completely wrong.
VL is completely mathematically modeled, while the Supernatural synthesis still uses samples. Samples are records. That means that there is still some static, repeating involved.
Yes, playing a convincing trumpet as an example on the VL requires more input from the user, ideally using a breath controller. On the other side, you'll get a more "natural" result to the real thing. There are simply things which the VL takes into account and which can't be handled equally by the SN because the later simply reproduces samples.
Let's take the saxophone as an example. The growl effect can easily be achieved on VL by using a controller (even key velocity) to trigger that hard kinda distorted sounding typical saxophone tone. You wont get that effect on the SN saxes if there are not samples included in the SN model. A lot of nuances simply can't be created with the SN model. Whether you need or don't need those nuances is another story, but don't forget, music is all about dynamic and nuances. Fact is, SN is not necessarily the better model. Let's not forget, it's the approach to push acoustic emulations further to the real instrument. SN gives us certainly much better results than the usual sampling approach we find on current sample based instruments. But it is still not on par with the virtual acoustic model.
realtrance
Posts: 128
Joined: 16:39, 8 April 2005

Re: Duh! I mean: no editing of acoustic tones?!?

Post by realtrance »

mojkarma,

I'm not sure... I don't want to underestimate VL, have played with VL-synthesis as provided by Yamaha on one of its plug-in cards in the past, not using an actual wind-controller with it, granted, but still... my impression is that mastery of the actual control over the nuances of performance takes quite a bit of time and effort.

While it looks like SN synthesis has a lot of the parameters built into the behaviour modelling technology, I don't think that's exactly the same as saying it's all just straight playback of samples?? A sample is a "static snapshot" of a tone, certainly, and its limitation, since the very beginning of sampling, is that it's always going to play back the same way. But once you get down to the granularity of multi-sampling that's now possible -- and this is even before we get to SN, behaviour modelling, etc. -- it's possible to include enough samples for all the variations of the way an instrument is going to play that, in terms of missing gestural detail, the difference between that and full-on physical modelling is almost moot. Moot to the extent that all but the finest masters of a particular non-electronic instrument will not notice the difference.

Now, I _think_ -- without knowing the details of what behaviour modelling technology actually is, but inferring that it's an evolution of AP synthesis, and knowing what that is and how it works on V-Synths -- that once you add in behaviour modelling, you're going well beyond what even very well-designed multisampling can provide, and you're at the point where in terms of actual physical performance of the particular instrument, you'll be able to do nearly everything the actual instrument can do, still, I'm sure, within limits (since the modelling processes must by definition be finite, whereas the actual non-electronic instrument's performative nuances are continuous and therefore, technically, infinite).

Yes? No?
User avatar
PauloF
Posts: 4201
Joined: 02:35, 16 January 2006
Location: Lisbon, Portugal
Contact:

Re: Duh! I mean: no editing of acoustic tones?!?

Post by PauloF »

mojkarma wrote:Honestly, you guys are simplifying the VL vs. Supernatural thing a little to much.
First, we are talking here about completely different synthesis methods. Saying that VL equals to some typewriter comparison is completely wrong.
VL is completely mathematically modeled, while the Supernatural synthesis still uses samples. Samples are records. That means that there is still some static, repeating involved.
Yes, playing a convincing trumpet as an example on the VL requires more input from the user, ideally using a breath controller. On the other side, you'll get a more "natural" result to the real thing. There are simply things which the VL takes into account and which can't be handled equally by the SN because the later simply reproduces samples.
Let's take the saxophone as an example. The growl effect can easily be achieved on VL by using a controller (even key velocity) to trigger that hard kinda distorted sounding typical saxophone tone. You wont get that effect on the SN saxes if there are not samples included in the SN model. A lot of nuances simply can't be created with the SN model. Whether you need or don't need those nuances is another story, but don't forget, music is all about dynamic and nuances. Fact is, SN is not necessarily the better model. Let's not forget, it's the approach to push acoustic emulations further to the real instrument. SN gives us certainly much better results than the usual sampling approach we find on current sample based instruments. But it is still not on par with the virtual acoustic model.
I must confess that I've never played on a VL-1, but after these posts I was curious to see what was the VL-1 all about. And after seeing a lot of demos I now understand what you guys are talking about when comparing VL-1s capabilities to Model sounds to the Roland's SuperNatural approach, although personally as a matter of personal taste (and from what I heard and saw) I would prefer the latter.

Regarding the "No Editing of Acoustic Tones" on the JUP-80, it was already explained here, but from what I could hear and read, you are wrong. It allows for some editing yes, but in a very subtle and innovative way, although limited to what Roland defined as the "most appropriate for each sound" that with the addition of the behavior modeling and can be saved to a live set on any of the 4 layers is a killer.

EDITED: For complete reference see the Jupiter-80 Parameter Guide, pages 25-31 for Live Set Tone Modify Screen (Supernatural Acoustic Tones) and pages 32 for Performance variations for SuperNatural Acoustic Tones and Pages 80-85 for Control Change (CC) assign list for SuperNatural Acoustic Tones, just to give a flavor of what you can edit.

Together with that the JUP-80 allows for the mix of SN Acoustic tones with SN Synth Tones where you can edit any parameter in a very user friendly interface (a la Gaia), together with the Dedicated MFX processors by TONE 4 upper + 4 Lower and that gives a big competitive edge.



I think we are in the presence of a very unique and powerful instrument, although it might not be the right board for everyone (no board is!!!!).
mojkarma
Posts: 618
Joined: 23:59, 8 August 2009
Location: Varaždin, HR

Re: Duh! I mean: no editing of acoustic tones?!?

Post by mojkarma »

realtrance wrote: I'm not sure... I don't want to underestimate VL, have played with VL-synthesis as provided by Yamaha on one of its plug-in cards in the past, not using an actual wind-controller with it, granted, but still... my impression is that mastery of the actual control over the nuances of performance takes quite a bit of time and effort.
My intention was not to point one synthesis as being the better one. I'm just saying that we simply have two completely different approaches to the question how to make acoustic emulations sounding more natural. Yes, VL takes some time. No wonder! If you want to play a trumpet patch, you have to know at least a little how the instrument creates its sound. Put simply, take the JP80, select the trumpet patch and play it like a piano, and it will sound nowhere near to a trumpet. On a general level, we all need to put some effort into learning how instruments sound which we don't play personally, if we want to emulate them. We all need to know how synthesis general works on our keyboards, if we want to go beyond playing factory programmed patches. I'm saying that not as a criticism to the lack of more parameters on SN sounds on the JP80, but more on a general level.
realtrance wrote:While it looks like SN synthesis has a lot of the parameters built into the behaviour modelling technology, I don't think that's exactly the same as saying it's all just straight playback of samples?? A sample is a "static snapshot" of a tone, certainly, and its limitation, since the very beginning of sampling, is that it's always going to play back the same way. But once you get down to the granularity of multi-sampling that's now possible -- and this is even before we get to SN, behaviour modelling, etc. -- it's possible to include enough samples for all the variations of the way an instrument is going to play that, in terms of missing gestural detail, the difference between that and full-on physical modelling is almost moot. Moot to the extent that all but the finest masters of a particular non-electronic instrument will not notice the difference.
I didn't say that SN is just straight playback of samples. But the base of the tone is at the end still a sample. That's simply a fact. It is obvious that SN goes beyond the simple playback of samples, but there are still limits. I already mentioned the saxophone growl sound. If there is no growl sample, the SN saxophone will not be able to produce that typical and often used characteristic sound. Further, there are sampled falls which are used for the SN Brass (BTW, I have the ARX03 on my FG). You have to trigger it with a controller. So, you already have to learn how to put that particular articulation on a controller which is most easy for you to use it and that may be a different hardware controller then what Roland has selected. The Fall articulation that you trigger is a short fall. There is no long fall. You can emulate it with the joystick, but it doesn't sound as good as a sampled long fall. There are other examples. As I said, on the first, it's certainly easier to go ahead and play a convincing acoustic emulation on the JP80 and it certainly sounds much better than everything what we have heard so far from our sample based keyboards. But it still has it limits. A typical trumpet scream FX sound is still nowhere near as authentic on the SN model as it is on the VL or a similar approach (like the MOSS from Korg).
Amazing One
Posts: 98
Joined: 10:48, 10 April 2011

Re: Duh! I mean: no editing of acoustic tones?!?

Post by Amazing One »

Everything I would have said has pretty much been covered by others (realtrance, PauloF, V-CeeOh, Mystic), great post guys!

Just a few things..

One (meaning, anyone of us) may have a desire for a musical instrument manufacturer to modify their current system concerning various elements of a synth so it performs in the way he or she is personally satisfied with, but differs from the manufacturers intended goals, thus putting excess expense on the manufacturers target consumers due to the surplus features one wants and the majority of target consumers (as suggested by market research) will likely find cumbersome, unnecessary and uninviting should they be implemented.

We are all free to have an opinion and hopefully they are taken into consideration with the musical instrument manufactures putting their best effort into pleasing the majority of their targeted consumers.

I do not agree with everything Roland does, but this Jupiter-80 appears to be a serious live performance beast of a machine, perfect for musicians that are skilled at playing the keys and have a need for an advanced, intuitive, highly expressive, user-friendly instrument with incredible sound, performance capabilities and straightforward methods to accomplish desirable but complicated task.

A number of musicians will find a synth that offers the Jupiter-80’s advanced, intuitive, user-friendly features a lot more appealing than a Jupiter-80 that offers a surplus of features they never intend to use with an added expense they are not willing to pay.

I think Roland did a great job at balancing cost (considering the advanced technology and exclusive nature of the Jupiter-80’s specific capabilities along with the state of the economy) with features and (from my view, concerning handling of complicated task) ease of use.

But I share V-CeeOh’s sentiment, it was never my intent to convince anyone that my way of seeing things concerning the Jupiter-80 is best for them personally.

As stated by others, we all have different needs, and there is plenty of gear to choose from, the Jupiter-80 looks like a darn good choice to me.

Thanks guys!
realtrance
Posts: 128
Joined: 16:39, 8 April 2005

Re: Duh! I mean: no editing of acoustic tones?!?

Post by realtrance »

mojkarma,

Agreed with what you're saying. I'm not sure one is going to get the same complexity of actual control over a large variety of expressive elements for say, the flute in JP-80 compared with what you could do with a wind/breath controller and VL...? It's basically the difference between MIDI-controller-based playback of a sample set, including growl tones, and actually influencing a physical model using exciters and resonators that behave pretty much like they do on the modeled instrument. There is no traditional physical modelling going on in the JP-80, rather a specific set of controls over sample set playback triggered by the player's use of keys, velocity, joystick, or whatever the CC-controller representing, say, growl on the tenor sax can be assigned to (including external breath controller).

Still, in both cases, be it physical instrument or behavior modeling, it boils down to MIDI input streams that trigger changes in output that are roughly parallel to those same changes in output that would occur on an acoustic instrument. It would truly be an interesting experiment to put the two approaches side by side and test them against each other, in a way that does justice to the particular strengths and weaknesses of each.
Amazing One
Posts: 98
Joined: 10:48, 10 April 2011

Re: Duh! I mean: no editing of acoustic tones?!?

Post by Amazing One »

The VL-1 definitely brought some innovative developments when it entered the synth scene, though having some cumbersome and impractical setbacks from my view as well.

Yes, you can use a wind controller, but there are various types of controllers on the market for any synth that has a controller or MIDI input such as the one’s on the Jupiter-80.

Furthermore, the Jupiter-80’s controllers (such as the modulation lever or other assigned controller) can produce instrument specific nuances and character changes from the various SuperNATURAL modeling engines depending on the amount of modulation applied (so we are not just dealing with keyboard performance), so in this way, it can perform the same task as the wind controller for the VL-1 but each synth produces different results due to the differences in their behavior modeling algorithms. From my view, the Jupiter-80’s technology and the result produced from the manipulation of the controllers creates a much more realistic result (concerning changes in nuance and character) for a much wider selection of instruments, not to mention that the VL-1 does not have Behavior Modeling so when using instruments such as guitar, strings, piano etc, I feel the Jupiter-80’s method of simulating a convincing acoustic instrument performance is much more intuitive and closer to the real thing.

In short, the Jupiter-80 is much more justified (in my opinion) in it’s belief that it is….the instrument (patch) being played.

But I give Yamaha credit for what they accomplished with the VL-1...

Also, the quote from Vince (keyboards product manager for Roland US) below seems to get overlooked.
Vince wrote:There is no "wave rom" as in what is in the Fantom-G. It's using exclusively SuperNATURAL-based sound engines with multiple behavior modeling algorithms. The 350 PCM waves mentioned are additional PCM waves along with the modeled vintage waves in the analog modeling section.
Devnor
Posts: 696
Joined: 20:22, 27 September 2010

Re: Duh! I mean: no editing of acoustic tones?!?

Post by Devnor »

ozy wrote:
did somebody here EVER play a VL-1? Once?

I use the breath controller only for major studio solos.

It's normally driven by velocity+aftertouch (so called "touch envelope")

Bof, whatever.

I got it.

Fanboy forum.

I was looking for serious info, but... whatever.

bye
Good luck with that envelope. IMO nothing could sound any less realistic than VL patches played w/o the BC. If physical models were all that cracked up to be there would be products at least from Korg or Yamaha but they dumped it too.
DJ RAZZ
Posts: 71
Joined: 05:38, 6 April 2011

Re: Duh! I mean: no editing of acoustic tones?!?

Post by DJ RAZZ »

I ma sure there will be plenty of editing within the vein of the instrument. But this synth is 2 or 3 trick pony. But those ponies are champions. That is why the GAIA on top would be a great addition. Add a Gi, AX-1, RC-30, SL-20 and that new arranger box thingy and call it done. Roland rig from hell.....
Mystic38
Posts: 1105
Joined: 14:04, 24 August 2009

Re: Duh! I mean: no editing of acoustic tones?!?

Post by Mystic38 »

unless you needed the extra keybed then is no advantage in the gaia at all.. the JP80 has a complete VA onboard.. and while you dont get the immediate tweakability of the gaia, you get better sound quality :)
DJ RAZZ wrote:I ma sure there will be plenty of editing within the vein of the instrument. But this synth is 2 or 3 trick pony. But those ponies are champions. That is why the GAIA on top would be a great addition. Add a Gi, AX-1, RC-30, SL-20 and that new arranger box thingy and call it done. Roland rig from hell.....
ozy
Posts: 169
Joined: 13:32, 1 December 2010

Re: Duh! I mean: no editing of acoustic tones?!?

Post by ozy »

Devnor wrote:Good luck with that envelope.
had some nice luck with it for 18 years, but... thanks anyway for wishing me good. It's always appreciated.
Devnor wrote:If physical models were all that cracked up to be there would be products at least from Korg or Yamaha but they dumped it too.
Right.

In the same vein: "If real analogue was any good, korg and roland would still be producing the Jupiter and the Polysix, and instead they dumped them. This means those synths were not so good after all".

That's some flawless logic indeed.

That's the same reason why they don't put polyphonic aftertouch in keyboards: it's because it wasn't so special.

Right.

Aristotle, I'm coming, get out of my way!


[ok, let's cut the cheap sarcasm: the "they had to cut corners in order to keepo the price down" argument is a bit lame, when we are discussing "game-changers" which cost 3000 or 4000 bucks. That's professional league budget, you gotta deliver professional for that kind of buck]
mojkarma
Posts: 618
Joined: 23:59, 8 August 2009
Location: Varaždin, HR

Re: Duh! I mean: no editing of acoustic tones?!?

Post by mojkarma »

Hi ozy,
basically I understand your request and it is as such not wrong in any way. But it has nothing to do with the price of the keyboard. Roland implemented what they call SuperNatural to give the users a possibility to get the highest possible results in emulating acoustic instruments combined with a minimal (let's call it in that way) effort.
The concept of the keyboard is not being a deep synthesizer with the possibility to change every single parameter one can imagine. Sure, somebody wants to create unique sounds by pushing a trumpet sample thru filters and envelopes created for a sweeping slow motion pad, but the JP80 is simply not designed to do this. It is primarily a performance keyboard. It has nothing to do with the VL vs. SN discussion. The way how they (Yamaha and Roland) modeled the acoustic sounds has nothing to do with what you ask. It's simply a design decision and has nothing to do with the question whether VL or SN is a better synthesis method. It's pure factory decision.
If the JP80 doesn't do what you want, it's simply not the right tool for you (as it is not for me because of other reasons).
Post Reply