Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Yamaha Avantgrand v/s Roland V-Piano

Thanks for your support, Kona: it's nice to know that my corner of the ring is well-reinforced .

Does anyone know if the V uses true physical modeling? A Korg rep stated that he doesn't think any Roland products use true physical modeling. He believes it probably relies on additive synthesis. If that is the case, it wouldn't be the case that any piano sound can be created using the V because an underlying sample is still the starting point.

good, my whining is rubbing off on somebody..

Additive synthesis and modelling can be practically the same thing. To produce additive synthesis you don't necessarily need an underlying sample. It can be as simple as a sine wave and a basic platform, then you add oscillations and attack/decay envelopes etc until you build up a profile of a piano note. Modelling uses certain algorithms to mimic a physical reality but it can still be implemented using an additive approach. I don't know how the V-Piano works, but I thought I'd mention that. Fwiw, I doubt the V-piano is using samples because I think they wouldn't have had the same issues with the midrange if there was a piano sample under it. I also think that anything "added" to a sample could be pulled back to reveal the basic sample, at least somewhat. Yet this is not the case on the V-piano. You could easily work out whether samples are involved by inspecting the hardware - you could expect to see some sort of mass storage in there somewhere which holds the samples.

Btw, I don't know that you'd want to take the word of any competing manufacturer assessing a rival product. They all have an axe to grind whether they realise/admit it or not.

The distinction was between additive synthesis and pure physical modeling. I always assumed the V was a purely physically modeled sound, but the Korg rep seemed to think that was unlikely. I wasn't saying that the V uses samples for playback but that the initial waveform is taken from an attack sample..... Although I wonder how that makes it different than SN apart from the additional adjustable parameters.

What, then, is 'pure physical modelling'? Additive and subtractive synthesis are well established means of building up a sound from mechanically or purely electronically generated waves (Hammond and Moog, for example). Digital sampling is another thing - I assume digital pianos mostly dissemble and re-assemble samples to a greater or lesser extent, some largely playing back samples in tact, others involving a lot more analytical processing (Yamaha vs Roland?). But now, increasingly, there is this term 'physical modelling' which makes one think of piles of clay and plasticine. But what is it if not 'synthesis' and does it really not use samples at any stage of sound creation?

i loved how when the v-piano first came out you had all these roland rep's/contract musicians talking about how it used physical modelling instead of sampling, as if that was as well understood as the sun rising each day. you could tell they didn't have the slightest clue what they were talking about other than what their literature said. not that i know how it works either, and i've often wondered how the root sound actually is generated....

How very common (everywhere) ...

If they are using genuine physical modelling, there won't be any sample content in the sound. Pure modelling is done mathematically - there is no playback of any kind, it's realtime generation according to the physical model. I suppose you could argue that Roland took a look at real sampled attacks etc, and of course they would have, but that's different to actually using them. More likely, they used that information to refine their modelling. If you are able to change the attack characteristics in a continuously variable fashion, it is evidence of modelling, not sampling.

Modelling is a word used a lot in the music industry and it is rarely defined in a satisfactory way - especially once you add the proprietary techno-spin hype onto it. In the end we have to go back to how it sounds - which is what musicians should be doing.

In the end we have to go back to how it sounds - which is what musicians should be doing.

Exactly.

I actually know a thing or two about modeling. If you really want to know more about it, read some of this below.

http://2uptech.com/intimate_control/RandallJones_MSc_FINAL2.pdf

You can start with the "A brief history" part

Excerpt

Most of the early work on physical modeling of musical instruments was focused on
strings. This is due to a combination of happy accidents: the equations describing the
vibration of an ideal string are straightforward to understand, computationally ecient to
simulate, and when used to make even the simplest models, produce sound qualities we
associate with stringed instruments.

thank you Kona. i will read up on it. i admit the roland guys pitching the platitudes of modeling is no worse than me trying to explain to my friends what the Federal Reserve is doing.

Yes I understand. My question was whether anyone can confirm that the V is based on pure physical modeling. The SN sounds use samples in the attack. The Korg rep seemed to believe that the V piano is based on the same underlying technology. A related point had come up on the Jupiter and it's VA section.

Hi Hideki,

First of all, never listen to what a "Korg rep" says about Roland's stuff as they are probably misinformed about the facts, anyway.

Here is Roland's word on that, copied directly from their product info:
(Note the words below, highlighted in red.)

--------------------------------------------------------------------------------

V-Piano: V-Piano®

The Piano Redefined

Breaking Barriers
The V-Piano soars above the limitations of past technologies with its revolutionary “living” piano core, allowing every note to respond and evolve naturally, seamlessly, and perfectly without requiring samples.
With the V-Piano, there is no velocity switching, and it provides a smooth, natural decay that must be heard to be believed.

--------------------------------------------------------------------------------

pv88, have you received your V-Piano yet?

James
x

And, I think to truly confirm what it is you are asking would require "dewster" (or, someone who can check the actual hardware of the V-Piano) to open it up, and, see if any samples are to be found. Not very likely, as such...

Here is a good article to read:

http://i.imgur.com/pfMdy

Or, just download the PDF, here:

http://www.hindawi.com/journals/asp/2004/981942/abs/

And, a simple layman's explanation:

http://www.applied-acoustics.com/techtalk/physicalmodeling/

Hi James,

Did reply to your question with a PM, instead.

Okay, thanks.

James
x

I can tell you that I am 100% certain the V-Piano does not use any samples, not even for the soundboard. Everything is emulated sound. Below is an excerpt of the pdf I posted above.



Physically-based modeling, or physical modeling, is a way to make sounds based on
the physics of mechanical systems. Compared to other kinds of synthesis such as FM or
sampling, it tends to be computationally expensive. Interesting vibrating systems, such as
musical instruments, are fairly complex; modeling the physics of these systems is much
more involved than modeling the sound spectra or waveforms they produce. Creating sonically interesting physical models that will run in real time has been a major challenge.
Despite this computational cost, however, physical modeling has been the most popular
synthesis approach in academic research since the early 1990s [62]. This popularity is
due largely to its promise to extend our acoustic world in perceptually novel yet intuitively
correct ways. Many researchers consider physical models to oer better prospects than
signal-oriented methods for the design of expressive digital instruments

The V-Piano uses four of the latest modified Fantom synthesis engines running four distinct/separate three dimensional physical models running simultaneously. What is interesting is that each of the four interact with each other and affect the way each other sounds. Each model uses a mesh type of design that can be manipulated by changing shape (hammer size) and physical property (silver and copper for instance) and soundboard size thus changing the produced waveform.

Just think how far this technology has come from just 10 years ago. The article I posted talks about "Intimacy" of the user and the instrument.

The Supernatural Roland piano uses recorded samples of three of the synthesis engines and using one modeled sound engine to create the illusion that the decay time does not loop. In essence, no real piano is used to gain the supernatural piano sound as well.

I own and have used the Yamaha VL plugin with my S90 for the past ten years which uses a breath controller which I blow into which is letting the physically modeled single or 1 polyphony sound to be generated. Archaic by todays standards, it does mark a big leap in the technology field and it is a shame Yamaha has abandoned it.

MIDI keyboards typically have acceptable
latency and jitter for intimate control, but the musical keyboard interface is not a good
match for the intimate connection invited by physical modeling synthesis. It’s evident that
Yamaha realized this. They included a breath controller—a little-used option for keyboard
players at the time—with the synthesizers as a strong suggestion that more intimate control
was desirable.
Keyboard instruments are certainly capable of expressive performance, but they do not
oer particularly intimate control. Tools aord certain kinds of use at the expense of others;
seen as musical tools, keyboard instruments trade intimacy for the control of harmonic
complexity. In order to oer this control, the piano mediates its soundmaking apparatus
through events. Each event, a keypress, triggers an excitation of a physical system, the
string, with an impact of a certain velocity. Compared to a guitar, say, the piano aords very
little control over a note once is it started and in general, we can say that signals provide
greater control intimacy than events. Given the ubiquity of the keyboard as synthesizer
controller, its use to control physical models is understandable. But in our search for greater
expressivity, it makes sense to look for more intimate kinds of control.
A haptic connection is another vital aspect of musical intimacy with acoustic instruments. Vibrations of the instrument sensed by touch are often an important part of the performer’s sensory experience.

In the copy and paste process, some letters go missing and the formatting gets all messed up. You can go read the pdf I linked to in an earlier post, this one is from page 24. This article was written in 2008 btw, so it is a little dated, however the actual material is not that far off as to where the V-piano has come. The article talks about 2D sound vs 3D sound where the V-piano actually ended up. THe intimacy factor plays a huge part as to why the V-piano works so well and feels so organic. That is probably what took Roland so long to get it right.

From "Medical Dictionary" at dictionary.com:


hap·tic definition
Pronunciation: /ˈhap-tik/ or hap·ti·cal Pronunciation: /-ti-kəl/ Function: adj1 : relating to or based on the sense of touch haptic mode of perception —Colin Gordon>2 : characterized by a predilection for the sense of touch haptic person>