Messe: Viscount Physis Piano

[Ashville.Guru]

Heh, apparently they've never heard of Roland....

Strongly disagree. They're rather careful about what they're saying in the video. I'm pretty sure they don't claim anywhere in the video, that this is the the first DP in human history to use physical modeling.

 

Sure, at 0:30 the guy says says "...to illustrate how what physical modeling is, and how this piano, and this technology is different from all the other DPs in the market, 'cause some of you may be familiar with physical modeling, some of you not...".

 

But he's not really making a serious claim here, he's just excited about his cookie demo...

 

To settle any doubts, note what he says at 7:02: "making this the only product on the market with multiple physical models inside.".

 

You think they'd have taken such pains to qualify and water down the claim, if they hadn't heard of Roland? They're certainly not obliged to mention the history of the market, while giving a product demo.

[Ashville.Guru]

At least it offers more than the V-Piano with EPs, Clav and mallets. And it'll likely be cheaper.

 

Don't forget the sampled sounds! The EPs, Clav and Mallets are just the other physical model engines. There are strings and other sampled sounds that can be "mixed" (layered?) with the pianos.

 

But seriously, how cool is that panel thing ? I especially like how you can make the entire display disappear. Easier to forget the 'digital' and connect with the 'piano'. Design, customize your own interface, to taste. I haven't seen a feature like this in any other keyboard, DP or not. Deserves props for this alone.

[Tobias Åslund]

No matter who's first or if their claims are true or not - this still seems like a very nice keyboard! The weight, IMHO, managable, even with the wooden keyboard.

It would be nice to hear the quality of the other sounds and the most important thing - how much the thing cost and when it'll be available.

My guess our Swedish distributor will do some serious promoting about this one. I can't wait to play one!

[Josh Paxton]

This company certainly seems like TLAs (three-letter acronyms).

 

After the demo, they have my attention, though I sure wish we got to hear more non-phaser drenched Rhodes.

[AnotherScott]

I agree with Ashville, their reference to being the only DP on the market with multiple modeling engines seems to be an implicit acknowledgment of Roland.

 

It's hard to tell for sure from what little we have, but from what I could tell, I'd probably prefer its sound to the V. Other than that, yes, it's got all those other models, plus a bank of rompler sounds, and we'll have to see about price. The weight is certainly a lot more portable than the V.

[Al Coda]

Looks impressive (and expensive)

 

Two models, plastic keys (4l lbs) or wood keys (60 lbs)

 

This is the best I´ve seen and listened to from Frankfurt Messe up to now.

 

I got the info it has 6 SHARC chips inside which is lots of DSP power,- comparable to JB Solaris.

I like this DSP technology because there will be connectivity.

As w/ my XITE-1, I expect a latency of a few samples,- maybe 10 to 15.

I also like the look and from watching the vid, I have the impression they have done a great UI.

 

Very intersting product for me.

 

A.C.

[Dana.]

To settle any doubts, note what he says at 7:02: "making this the only product on the market with multiple physical models inside.".

How many does Kronos have?

[Ashville.Guru]

To settle any doubts, note what he says at 7:02: "making this the only product on the market with multiple physical models inside.".

How many does Kronos have?

Correct me if I'm mistaken, isn't the Kronos piano sample-based (with a bit of modeling thrown in)? If yes, the answer is zero (at least as far as piano-based instruments go), and his statement stands.

 

Such hybrid technology(sample+modeling, also seen in some software pianos) is miles apart from that of a completely physically modeled piano. Note that I'm not saying one is better than the other, just that they're different animals.

[AnotherScott]

Correct me if I'm mistaken, isn't the Kronos piano sample-based (with a bit of modeling thrown in)?

The Kronos piano is sample-based, but its Rhodes and Wurly are modeled. They actually call their synths (i.e. Polysix and MS-20) modeled, but I don't think that qualifies as "physical" modeling. The CX3? Maybe...

 

But in context, he has arguably qualified his assertion by saying, in effect, that it is the only product in its product category (DPs) with multiple modeling engines, so he's probably still on safe ground!

[Dana.]

The only time I see the phrase "physical modeling" on the Kronos page on Korg's website is in reference to the STR-1 Plucked String Synthesizer. So, that's one, maybe two with the EP-1.

 

Anyway, I hope they find a distributor in the U.S. that's worth a damn so that I can play one.

[JeffLearman]

There really is no such thing as true physical modeling. It's all just mathematical models of abstractions of physical properties.

 

CX3 should count as modeled. I didn't realize the Kronos Rhodes and Wurly are modeled. I still haven't heard a modeled Rhodes that's half as good as a good sample set. Guido S's is still the best that I've heard. Lounge Lizard is pathetic as an authentic Rhodes (though musical and useful as its own thing, just as a Rhodes is a pathetic copy of a piano but way cool). If Pianotech's rhodes is authentic, it needs lots of tweaking from the demo "Play" version. I haven't had a chance to play a Kronos.

 

I think they're guilty of marketeering, which makes claims that are stretched to the greatest possible arguable position, rather than anything near the objective truth. Which is par for the course.

[AnotherScott]

The only time I see the phrase "physical modeling" on the Kronos page on Korg's website is in reference to the STR-1 Plucked String Synthesizer. So, that's one, maybe two with the EP-1.

Oh yes, good catch, the STR-1 is physical modeling.

 

And you're also right to question the EP-1. It is not strict modeling, it is something of a hybrid. Whether you count it as zero, one, two, or six is open to debate. EP-1 has 6 models... but 4 are Rhodes, 2 are Wurly. (The Physis is counting the Rhodes and Wurly as separate models, which makes sense, their physics are different.)

[AnotherScott]

CX3 should count as modeled.

It's definitely modeled, I'm just not sure it qualifies as "physical" modeling, as there is no mechanism in the original that generates sound unless you apply current to it. So I'm not sure whether I'd put the tonewheel modeling in the category of "physical" like a hammer and string/reed/tine, or if it's more like modeling a Polysix, where in the original, nothing physical is happening at all, it's all electrical manipulation of circuitry. The spinning tonewheels are kind of in between, to my mind.

 

I didn't realize the Kronos Rhodes and Wurly are modeled.

I should clarify, as I begun to in the previous post. It is not straight sample playback (or even modified sample playback with filters and layer crossfades) as you would normally characterize most sampled pianos, But the sound does actually have samples at its core. As Rich F. explained in the korgforum:

 

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

EP-1 does in fact start with a sample library, but it goes deeper than traditional multisample playback. MDS (Multi-Dimensional Synthesis) lets us separate a sample into different elements- in this case, the pitched content, the hammer sound, and key release noises. We assemble these elements together in a way that gives us precise control over how the sound moves through time.

 

The practical upsides of MDS are:

1. There's no velocity-switching, so it reacts dynamically throughout the entire velocity range, just like the real instrument naturally would.

2. You have control over physical properties of the instrument, like hammer width and attack brightness.

3. We can control time-based elements like decay and release without having to introduce more processing.

 

So... Is it modeling? I suppose this depends on your definition of the word, but since it goes so much deeper into the behavior of the instrument than a traditional sample-playback engine would, the word "modeling" should be appropriate.

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

 

So... there is a semantic issue here. And I can't say I fully understand it, either. With no velocity switching, it sounds like he's saying there is only one pitched-content sample available to generate a given note, and yet the character of the note does change very substantially with velocity, including very authentic "bark" which I can't see how you could possibly generate out of the same sample as the mellow tone if you were actually simply manipulating samples on playback. So something pretty sophisticated is going on. Kronos EPs are as good as or better than any other EP sound I've heard in a hardware emulation (I haven't played with software EPs). I'm not a big fan of their 4.7 gb acoustic pianos samples that so many people admire, but to my ear, they sure got EP right.

 

 

[Dana.]

I still haven't heard a modeled Rhodes that's half as good as a good sample set.

The Yamaha CP1/5/50 has the best modeled Rhodes sounds I've heard -- better than GSi, Lounge Lizard, Pianoteq, and better than Kronos.

[AnotherScott]

The Yamaha CP1/5/50 has the best modeled Rhodes sounds I've heard -- better than GSi, Lounge Lizard, Pianoteq, and better than Kronos.

Thanks for reminding me, I haven't heard the CP1/5/50 EPs yet either. (It wasn't a priority since I knew I'd never gig with anything that heavy anyway...)

[Dave Ferris]

The CP5 and Nord Piano's Rhodes are both very good.

 

On the CP5 I tend to play the Simple RII or Driven RII. For most situations in the past I find them to be almost too soft, their detail and overall coolness factor gets lost in the signal/sound chain. However with these API pre amps I've been borrowing, the sound is thick and really in your face. I'm going out of the balanced xlr outs. It totally changes the dynamic of the sound for me.

 

On the Nord, the Sparkletop really comes to life with the API, quite amazing.

 

Also the Nord's XL Yamaha BR. G & Bosie are nothing short of stunning.

[humannoyed]

For those of us who are tweakers PM holds great promise. How many of us have been bothered by some anomoly in a piano sample that we had no control over to fix.

 

Still sample technology has been around a while and is far enough advanced to still hold the lead in authenticity. I think therefore we will see more hybrids in the future, still mostly sample based but perhaps the balance leaning more and more toward PM as time passes.

 

Those pursuing PM exclusively are the trail blazers. I am appreciative of their efforts and attention to detail, but they have their work cut out for them with our high level of expectation.

 

 

 

To settle any doubts, note what he says at 7:02: "making this the only product on the market with multiple physical models inside.".

How many does Kronos have?

Correct me if I'm mistaken, isn't the Kronos piano sample-based (with a bit of modeling thrown in)? If yes, the answer is zero (at least as far as piano-based instruments go), and his statement stands.

 

Such hybrid technology(sample+modeling, also seen in some software pianos) is miles apart from that of a completely physically modeled piano. Note that I'm not saying one is better than the other, just that they're different animals.

 

[Dana.]

Dave brings up a good point about the details getting lost in a mix. I'm not sure there's an advantage of one over the other (sampled vs. modeled) in terms of cutting through. I was at a show last week and the keyboardist was playing a Nord Electro 3 (I think it was the Suitcase Rhodes sound). In the mix (two guitars, bass, drums), it had no presence. As listeners out in the audience, what boards/samples have we heard that cut through?

[Charleston]

Around 3:04:

 

Inside Physis there are 6 internal processors running in parallel

 

...24 billion computations per second

 

..all these physical characteristics of a piano, we use 100 of these ingredients, for every note. And the internal processors are now manipulating those parameters in real time as I play. So the reason why this piano, this technology, didnt exist before today, is because the computer processing power hadnt reached this level. So thats why were able to produce Physis today.

 

Hmmmmm....aside from Physis, what hardware piano (or modeler) has the next highest level of parallel processors (or processing)?

 

Also, I haven't kept up with the "best of the best" laptop/towers, so do they compete with Physis in the processing area?

[Ashville.Guru]

Having physical modeling in a DP is awesome in its own way. But that's arguably still in subjective, speculative territory.

 

OTOH, there's a lot of reason to be excited by the full-length, touch-panel interface. It opens up a world of future possibilities in the world of DPs, especially with controller features. Consider this: future OS updates could add stuff like

• A dozen sliders.
  
• Virtual pitch/mod wheels.
  
• A full-length ribbon.
  
• Slidable split points.
  

The only limitation is the imagination and motivation of Viscount's software developers. Which is why it would be mind-blowing if Viscount makes the interface open-architecture; tone-generation would still be under their control, they just need to release a restrictive API for plugin development. A whole lot of 3rd-party developers could start creating cool controller plugins. At very little extra R&D cost to Viscount, their product value would hit the roof...

 

... and I'm hoping that even if Viscount drops the ball on this (knocks on wood), other manufacturers run with the idea! Hardware sliders/knobs/buttons could very well become a novelty thing. Here's hoping that the Physis is the forerunner of that revolution!

[nickd]

It's Analog Devices SHARCs in the Physis, according to an earlier post.

 

Presumably "24 billion computations per second" means 24GFLOPS - which would be 4GFLOPS per device. But I had a look at the Analog Devices website and it seems the top performing SHARC "only" does 2.4GFLOPS. Although I confess I didn't read the datasheet for all of them - I picked what looked the top-end one and had a look through its specs.

 

I wonder which processors they are actually using?

[Moonglow]

I thought GEM was one of the first companies (Maybe even the first?) to do a really good job with physical modeling for piano.

 

The Physis Piano sounds good, but I really can't appreciate it as being anything substantially different/better than other high-end DPs heard over computer monitors (duh). Look forward to trying it out. As others have said, I wonder if the price point will be much lower than the V-Piano or CP-1.

 

The faceplate to the stand in that video reminds me of the Farfisa VIP series!

[JeffLearman]

I like the sound of the piano at the end.

 

Regarding the user-configurable interface ... I don't quite get it. The display is only 480x272. That's great for a keyboard's display panel in the center. But clearly all those other "buttons" have to be permanent, not configurable. Am I missing something here?

[mate stubb]

keyboardist was playing a Nord Electro 3 (I think it was the Suitcase Rhodes sound). In the mix (two guitars, bass, drums), it had no presence.

 

A real rhodes would get lost in that mix too. For doing battle with 2 guitars, you usually must reach for the Wurli.

[mate stubb]

I like the sound of the piano at the end.

 

Regarding the user-configurable interface ... I don't quite get it. The display is only 480x272. That's great for a keyboard's display panel in the center. But clearly all those other "buttons" have to be permanent, not configurable. Am I missing something here?

 

After watching their vid, the configurable interface seems to consist of turning lit button groups on or off. Not "I can put any control anywhere I want."

[Dana.]

A real rhodes would get lost in that mix too. For doing battle with 2 guitars, you usually must reach for the Wurli.

Because it's got more bite to it? Wouldn't a Rhodes with a pronounced bell sound (i.e., Nord's Bright Tines or one of the Dyno sounds in the Motif) fare equally as well?

[JeffLearman]

As listeners out in the audience, what boards/samples have we heard that cut through?

I find the NE2 Rhodes is much more easily buried in the mix than either my Stage 73 (played through PA monitor rig or through Fender Blues Deluxe) or my Rhodes soundfont (played through my PA monitors).

 

I like the NE2 Rhodes; don't get me wrong. But it's hard to control the velocity on the semi-weighted keyboard, and it just doesn't have the punch or image that I miss from my older setups.

 

You may not like how my Rhodes soundfont sounds -- it's very bellish, EQ'd up the wazoo, and the low notes kinda fart in the top velocity layer, and there are at most 5 velocity layers. It has its drawbacks. But man does it jump out in a mix, especially when played in stereo. Since the subtle imaging effect was added using mid-side, it sums to mono perfectly. However, it's not a well-rounded do-everything Rhodes tone. It does what I crafted it for, and that's all. Still, I miss it!

 

BTW, my experience includes scores of blues jams with two guitarists playing too loud and someone else playing my rig. I know exactly how it sounds from the audience. Of course, some players make it sound better than others!

[Ashville.Guru]

Regarding the user-configurable interface ... I don't quite get it. The display is only 480x272. That's great for a keyboard's display panel in the center. But clearly all those other "buttons" have to be permanent, not configurable. Am I missing something here?

 

After watching their vid, the configurable interface seems to consist of turning lit button groups on or off. Not "I can put any control anywhere I want."

Note that I was not referring to the current implementation:

It opens up a world of future possibilities in the world of DPs, especially with controller features. Consider this: future OS updates could add stuff like...

Regarding 480X272, I'm sure it will be an issue if someone wants to put up pics of their girlfriend underneath the sliders. Otherwise, not so much.

[Ashville.Guru]

I'm not sure there's an advantage of one over the other (sampled vs. modeled) in terms of cutting through.

 

There's always a potential advantage for modeling in terms of cutting through. As there is in terms of... everything else. But then, serious emphasis on potential.

 

Consider Mr. Viscount Demo-Guy's cookie example. Once it's baked, there's very little you can do to alter the taste. If your Nord samples aren't cutting through, there's very little you can do to make them (EQ can only take you so far).

 

As humanoyed points out, modeling is for the tweakers among us, those who like to roll their own. There are a gazillion bazillion uncountabillion different sounds that you can generate with 100 different parameters. Somewhere in that astronomical number is the ideal piano sound for you: it floats your boat, and cuts through the mix very nicely.

 

But of course, finding it will take time and effort. Just like a synth, you'd need to play with the parameters, and train yourself, so that you understand how changing each parameter affects the sound. Once you do that, you're likely to find that sound that cuts through best. The same applies for any other aspect of the tone.

 

The whole point is that you simply can't do this with samples. You're stuck with the sound that is provided.

 

If you're not a tweaker, and expect things to work out-of-the-box, the modeled piano is perhaps simply not for you.

[JeffLearman]

A lot of Rhodes pianos didn't sound that great right out of the box, either. ;-)