CP1

CP5

CP50

"Spectral Component Modeling" sounds like an interesting technical term...
Time for a new generation?


Peace

Thanks for the link, Richard....these look pretty slick.

Yamaha has developed an uncanny ability to nail the market right in the sweet spot every time.

Snazzy

Wow, interesting.... Comparison chart here: http://www.yamahasynth.com/jp/library/cp1_5_50_en/compare/comparison_eng.html
Video demonstration describing this "Spectral Component Modeling"
http://www.keyboardmag.com/article/yamaha-cp-1-stage/December-2009/105233

I wonder when they'll be in stores... the CP1 and CP5 look kinda nice.

Seems the CP5 is the strongest one.

I'll take a wild guess.

I think they are finally employing MP3-like spectral (FFT or DCT) compression to significantly reduce (~1/10) the size of the stored samples. I haven't researched it much, but it is probably the case that the compressed domain is linear, so they could scale / mix samples there and then run them through a single decompression stage.

If that is the case, then this is really good news. I'll take modern lossy compression (at an adequate bit rate) over stone age compression like note stretching and looping any day of the week.

However, if I were building a DP I don't think I'd go this route (particularly with flash memory so cheap). But I've thought about it a bit.

So what's better:

NW-STAGE keyboard (Wooden synthetic ivory weighted keyboard)

-OR-

GH keyboard

?

I don't think I would put a vacuum fluorescent display in a stage-type instrument. They are not as rugged as LED or LCD, require high voltages to operate, and they dim over time.

I don't understand what the numbers represent in the "SCM Piano Block" section of the comparison chart - is this memory? Whatever they are, the CP1 has more than the others.

Wow, these look very sexy!

I have my eye on the CP1 - that 55x2 character display is gorgeous!

Also, I note that all three instruments include the S6B acoustic piano sound, which I am assuming is similar to the highly regarded Natural S6 voice previously only available in the S90XS/S70XS.

And just a handful of vintage electric pianos too, rather than hundreds of poor quality sounds that no musician would ever consider using - good decision Yamaha!

Is 'NW STAGE' a new action? I haven't heard of this one before?



It's not immediately clear which instrument offers the highest specification, some are stronger in different areas rather than being arranged in a hierarchy - again, another interesting concept.

I'm really looking forward to trying these new models!

Cheers,
James
x

"Yamaha even developed a special NW-STAGE wooden key action specifically to meet the demands of top pros." I guess this is a new action?

The piano (and other) samples over there sound pretty good, though I would like to hear individual notes decay over a long period of time (I really wish that was part of the Purgatory Creek MIDI demo). Demos are usually so frenetic that any compression issues are effectively hidden (the whole point of a blizzard of notes, I guess). To their credit, they let the sustain at the end go on for a while before damping the strings.

Anyone got any info on when these will be released into the wild?

I'll guess it's a little different. I think they have broken the samples down into "components" like hammer strikes and the "bright" sounds near the ends of the string and the mellow sounds near the center of the strings. then on playback they can add weighted sum with more of the center string sound for a low key velocity and more of the end string sound for the high velocity string. This is almost the way physical modling works but I think Yamaha is doing it with sampling. The proportion of end to mid string sounds would change over time as the note sustained. Yamaha talks about "12 components" so rather then just end and center strings sound they might have 10 other sounds and then they get blended with time varying weights.

I am pretty sure this is what Roland does in the RD700 and FP7 too. Roland calls is "performance modelling".

None of these new CP series pianos have speakers. So the dynamic range and the level of bass power is entirely up to the customer and his budget.

Last night I was at a live music performance, 2,000 seat auditorium. The dynamic range is amazing when you put 25 instruments on stage. I noticed the percussionist, when he did the cymbal hits his hair would literally blow backwards. I wonder what speakers it would take before the "crash cymbal" sample in my Roland synth would make my hair blow backwards when I press the key? Reminded me of that old Maxel ad.

CP-1 next January and the other two later that year is what I've read.

"Spectral Component Modeling" - I'm sure someone in the department of obfuscation got a raise for that one.

The hell of it is, if Yamaha would just come out and explain, in plain english, how the hell they are doing the pianos in these things I would either immediately sit up and take notice or strike them from the list. I think they could do it without giving anything proprietary away (DPs are notoriously behind the technology curve anyway) and it would make it so much easier on us tech savvy consumers.

I'm starting to think they just hate us.

If the consumer was that tech savvy, they wouldn't need it explained to them.

Snazzy

dewster, the Yamaha America website seems to make a little more sense than the YamahaSynth.com site (I get the impression that the later is still in development).

http://www.yamaha.com/yamahavgn/CDA/List/ModelSeriesList.html?CTID=205800&CNTYP=PRODUCT

Take a look at the brochure too, it's very sleek, and includes an explanation of SCM. Clearly, Yamaha have adopted physical modelling for these new instruments.

According to the Japanese press release, the CP1 will retail at 525,000 JPY which is approaching the V-Piano price-bracket.

Cheers,
James
x

Hey, thanks for the pointer K-James! SCM is still largely a mystery to me though, I need more info to sooth my savage techie breast.

A lower price would sooth it too.

It's just marketing speak. It doesn't HAVE to mean anything.

They might just as well have called it "sonic projection synthesis". Or "modulated audio stratification". Or "bibbidy bobbidy boo".

It just has to sound good, to attract attention. Nothing more.

Correct but they do talk a little about it on the web site and give enough hints that we can make slightly informed guesses.

For example when they talk about components of the sound. This implies that what is stored is not a simple recording like a CD but that somehow they have broken the recording down some how into well, "components". What these are we don't really know.

They print in the FAQ

This really says they are doing some form of modeling that is certainly not done in their current line of DPs We don't know the details but I'm sticking with my guess that Yamaha is using a hybrid of sampling and models and not a pure physics model.

What if Toyota told you the car you were considering purchasing had an engine that works via the new mysterious "bibbidy bobbidy boo" processes? You'd say "F that, I'm getting the Honda that uses internal combustion!"

It shows Yamaha don't respect their knowledgeable customers, which I have experienced directly.

If I had some kind of idea of what it was that they were going out of their way to not explain to me, that would be one thing, but I don't even know that.

What are we supposed to do, read their frikkin' minds?

Honestly, corporate engineers need to take back a tiny modicum of control over how the products they so tirelessly labor over are represented by the suits.

These details MEAN something to some of us, and could easily stimulate sales, if they weren't endlessly obfuscated.

You may be 100% right. But this could also mean that they are only adjusting the cutoff and Q of a filter. Market speak can mean anything, at this point no one expects it to be truthful.

But how does Yamaha's crazy talk help me, the curious consumer, who wants to buy something based on solid engineering?

That's just what they want us to do. Make guesses. When we infer things that they did not state, we render them harmless from any claims of deception. That's standard procedure for marketing.

Sounds like the same display Kurzweil uses. You're right, they do dim over time (though mine is 10 years old and has yet to noticeably dim), but they LOOK cool!

haha

We have an old Ensoniq ESQ-1, in the studio with vacuum fluorescent display...it dates back to around 1986 and the display still works as new.

Not bad for a 23 year old instrument.

I agree with the above poster...I think Yamaha's displays look cool, and will be highly visible.

Snazzy

Having watched the Keyboard Mag video and read some of the blurb, I'm thinking that what's going on in these new pianos seems to be a more advanced form of what happens in my GEM Prp800. There, the piano samples are modified by the modeling algorithms so as to create a complex interaction that simulates the resonances and harmonics present in a real piano. The new Yamahas appear to be using multiple velocity samples and re-blending them using the models, whereas the GEM uses just a single velocity sample (and so has absolutely no audible velocity switching). In both the GEM and Yamaha, the electric pianos are completely modeled without the use of samples. If my theory is correct, the text below, copied from the GEM manual MAY shed some light on what's going on. Of course, I'm no techie and therefore could have this completely wrong - in which case, my apologies - but at least it explains the principles of modeling in layman's terms.

FROM GEM's MANUAL:

Physical Modeling is a method of sound synthesis based upon a mathematical model which describes the physical construction of the instrument being simulated. Unlike sampling technology, in which an existing sound is simply recorded at a particular moment in time, (to remain essentially the same for ever), the sounds produced by a Physical Model continuously react and respond to the player’s input, maintaining all the little nuances and imperfections which provide the most reliable subconscious assurance that the instrument being played is the real thing. A sampled sound can be likened to a “snapshot” of a particular moment in time during which an instrument is being played. It’s like taking a photograph. The goal of Physical Modeling is this: Instead of simply recording the final audible product of an instrument like an electric piano, we replicate all the elements which are incorporated into its construction. If the physical model is constructed carefully and accurately with meticulous attention to detail, the resulting sound should be exactly like the real thing and, more importantly, the experience of playing and interacting with the instrument should be similarly convincing. In the Rp 800, all the acoustic piano sounds are created by combining sample playback with Generalmusic’s Natural String Resonance, Damper Physical Model, Advanced Release and FADE technologies, (described below). Other instruments such as RHODES, WURLI and CLAVINET are created using pure Physical Modeling.

Natural String Resonance
This physical modeling technology, patented by Generalmusic as Natural String Resonance, allows all of the complex harmonics normally produced by a piano’s soundboard to be faithfully re pro duced. This means that a note’s individual sound will always be slightly different depending upon which other notes are currently being held, (and consequently which strings are un-damped and free to resonate in sympathy with the note played). If you hold down a low C and let the note decay, the strings for that note are still un-damped for as long as the key remains depressed. If you now strike another C higher up the keyboard, (staccato), you will hear the sympathetic resonance of the low C strings in response to the new note played. This natural effect replicates exactly what happens inside a grand piano. If you experiment with different combinations of notes you will hear harmonic colors particular to each. Because this effect is produced by physical modeling and not by samples or DSP effects, the result is a musically and technically accurate simulation of a piano’s soundboard and virtually infinite combinations of harmonics can be produced.

Damper Physical Model
Another technology patented by Generalmusic is Damper Physical Model. Whenever the damper pedal is depressed, the damper physical model simulates the effect of sympathetic resonance being produced by the strings which the action of the pedal has now left free to resonate. Use the damper pedal to hear the effect of the Damper Physical Model by comparing the sounds of notes played in the highest octave of the instrument with and with out the damper pedal depressed.

Advanced Release Technology
The particular sound of a piano string being stopped by a damper while in motion is replicated by Generalmusic’s unique Advanced Release Technology. Sample based electronic pianos traditionally use an envelope generator to control what happens when a key is released. This simply allows the sample loop to continue playing for a set period of time until its amplitude is finally reduced to zero by the envelope generator. In an acoustic piano, vibrating strings are silenced when a felt damper comes into contact with the moving string. When this happens, depending on how hard the key was struck and the length of the string itself, certain frequencies are damped earlier than others while some other frequencies are even accentuated, (anyone who ever studied how to produce harmonics on a guitar will recognize this principle). This produces a distinctive harmonic “ring” as the different frequencies in the string’s tone dissipa te through out the piano soundboard. This Advanced Release Technology in the Rp700 series simulates these phenomena with complete accuracy throughout the 88 note range.

FADE - Filter Algorithm Dynamic Emulation
Reproduction of the complex harmonic and dynamic changes which take place as you increase or decrease the velocity of a key-strike on a piano have always presented a serious problem for traditional sample-playback technology. The only practical way to replicate these`changes has been to select three or four distinctly different levels and switch between these according to the velocity with which the key is struck. This produces the unnatural effect of having clearly audible steps between different velocity levels, further diminishing the authenticity of the sound reproduction. Unlike the velocity-switching methods used in other electronic pianos, Generalmusic’s unique FADE technology utilizes only one specially configured sound source per note. At the heart of the FADE engine is an extensive database which can be used to look up the precise harmonic content of any note played at any velocity level. Whenever a note is played, the FADE engine analyzes the velocity of the key-strike and constructs, in real-time, a model of the necessary harmonic content for that particular note played at that velocity. The note’s sound source is processed by the FADE engine with appropriate harmonic content being added or subtracted accordingly. In practice, FADE technology provides seamless transition from pianissimo all the way through to fortissimo for each note without any audible switching.

video demo:

http://keyboardmag.com/article/yamaha-cp-1-stage/December-2009/105233

The CP1 is listed at $5999.00 USD...street would be around $5000 USD.

The CP5 list is $3299.00 USD, so street should be around $2600 USD.

The CP50 is listed at $2,199.00 USD so street should be around $1700 USD.


Snazzy

C1...Estimated Arrival in US 12/30/09

C5 ETA will be 3/15/10

C50 ETA will be 3/15/10

That is not exactly the way I heard it on the video, You have to listen to all three videos in sequence.

What I heard was there are four different classes of piano sounds, (1) acoustic piano, (2) Synthesizers, (3) Rhodes type and (4) reed type. All of these in their original incarnation are "real" analog instruments.

My interpretation is that 1, 3 and 4 are sample based and the synth sounds are created synthetically. Each of the four clases use a different sound generation system

I think he specifically said the synth souds were not samples, but he makes a distinction between synths and E. pianos.

When the yamaha guy talks about the electric piano he says the sound can come from hits on different locations on the string or tine.

Then he goes non to talk about digital modeling of the amp and speaker and various stomp boxes. I assume this is the same "electronic modeling" we see in various DSP gutiar amps. It's is not the same as modelling the sound of the piano. Those old e. pianos where a lot like electric guitars in that a lot of their tone was a result of the tube amps and speakers.

BTW. Most guitarists have the same opinion of "modeled amps" as pianists have of DPs. They are OK (maybe) but not at all like the real thing.

The CP1 is a vey complex instrument. It looks like it can do a lot more than a Roland v-piano and sells for $1k less. Bt I doubt many acoustic piano plays will rush to it because 3/4 of the the CP1 does is Rhodes and synths and simulation of old tube based amps and built-in EQ.

I'd say the C5 will be the most versatile and probably the most popular.

It has a mic input for the solo performer, as well as the other sounds one may use performing in a band. It has FX from the Motif XS workstations and rhythm patterns for a wide range of musical applications(arranger?). Also the NW-STAGE wooden weighted keyboard. Pitch wheel.Record and playback functions for MIDI and audio.


I couldn't find the weight.

It is around the same price as Roland's RD-700GX.




The C1 will compete with Roland's V-Piano.

At 60 lbs, it's a bit lighter, and probably more portable than the Roland. It weighs about the same as my CP-300.

It also has master-keyboard functions which have been specially crafted for use on-stage, up to four virtual zones can be setup on the keyboard and assigned to four different tone generators, including other MIDI instruments. It also has a pitch wheel something Roland neglected to put on the V-Piano.


The C50 is hardly entry level; it skips the wood keyboard, and is probably the lightest but still retains the SCM (Spectral Component Modeling) technology of the other two. It also has a pitch wheel. Also has rhythm patterns.Record and playback functions for MIDI and audio.


This is going to get very interesting.

Snazzy

The C5 is 25.2kg (roughly 55lbs)



You may well be correct, but it's not completely clear from either the literature or the video. Either way, the Rhodes and Wurlitzer sounds are pretty good to my ears.

Thanks Voxpops...I could live with 55 lbs, I believe, especially if it sounds better than my CP-300.

Won't be out till March next year, but I imagine playing the C1 will give a good approximation of what it will sound like.

These are a complete surprise; especially for Roland, I imagine.

I'm interested in what "rhythm patterns" means...maybe a basic arranger function, or preset drum/arpeggio patterns as on the Motif.

I think the keyboards are all wood this time (including the black notes) as the action is supposed to be all new. No doubt these will end up in the Clavinovas.

Snazzy

CP-1: 27.2 kg (60 lbs)
CP-5: 25.2 kg (56 lbs)
CP-50: 20.9 kg (46 lbs)

from http://www.yamahasynth.com/jp/library/cp1_5_50_en/compare/comparison_eng.html