Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Roland FP90 versus RD2000 - and polyphony limitation questio

I know these are different beasts. But why would the RD2000 have so many more sounds, access to a sound library and have a polyphony limit of 128, while the FP90 has almost a couple hundred more in polyphony capability. This makes no sense to me. Can someone explain this?

What Roland says:



The _modelled_ piano sounds are in the "V-Piano Technology" sound engine, with unlimited polyphony.

The _other_ (non-modelled) sounds are in a different, sample-based sound engine. And it does have a "conventional" polyphony limit (I assume "128 voices" means "128 simultaneous notes").

For the FP-90:



So the _modelled_ sounds are (once again) "unlimited polyphony". And the sample-based sounds have a high limit (higher than the RD2000).

My understanding:

. . . You are unlikely to run out of polyphony using sampled sounds,
. . . with real playing, on either DP.

Thanks Charles. If one has four or five sounds layered though (and maybe playing back through a computer someone may use more), doesn't that cut into the polyphony? It just seems like 128 was a low number for today's technology.

Yes, layering sounds requires increased polyphony. Whether 128 isn't enough -- that depends on whether you _do_ layer four or five sounds, and what you're playing.

If 128 is enough, the only thing Roland gains by putting in 384 is "bragging rights".

They probably have the same processing power... the fp90 can only do 2 sounds at the same time... the rd 2000 can do 8 sounds, which also have much much more dsp power behind them... dsp power requiers quite a lot of cpu... thats why the rd2000 probably has less polyphony

The easy assumption is that the "fully modeled SN pianos" are more or less equivalent to the "RD-2000 V-Piano engine." Both have unlimited/full-keyboard polyphony.

Everything else? Kind of a crapshoot.

The FP-90s non-piano tones could very well be generic fully-sampled PCM, and whatever processor is involved can do 384. The RD-800's hybrid SN engine, with sampled attack and modeled decay, does the same 128 as the RD-2000's non-V-Piano engine.

So I'm not sure that I see any discrepancy here? In short, the RD-800 sounds may be rendered at a higher fidelity than the FP-90's, which would further differentiate the products?

Thanks Gombessa -this is the type of response I was looking for as it did not make any sense why the FP90 has 384 which it will never need and the Rd2000 had a 128 limitation. But if they are apples and oranges in terms of processing usage and fidelity - then the 384 on the FP90 - while not needed for that keyboard - is just fluff.

Also the RD-2000 has 192kHz processing on the sampled side of things - this gives a greater dynamic range versus 44.1kHz. Means there is less compression to the sound vs a real piano and your play can have a greater range of dynamics.

Are you sure about that?

I know that...


...however does this necessarily mean that the instrument's onboard samples were captured at this frequency?

Kind regards,
James
x

You are right James - it is a bit unclear.. though I am hardpressed to think of a reason for the system to support these bit/sampling rates if none of its voices make use of it.

When running a DAW or external VST/effect on a PC/Mac that allows higher frequencies, I expect.

Kind regards,
James
x

the native resolution would not be of import with external vst - audio would come from the pc's or external module's engine wouldn't it?

Oh wait - the 8 layers of sounds/zones possible internal and external combined then outputting from the RD-2000 to the amplification - yes, that would be useful ^^

Eying the mentioning of max. 44.1kHz 16bit depth for USB recordings and the lack of expressly mentioned higher bitrate SN/pcm soundsets, I think it i safe to say you're right

The Main reason for the older SuperNATURAL with 128 notes is compatibility with the RD-800. The RD-800 is a successful workhorse.

Roland gave us a few new "Acoustic"- V-Pianos and Electric Pianos - RD-1000 + MKS-20 plus the CE-1 and Dimension D effects + Axial with everything else a rehash of the RD-800.

Most likley a smart move by Roland. Time will tell.

There is no relationship between the sample rate and the compression. The only things which can reduce the dynamic is the sample depth (16bits is roughly 96dB), and the way samples are recorded.

192kHz can be interesting if you want to avoid aliasing artefacts. With samples, you can use antialiased 44.1kHz samples. With modelisation it can be harder, but it depends of the formulae involved. (And Roland hasn't publiblished their formulaes).

http://www.sonicscoop.com/2013/08/2...-knew-about-bit-depth-is-probably-wrong/ for those interested ^^

I don't get your comment - how can polyphony have any influence on compatibility with the RD-800 ?

surely you could have exactly the same tones but with the ability to have more of them at once ...

Cuz some of the guts inside the RD-2000 is RD-800 guts that have 128 note polyphony!

At least for supernatural engine the polyphony doesn't mean notes but voices. I have roland FA08 that has 128 polyphony on the paper. The Supernatural piano runs out of polyphony very rapidly when you step on the gas as each note eats up 4 voices. That gives you in fact 32 note polyphony. The 384 FP90 voice polyphony equals to 96 notes polyphony, which is adequate for piano. You can't directly compare polyphony between manufacturers or even between different engines of the same manuf. number by number.
Yamaha 64 note polyphony in fact beats Roland 128 voice polyphony by being able to play twice as much notes.
The 384 voice polyphony is not bragging in case of supernatural, just misleading what the numbers means.

Then OSCAR1 - Roland should have increased the polyphony on the RD2000 if they wnted it to be competitive for the next several years. Was it that much more costly to say - got to 256?