Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards Roland DP's: how many samples per key ?

For the Roland HP range: i frequently see this text in the specs:

" ... individual multi-samples per key ..... ".

But they never say how many samples per note/key (Roland samples every note unlike Yamaha), i didn't found info on the web neither.

The new Yamaha CLP300 range has:

>>> 30 sampling banks: (rest of the notes interpolated)

3 samples per note: MF - F - FF
(CLP320 and 330)
4 samples per note: MP - MF - F - FF
(CLP340)

>>> 50 sampling banks: (rest of the notes interpolated)

5 samples per note: P - MP - MF - F - FF
(CLP380)


So who knows how Roland manage this?

Or Kawai?

Well it seems that Kawai uses another technique, called "Harmonic Imaging".

http://www.kawaius.com/main_links/digital/harmonic_imaging.htm

Like samplebased software piano's as "SYNTHOGY IVORY" and the new "GARRITAN STEINWAY" you have the 'Virtual Sound Technology' like "PIANOTEQ" uses, which is in principle very good and gives more nuances in the sound, so a pro player can put more nuances in his playing, but the sound itself is still not so good (yet) and natural as a sample based piano program.

A famous pianist like HUGH SUNG says that he uses PIANOTEQ for putting more nuance's in his play:

http://www.hughsung.com/blog/index.php?query=muse+receptor&amount=0&blogid=1

I think this sound issue is the reason why some people don't like the KAWAI sound compared to YAMAHA, ROLAND or even CASIO.

There are also some models of the GEM digital piano that use such a 'Physical Modeling Technology' called DRAKE.


Found it all on the net, but what about ROLAND ?

This article may help better undstand roland.

The article goes fairly deep into how Roland makes their sounds. While the information is a bit out-dated, the latest stuff they talk about is still in production, or is only just now starting to be replaced, so the article is still quite valid and illuminating.

The article also has a brilliant section on keyboard and keyboard weights, which illustrate why it is I find Roland's touch to be so real, while other digital pianos can feel so fake to me.. it's not all just about key weights, just as it's not all about samples..

I don't know how many samples per key as regards dynamics. I've read, although I can't find the link now, that Roland has one hundred steps from 'no sound' to 'all sound'. I don't think that means 100 samples, but that there are 100 shades of sound between no sound and all sound.

To me, it's seamless, and with the tonal variations these pianos have, you wind up with a very expressive instrument.

There's so much more than just samples that go into making a good fake piano.. samples is but a tiny part, really.

Numbers can be deceiving. Sometimes the car with the lower horsepower rating is fastest around the track.. numbers don't tell everything.

Do read that article. It'll explain in fairly deep detail how Roland does what they does, and that includes the keyboard itself.

OK. But why are there so few positive comments on Kawai's "harmonic imaging" if it is a superior approach?

Every manufacturer claims to use a "superior" approach. The proof is in the final product.

Thanks stringless for that great link!

Because the main piano sounds so ugly and uneven compared to other digital pianos, that even if they used a superior technology than other manufacturers, it wouldn't be noticed. And by the way, their physical model has some bugs (resonance not working like it should, bugged sostenuto pedal, etc).

I have a Kawai CA-51 and replaced its clunky sound with Pianoteq (by attaching the DP to my PC). Now THAT is superior in every way to everything I've heard so far. Best money ever spent.

Shirakawa,

If you believe there are bugs in the CA51's software, I would be grateful if you would please send me a private message detailing the problems.

Kind regards,
James
x

who cares?

Maybe the answer to your question is in your very own signature.

About the "100 steps", isn't it some marketing trick to hide the fact that they could not produce a keyboard that would enable the 128 (or 127?) variations that Midi allows?

I play from time to time a less than 3 years old Roland Digital (a nice expensive one with a large LCD and a lot of sounds). The action is not that great, but what annoys me the most is the feeling of some short latency between the moment I play some keys and the moment the sound is produced.

Vincent, I think some Rolands allow you to adjust the "latency" period. Have you tried this? Which model do you have?

I don't know the model (it is not my DP) but I will check this feature out next time I go to friends' place. Thanks ofr the information.

Wow, such a sutble bash.

Here's a not-so-subtle one:

At least with the Roland you get audible and palpable changes in tone quality as you play louder or softer. So much more tonal color that I wonder if the other digital pianomakers' only tone is "Monotone Boring Tinkly Grand Piano." Piano 2 on the other maker's offerings is probably "Dull Coma-Inducing Boring Budget Upright"

I even SAID I wasn't sure of the number, and that I couldn't remember where I got it. Thanks for ignoring that bit. Really, thanks. Only you turned my words and used them to illustrate some supposed inferiority on the Roland.

Now that we got the daily bash out of the way, perhaps the older roland you speak of with the delay had something spilled in it? I've never heard of a lag in an electronic keyboard instrument, be it piano, digital piano, organ or digital organ. No-play, sure (dead keyswitch).. but never a 'delay'. That's for mechanical pianos into whose actions have fallen countless gummy bears and pencils and gumdrops and sandwiches and whatnots. That'll do it for sure.

Maybe you just like to bash and offer anecdotal evidence of suckage where there is none. Does it make you feel bigger?

They got pills for that these days, you know. No need to get on a forum and bash products you don't like by twisting words around.

No need for personal attacks, as we do not know each other and probably can't have any influence on each other's life.
I did not pick on your post but on what Roland communicates. I am sorry you took it any other way.
Here is why:
It happens that the Roland reseller I visited in town had a very similar way to describe the difference between Roland DPs and the competition.
"Each key sampled at several levels ("100" is the exact number the sales person gave me) for the full duration of each note".
He compared his higher priced Roland DPs with a beaten up (the keyboard was broken, I could hear) Yamaha YDP-223. "Yamaha? Not all the notes sampled, loops to get the notes sustained, ... Junk!" and denied any multiple level sample (p, mf, f or more) for the Yamaha DPs.

The problem is that the piano sample size for a Yamaha P-120 is about 20 MB ... just for the piano voice #1.
So for the Roland, what is the size of the samples set? That will in tens of GBs, correct? Technically possible, but unlikely.
So it is my perception that your post is close to the Roland's marketing message, and it is that message that I dare to trust with reserves.
I am not in the piano business. As a end user, and not even a pro at that, I have no preferred manufacturer, and I would buy, if in need, any instrument that fits my budget and inspires me. I really respect Roland as a synth manufacturer. I considered a FP-3 and a FP-5 at one point but so far, after spending hours trying them, I am not crazy about their DPs.
I am not in love with any one manufacturer and that allows me to just try anything available with no preconceived or interested opinion.

I hope this clarified my previous post and will end this misunderstanding. If not ... you can like it or not, this is my just my opinion that I post here. And I will continue to do so because this forum is also about that too: John Doe sharing his point of view with other John Does.
Thanks to eJohn who gave me something to search for and try to improve my experience with my friends' Roland DP.

My apologies to the rest of the forum members for the noise.

One little unscientific test you can do with your digital to check dynamic range is to connect it to a recorder that has level meters or to a computer which is running sound-editing software that displays the record level.

I have a mini flash-recorder. If I connect it directly to my dp and push record, I can attempt to play an isolated note at many different levels of volume by varying my attack. If I simultaneously watch the record level meter(s), I can get an idea of the number of distinct volume levels that the dp can differentiate in my attack and communicate through its output. If I play back the recording, I can compare what my ears hear with the perception of the meter.

I realize this is not scientific., but it did tell me two things about my Roland FP7.

a) the dynamic range capabilities are different depending on what range is sampled (bass/mid/treble)

b) the overall dynamic range of a Roland FP7 is excellent for a digital piano.

I'm a Casio owner myself (Ivory interfaced but still), however if I were in the market for another digital, I'd probably go with Roland FP4 or FP7. I've only heard good things about them on this forum, and the individual sampling is a big selling point for me.

I finally found the answer to my initial question myself.
Must have overlooked it.

- 3 level sampling for EVERY note/key for the HP-201

- 4 level sampling for EVERY note/key for the HP-203,204 and 207

Found it on the European Roland site HERE

For the US things are a bit different:

- 4 level for the HP-203 and HP-207 (there is no HP-204)

- 3 level for the HP-201 i think: it's not completely clear.

US Roland site HERE

Only for the FP-4, FP-7 and RD 300 and 700 range i still found no info.

Vincent L., may I ask if you have a link to the source stating that the P-120 utilises a 20 mb sample source for sound #1? Are you measuring megabytes or megabits?

turandot, while this method may approximatley measure the dynamic range of a digital piano, it does not take into consideration changes in timbre when playing from soft to loud.

It is relatively simple for a digital piano to play a mezzoforte sample at half volume in order create a pianissimo tone, or to play the same sample at twice the volume in order to create a forte tone, however the characteristics of the recreated sound are not an accurate representation of an acoustic piano.

Kind regards,
James
x