with roland having gone physical modeling, I really wonder how much longer sampling technology will last.. seems yamaha isn't so into physical modeling, or maybe just holding off... I mean, the sampling thing is getting a little ridiculous don't you think? I mean, not the approach itself, but going 100 layers, etc...
I guess it's not like physmod will take over the world, sampled pianos are great for when you already know what sound you want, and it's already been sampled just the way you like it.
My opinions are obviously "going against the grain" with people that have invested money, time, and devotion to samples.
This isn't meant in a mean spirited way, it's just that I've concluded after listening to dozens of sample songs that the sample manufacturers have posted on their websites, that they don't seem (to me at least) to stand up too well. This in turn is based on the premise and assumption that they have only posted samples that make their products sound good. When I hear one note that consistently goes "boonk" my faith evaporates.
My conclusions are also based on the huge improvements in a couple of years that the physmod systems have already demonstrated. This isn't to say that they sound perfect or even great. But since the concept is relatively new, it likely will improve the most with time.
The sample manufacturers are going to battle hard and they already have; one of the most subtle and effective demos is of the BMDO, where they play their rendition (part of one of Chopin's Scherzos) first, then another sample rendition immediately after - that is obviously inferior. Strangely this is the only site I've found where the manufacturer posts competition samples too.
My gut feeling is that Roland has scored a winner, and the other makers of DP's will be forced to play catch-up if they haven't already done the research.
If this approach succeeds, then Roland will do well in sales, and it will be interesting to see what those of us with older keyboards and samples will do.
Samples will not suddenly die out - there is far too much invested in the system, and for many it is quite adequate. I just don't think they are the way of the future.
I have not heard the Roland, but from the other physical modeling pianos I've heard, they have many more notes that "...go boonk."
I downloaded it for a trial; it lasted on my HDD for about the sustain length time of a staccato C9.
At the price of $68.95 US one can't argue that the price isn't right, but after using Pianoteq, it elicited a few smiles and chuckles from me.
I suspect that in a few years, they will improve it to where it's usable.
That's one of the only audible videos I've found of the new VSL piano so far (most of them have lots of rumble).
Regarding the comparisons on the site for BDMO, I've found them to be some of the most helpful comparisons I've encountered and I like the new trend of posting the results of many competitors. The one thing they do differently, that I also like, is to include several recordings of real performances (as opposed to sampled pianos) to compare against as well.
It was interesting to hear people's preferences: not everyone I talked to preferred the BDMO's sound in those demos, though many did of course, and I think that the demos are very fair and representative (as I own or have used pianos from 3 or 4 of the libraries represented).
Is the new VSL piano available as a sample, or only with the hardware? I'm trying to figure out how a regular computer could manage 100 velocity levels.
Glenn NK, I seriously doubt that pianissimo's "physical model" is all that serious.. I should ask the developer directly, but I'm pretty sure it comes from a totally different school of thought than pianoteq or v-piano. Afterall, there isn't a standard for what to call "physical model" or not. Pianissimo is good for what it is, but comparing it to pianoteq is like pitting a lizard against a tank.