Topic: Modeling - it's getting there, slowly but surely!

I remember the day the Roland RD-1000 arrived at the local music store in San Antonio circa 1986. My friend (a Kawai EP-308 owner) and I rushed down to hear what the Roland rep described as a serious challenge to Kurzweil's rig (which were going for about 12,000.00 USD at the time).

While certainly not a sampled grand in quality it did a great job with the Yamaha CP-70 and the Rhodes sound - and since our back's were begging for relief (Both the EP-308s and Yamaha CP-70s are HEAVY beasts) he bought one.

Soon I took my Yamaha cp-70 owning buddy to hear the RD-1000 Austin got in and BAM! Sale #2.

I ended up with that RD which is still my go-to controller (real wooden keys!) it's awesome and durable but I've been using other sound sources for decades.

When I heard the original Pianoteq 1 way back when it really reminded me a lot of the original Roland SAS engine. Very dynamic and responsive but not a sweet wooden piano sound.

What shocked me though is a couple years ago I dragged my old upright down the sidewalk from home to studio (40 feet) only to discover that there was still no comparison between my racks and VIs of Roland, Kurzweil, Yamaha Ivory, Mach V etc and the real piano. None. Not even in the same world.

In fairness - I used to work in a studio which had a small Kawai grand and it never sounded very good, so simply being real doesn't mean it sounds good. But realizing that even the latest greatest Ivory was being blown out by an old upright was a shocker. 30+ years of sampling technology. From megs to hundreds of GIGs and still an experience totally unlike sitting down with the real thing.

So, on the verge of a major real piano purchase I reinvestigated all options. Kudos to Pianoteq for sticking with the plan, SO much nicer than Pianotech 1 and some of the overtones are downright scary real. Still we are not "there" but ever inching closer. I'l tell you what - it was a thrill to sit at my original RD-1000 and experience the latest modeling technology from the very same keyboard we rushed out to hear in '86.

I'm curious - it still technically Structured Adaptive Synthesis? Anyway, I'm a convert and will be purchasing Pianotech 6 soon. To me all attempts everywhere have some definite flaws (I'm still planning to purchase a large upright) but now, just using the Pianoteq demo and blending that with some samples, WOW! Definitely the most satisfying non stringed piano experience ever.

Again, thanks to all the folks at Pianoteq for staying the modeling course...

Re: Modeling - it's getting there, slowly but surely!

Pianoteq doesn't use SAS at all, it never did (obviously, since that's old proprietary Roland tech). Proper math algorithms all the way.

Last edited by EvilDragon (14-03-2018 08:53)
Hard work and guts!

Re: Modeling - it's getting there, slowly but surely!

EvilDragon wrote:

Pianoteq doesn't use SAS at all, it never did (obviously, since that's old proprietary Roland tech). Proper math algorithms all the way.

Thanks for clarifying, I didn't remember that it was a proprietary technology and was thinking it was a technical term for sonic modeling - especially after I heard the similarity in the early Pianoteq sound.

Re: Modeling - it's getting there, slowly but surely!

EvilDragon wrote:

Pianoteq doesn't use SAS at all, it never did .... Proper math algorithms all the way.

BTW - and SHAME on Roland for using those perverse improper SAS algorithms! What were they trying to push on us?

Re: Modeling - it's getting there, slowly but surely!

http://ssl7.ovh.net/~pianoteq/philippe/

Re: Modeling - it's getting there, slowly but surely!

This is an amazing link! Thank you! I wonder what path leads Phillip to the synthesis of the pianoforte. This inspires mathematics and development.

Re: Modeling - it's getting there, slowly but surely!

Thanks!

Re: Modeling - it's getting there, slowly but surely!

Pianoteq developers team member Dr Juliette Chabassier's papers (and PhD thesis) are well worth reading if you are technically inclined and really interested in some of the nitty-gritty.
https://scholar.google.com/scholar?hl=e...sier+piano

3/2 = 5

Re: Modeling - it's getting there, slowly but surely!

Great direction! On the above link I saw the work:
https://scholar.googleusercontent.com/s...s_ylo=2017
Completely fresh work. It clarifies the existing model of the string. Clarify the interaction with the bridge. The author puts the following questions, which should be investigated. Terrific! As I see, this is an actively researched direction. And the piano hides many more discoveries and clarifications. Therefore, the following years will definitely be productive in the direction of clarifying the piano sound. I'm inspired by great hopes!