silly physical models

Official support for: xoxos
Post Reply New Topic
RELATED
PRODUCTS

Post

but first, a few updates..

pling4 is in alpha (instruments forum) and is kinda nice.

a suite of commercial effects has been released (effects forum) some are nice, some are "just sort of there anyway" :)


i've had a rebirth with hermite interpolation.. i'd discarded it when i first started developing as cpu intensive. however, for the cpu cost (i must have been stringent then..) it had significantly better reproduction of high frequencies, such that the karplus-strong implementation in pling4 shows little to no change in amplitude decay across the scale.


on a whim, i have updated piamo with the same interpolation method. there's a fair bit about that model that could be improved.. the hermite version uses about 3.5% of my 1.6g cpu per note (linear uses ~2.5%). someday i'll have another go at it and probably be able to knock about half of hte cpu off there, but there's lots else i'd like to change so not today.


if you are new to piamo, it is a simple modeled piano.

pianos are difficult to model - their complex forms produce complex harmonic resonances, which are going to be judged critically by musicians. the simplified harmonics of a physical model rarely impress unless samples are used in the impulse or using other methods.

but, being me, i wanted a mathematical model without samples.

i think the worst thing about it is the very simple cabinet model - it's more like a reverb. you have to use your imagination that it's a wooden cabinet :) that can probably be improved at present..

the keenest thing, from my perspective about piamo is that instead of using impulse samples, the hammers are modeled with mass-spring models, which interact with the strings.. felt piano hammers usually bounce off the string a few times when a note is played, which produces harmonics.. without having access to conduct analysis of this interaction, i used low/allpass filtered waveguide strings and third order mass-springs.

the frequency of each order can be set from the gui. there are some things i can improve about my model (current mass-springs have a higher dynamic response for higher frequencies which can be unrealistic) - personally, with the slight amount of knowledge used to implement them, i think the timbre of the hammer-string interaction in this model is impressive.

anyway - as said, it doesn't sound very different, (it sort of sounds like a simple, plasticy, piano with rounded corners..) the original version is probably the more serviceable, but there it is :)

both versions are included in the zip, install only the one you wish to use.

http://www.xoxos.net/vst/piamo.zip
you come and go, you come and go. amitabha neither a follower nor a leader be tagore "where roads are made i lose my way" where there is certainty, consideration is absent.

Post Reply

Return to “xoxos”