... In a gig you probably want the numbers constrained in spacetime in a reasonable way so that the audience don't have to wait millions of years for it to come back to a useful range.
Probably...the thing about Perlin noise (as I understand it now) is that it is predetermined, backwards from the idea of a random walk: you input two or three numbers and get as output a random number that is always the same for the same input, always similar for a nearby input, and uncorrelated for distant inputs. So it's like a surface that you can do a random, or any other kind of walk on. It's usually used in computer graphics to make terrain and texture. In 3D it can make 3D textures like marble or bread.
I suppose a pd 3D perlin object would have three inlets (plus messages to set parameters) and one outlet, and the random values would need to be calculated once and for all at init, or whenever the parameters change, so it would be a memory hog.
Martin
On Wed, 15 Nov 2006 11:52:31 -0500 martin.peach@sympatico.ca wrote:
... In a gig you probably want the numbers constrained in spacetime in a reasonable way so that the audience don't have to wait millions of years for it to come back to a useful range.
Probably...the thing about Perlin noise (as I understand it now) is that it is predetermined, backwards from the idea of a random walk: you input two or three numbers and get as output a random number that is always the same for the same input, always similar for a nearby input, and uncorrelated for distant inputs. So it's like a surface that you can do a random, or any other kind of walk on. It's usually used in computer graphics to make terrain and texture. In 3D it can make 3D textures like marble or bread.
I suppose a pd 3D perlin object would have three inlets (plus messages to set parameters) and one outlet, and the random values would need to be calculated once and for all at init, or whenever the parameters change, so it would be a memory hog.
Martin
I knew I'd seen it in applications context before, thanks for that.
Here's what looks like a pleasant overview
http://freespace.virgin.net/hugo.elias/models/m_perlin.htm
Andy
On Thu, 16 Nov 2006, padawan12 wrote:
Here's what looks like a pleasant overview http://freespace.virgin.net/hugo.elias/models/m_perlin.htm
It's pleasant except for the part where he claims that linearly interpolated Perlin noise looks like crap: I wouldn't make so bold a statement about that.
One nice thing about linearly interpolated Perlin noise is that it's easy to make it recursive, dividing the space into equal quadrants. Thus, when seen in terms of Haar wavelets, it becomes a very simple thing.
_ _ __ ___ _____ ________ _____________ _____________________ ... | Mathieu Bouchard - tél:+1.514.383.3801 - http://artengine.ca/matju | Freelance Digital Arts Engineer, Montréal QC Canada
-
martin.peach＠sympatico.ca -
Mathieu Bouchard -
padawan12