Deviation Actions

greenhybrid's avatar

When the days slowly settle II

By greenhybrid
3 Favourites
35 Comments
1K Views
I wish I had a nice ship model and a mesh loader in picogen :S

download-size: 1680*1050
Image details
Image size
1680x1050px 1.11 MB
Published:
© 2008 - 2021 greenhybrid
Comments35
Join the community to add your comment. Already a deviant? Log In
hmm, I can't see my comment here, and I see a duplicate of the first comment with the name "nobody"

I do believe I may see a bug.
kram1032's avatar
A ship made of Perlin-noise? :D *hint*

Really nice!
greenhybrid's avatar
fractal spaceships, maybe :D

i think i'll leave that to aexion :P
kram1032's avatar
yeah, work together and get a fractal importer :D
kram1032's avatar
yay mib :D
I miss a fractalizer or defractalizer weapon in the movies :D
(Either make something SO fractal that it dies on complexity or SO, uh, let's say "euclidean", that it dies on order :D)
lyc's avatar
so, i take it you have a new ~20inch screen? :D

very good render! the water reflection is nice, surface is FBm right?
greenhybrid's avatar
22inch, yay :D
I increased res from 1280x874 to 1680x1050, plus the additional contrast. super-awesome :)
--

I am not sure, isn't Perlin-Noise a form of FBm?

If so, then the answer is yes :)

It's not too easy to find information on how Perlin-Noise and FBm are related :|
Interesting... Do I see a bug in DA?


regardless... Picogen you say?

Hmm... Interesting...
greenhybrid's avatar
Where do you see a bug? Gonna hunt it down if they give me root access :D

Yes, picogen is the terrain rendering software I am writing, it's very soon, but those are the first results.

Maybe you can see more at [link] , or on [link] for a quick introduction :)
lyc's avatar
other way around, fbm is kinda like perlin noise in that you are adding together octaves of noise. in any case, what you've used there looks pretty cool :)
greenhybrid's avatar
thanks :)

i thought fbm (coming from biology?) was older than perlin, or can perlin be regarded as a generalization of fbm?
lyc's avatar
the basic idea is to add noise with exponentially higher frequencies, and then during evaluation cut off the addition of higher terms beyond the nyquist limit, replacing them with an average value (remembering to lerp out the last term with the fractional component of your octave number to make it smooth).
greenhybrid's avatar
lyc's avatar
since you like freedom, spare a thought for those in china who may not be able to fully participate: [link]
greenhybrid's avatar
interesting (but then, this is the sort of freedom (w.r.t. availability of information) you like, too, not?)
afaik, the ppl in north korea have a huge intranet, so even worse, but *afaik*
View all replies
greenhybrid's avatar
erm, hmm, the cutoff+average is also the trick of aa'ing terrain to smooth out edges of triangluated terrain, you showed me once, am i right?
lyc's avatar
the exe thing i showed (dl: [link] ) was cheap-and-fast prefiltering, this is pseudo-analytic aa (pseudo bc i dunno what the mathematical basis for lerping out the final term is).
greenhybrid's avatar
yeah i remember that :)
i wonder if perlin functions (with interchangeable base functions, like i want to implement next (i.e. i am going to have a 'balls' function which i want to use as a base function for my clouds)) can be aa'ed analytically without much pain.
greenhybrid's avatar
ah, nyquist limit is then what i am looking for when it gets noisy (frequency > screen-res, e.g.)? :)
lyc's avatar
yes, since anything above that will alias, and thus needs either supersampling or analytic frequency clamping as described.
greenhybrid's avatar
thx for pointing that out. while somehow obvious, it's something one (==me) can easily forget to think about. for my clouds this will be harder to analyze, unless i get unlazy and implement ray differentials as pharr describes (heh, i should read more pharr (esp. the sampling chapter) instead of books about WW II ... :P)
View all replies
Join the community to add your comment. Already a deviant? Log In