Planetside Software Forums

General => Image Sharing => Topic started by: cyphyr on September 30, 2009, 06:15:45 PM

Title: Cirrus cloud studdy experiments
Post by: cyphyr on September 30, 2009, 06:15:45 PM
I''ve been trying to get a wispy "Angle Hair" (I think) cirrus clouds.
A way to go I think yet :)
Richard
Title: Re: Cirrus cloud studdy experiments
Post by: Tangled-Universe on September 30, 2009, 06:20:48 PM
Cool :) The first one will probably also be very nice as blendshader for cumulus like clouds!
Title: Re: Cirrus cloud studdy experiments
Post by: Henry Blewer on September 30, 2009, 07:20:08 PM
They both are unusual and look great. A voronoi diff function plugged into the blend shader works for me. Then I used ridged noise stretched somewhat in the direction they tend to 'travel' along. The ones I make are not NWDA, but they look good. I often make them very wispy and soften the edges.
Title: Re: Cirrus cloud studdy experiments
Post by: CCC on October 01, 2009, 12:51:32 AM
They look very marbleized but what is needed is a self-similar dissipating edge iteration for that feathery cirrus look. This is very hard to get down. As for these clouds, good for a more alien/fantasy look.
Title: Re: Cirrus cloud studdy experiments
Post by: inkydigit on October 01, 2009, 05:07:49 AM
pretty trippy looking, and yes they might make nice blenders/warpers
? for cumuli
Title: Re: Cirrus cloud studdy experiments
Post by: cyphyr on October 01, 2009, 08:59:42 AM
And another one :)
Don't take these as finished in any way, they are what it "says on the tin", just studies. Two dudes with the same suggestion, using as a blend shader for cumulus ... how would you do that ??
:)
Richard
Title: Re: Cirrus cloud studdy experiments
Post by: Tangled-Universe on October 01, 2009, 09:28:06 AM
Quote from: cyphyr on October 01, 2009, 08:59:42 AM
And another one :)
Don't take these as finished in any way, they are what it "says on the tin", just studies. Two dudes with the same suggestion, using as a blend shader for cumulus ... how would you do that ??
:)
Richard

Well...I guess you know how to use a fractal as blendshader for another fractal :)
Title: Re: Cirrus cloud studdy experiments
Post by: j meyer on October 01, 2009, 10:00:57 AM
 8) Nice cloud studies!
Title: Re: Cirrus cloud studdy experiments
Post by: cyphyr on October 01, 2009, 10:08:13 AM
Maybe I'm just being dense but why would I want to use a cirrus to blend a cumulus?, I'll give it a try but surely the scales are all wrong...
:)
Richard
Title: Re: Cirrus cloud studdy experiments
Post by: dandelO on October 01, 2009, 11:07:38 AM
The cirrus cloud fractal scale would be massive in comparison to the cumulus, that means that the cumulus fractal will only appear where the cirrus fractal describes, not the cloud layer, only its fractal is what Martin means.

Try and make a new layer of clouds(low), disable the cirrus layer and use its fractal as the blend shader to the cumulus fractal...
Title: Re: Cirrus cloud studdy experiments
Post by: Henry Blewer on October 01, 2009, 11:41:57 AM
You could just plug in a power fractal with the same scales, alter the noise. So if the cirrus layer uses ridged noise, try perlin or billows. Getting really extreme would be to use functions, like voronoi. If you use a function, remember to use a get function (position?) for the voronoi input.
Title: Re: Cirrus cloud studdy experiments
Post by: Tangled-Universe on October 01, 2009, 03:12:20 PM
Quote from: cyphyr on October 01, 2009, 10:08:13 AM
Maybe I'm just being dense but why would I want to use a cirrus to blend a cumulus?, I'll give it a try but surely the scales are all wrong...
:)
Richard

Yes you're being dense Richard ;) The fractal shape of course, not the cirrus...Like Martin already said.
The scales probably need to be a bit bigger, but the warping will give interesting effects.
If that doesn't work to your taste then njeneb's suggestion could work as well, though it would give a different kind of result.
Title: Re: Cirrus cloud studdy experiments
Post by: Henry Blewer on October 01, 2009, 03:29:30 PM
I have found the voronoi functions do produce odd results.
Title: Re: Cirrus cloud studdy experiments
Post by: cyphyr on October 01, 2009, 05:54:06 PM
Voronoi functions can produce some nice results, I was playing with them a year or two ago but the render times were horrendous! I'd like to try to combine them with these warped fractal cirrus and see if even more "realistic" high cirrus clouds.

@TU :) Sorry but I'm still confused. Do you mean to use cirrus (ish, I know) clouds as a blend shader for cumulus clouds (500m to 3500m ish) or to use
cumulus (+its attendant low level scale cloud fractal) at the cirrus height (about 12000m I think)

:)

Richard?
Title: Re: Cirrus cloud studdy experiments
Post by: dandelO on October 01, 2009, 06:17:58 PM
I'm sure Martin means to just use the fractal shader you have for your cirrus cloud as the blend shader to a cumulus layer's fractal, lower down in altitude. The density fractal is really just a 2D black/white map, it has no altitude value at all.
The depth and heights come from the cloud layer. So you use the cirrus fractal as the blend shader to a cumulus density fractal.

I've been trying much the same thing to get more levels of detail in clouds, I'm looking to make cumulus-fractus, large dominant features and smaller wispy edges. I'm using large fractals with normal coverage and following that with a smaller fractal with higher coverage, the small fractal is blended by the first, larger fractal so it only appears where there are clouds to begin with.
Not exactly what Martin's saying but of the same idea.
Title: Re: Cirrus cloud studdy experiments
Post by: Zairyn Arsyn on October 01, 2009, 11:51:07 PM
cirrus4 looks unreal, and the least generic, :)
cant wait to see more!
Title: Re: Cirrus cloud studdy experiments
Post by: cyphyr on October 04, 2009, 05:00:50 PM
TU and dandelO both suggested I use the cirrus layer as a blend shader for a Cumulus layer. Heres a couple of very quickly set up renders following up on the idea :)
Interesting results and I think I'll play further. The first is the Cirrus layer (from Cirrus 3 (http://forums.planetside.co.uk/index.php?action=dlattach;topic=7679.0;attach=19589;image)) plugged directly into a cumulus layer and in the second its used as a blend shader for a Perlin Billows Cloud fractal. Going to try some renders from above next ... getting some ideas :)
enjoy and C&C (but do note this is just trying stuff out :) )
Richard
Title: Re: Cirrus cloud studdy experiments
Post by: Henry Blewer on October 04, 2009, 05:45:21 PM
Both of these are really fantastic. I like the wispy feel of the first, but the light of the second seems more natural. Try decreasing the cloud density by about 1/3 on the second image. This will allow more 'glow' to show through.
Title: Re: Cirrus cloud studdy experiments
Post by: Kadri on October 04, 2009, 07:37:51 PM
The first ones where interessant,but these are very nice.Just curious what was the render settings?İf rendered with high settings how would they look?İ am thinking about the details?

Kadri
Title: Re: Cirrus cloud studdy experiments
Post by: CCC on October 04, 2009, 09:16:37 PM
Sense the edges are of a cirrus blend, this would work well for anvils from super cell thunderheads.