Planetside Software Forums

General => Image Sharing => Topic started by: Dune on December 07, 2022, 06:38:25 AM

Title: Tundra and light settings... and stones test
Post by: Dune on December 07, 2022, 06:38:25 AM
I tried the lighter secondary atmo in this test, and it works good on terrain and objects. But I have the idea it also affects the clouds.

Another thing I tried here was sinking the fake stones into the snow, as if snow were melting around the stones (due to their heat absorption). But that's hard (literally), because the fake stones are derived from hard voronoi noise (smoothing won't work), and edges will show that voronoi shape. Would be great if fake stones could be based on soft spots/dots.
I did try stones from a soft dotty mask (raise, then compute normal, then lateral displacement), but coloring them is difficult.

These are not those stones, btw, this was my first try getting stones in.
Title: Re: Tundra and light settings... and stones test
Post by: KlausK on December 07, 2022, 07:06:26 AM
Would`nt it be easier to create some nice stone/rock obj`s and import them?

ps:
The scenery reminds me of my vacation in Sweden in Sep-Oct this year. As if you made it after this...see for yourself ;)

CHeers, Klaus
Title: Re: Tundra and light settings... and stones test
Post by: Dune on December 07, 2022, 07:58:57 AM
It would probably, at least for the front stones. But I am also looking for the perfect way to have the snow melt around the stones, so that should be procedural.

Nice ref!
Title: Re: Tundra and light settings... and stones test
Post by: pclavett on December 07, 2022, 01:15:39 PM
Beautiful render of tundra Ulco ! Love the way the snow was set up ! The vegetation is also quite nice !
Title: Re: Tundra and light settings... and stones test
Post by: Dune on December 08, 2022, 03:27:16 AM
I found a way to do it, but I need to perfect it yet. For anyone interested in some trials (I'm not posting a tgd, that would be too easy); use a ( blue node ) get position in texture, a perlin 3D and a constant scalar for scale, then a smooth step, then a color adjust unclamped. From the smooth step displace up by displacement shader, followed by a compute normal (very small patch), fed into child of surface shader, which is masked by the color adjust (which needs some tightening of WP and BP), give the surface shader some (rather much) offset laterally. Then feed that surface shader into the main input of another surface shader, masked by the same color adjust, and have the initial displacement shader fed into its child. Then give the second surface shader some negative vertical offset.

Now I want to try have some stones 'oversnowed', and more control over placement.

These stones are only on the snow here (displaced in test 2), as that's where I sat them on. So on dry land I made a different set, consisting of 'ordinary' fake stones. 
Title: Re: Tundra and light settings... and stones test
Post by: Hannes on December 08, 2022, 03:57:30 AM
Wow, that looks insanely cool!! Clever solution!
Title: Re: Tundra and light settings... and stones test
Post by: masonspappy on December 08, 2022, 04:13:53 AM
Wow. Some people are "next-level". Dune you are "next-next-level". :o
Title: Re: Tundra and light settings... and stones test
Post by: Stormlord on December 08, 2022, 07:24:33 AM
Superlevel!

STORMLORD
Title: Re: Tundra and light settings... and stones test
Post by: pixelpusher636 on December 08, 2022, 09:20:09 PM
The rock looks good but that veg looks fantastic. The reds really add to the realism of the scene.
Title: Re: Tundra and light settings... and stones test
Post by: Dune on December 09, 2022, 02:59:02 AM
Not entirely satisfied with the stones yet (too flat), so the experimentation is not over yet. Strange things happen with some settings and combinations; like that the stones are okay in the snow but raised on pyramidic mounds in dry areas. Or that the smoothing and lateral displacement displaces them up and sideways in a kind of circular way. The first has to do with a transform shader (world) I used, which kind of settles the displacement, not to be smoothed anymore, the second may have to do with the patch size of the initial compute terrain, or the fact that the localized (by stone mask) compute normal isn't exactly located at the stone location due to a missing XYZ scalar. Which I can't use for other reasons (like the inability to smooth later). Pffff, wish it would be easy. Intriguing though.

This is RT rendered, btw, was a bit faster for testing.
Title: Re: Tundra and light settings... and stones test
Post by: pclavett on December 09, 2022, 08:13:13 AM
Looks pretty good to me ! Love the veggies and the way you placed the humans with one looking over the small distant group ! Great work !
Title: Re: Tundra and light settings... and stones test
Post by: sboerner on December 09, 2022, 09:03:25 AM
Interesting! Thanks for the detailed description of your solution . . . I will be giving that a try. I've had trouble getting fake stones and intersect underlying to play well together, but on the other hand I haven't spent much time trying to get it to work. You've inspired me to try again!
Title: Re: Tundra and light settings... and stones test
Post by: Dune on December 10, 2022, 02:14:58 AM
I have to warn you, it's not really easy to get the parameters right. There must be a better way, so I'm thinking of lowering a soft part first and then raising a hardened section with some laterals. Texturing has to fit though. I also tried an image map with some 16-bit soft white dots. That works nicely. And you can also use a perlin fractal or voronoi fractal instead of the perlin 3D, but with stretched Y, which also works better.

Vector displacement (map) is another option.
Title: Re: Tundra and light settings... and stones test
Post by: Dune on December 10, 2022, 04:20:43 AM
It's indeed better to use an image map (PF don't work well) and make a small dent first to lower the snow hole, then reduce gamma from image map and make slight upward displacement in the center, then a compute normal and then a displace lateral and another displace by normal for fine-tuning, both fed by the image map, through a color adjust.

So the small central dent will be displaced up and outward a bit first. If you don't make the small dent up first, the lateral/normal displacement will displace from the concave sides and intersect. Ugly!
Title: Re: Tundra and light settings... and stones test
Post by: Hannes on December 10, 2022, 07:44:44 AM
This looks quite well. Sometimes image maps are the best.
Title: Re: Tundra and light settings... and stones test
Post by: sboerner on December 10, 2022, 08:29:23 AM
Yes, that looks very nice. Photoshop has several scattering dot brushes so you could create an image map very quickly.
Title: Re: Tundra and light settings... and stones test
Post by: masonspappy on December 10, 2022, 08:43:52 AM
Very realistic. Just like the stones I sometimes saw when I lived in places where it snowed.
Title: Re: Tundra and light settings... and stones test
Post by: mhaze on December 10, 2022, 12:57:03 PM
Clever! Great images too!
Title: Re: Tundra and light settings... and stones test
Post by: Dune on December 11, 2022, 04:23:35 AM
The mask is very simple black with a few round soft dots, so no fancy brushes needed. You can make them elongated of course, but a bit (not too much) of warp in TG also helps.

Two versions of the stone mask applied, but the smaller needs a tighter mask for texture, as you see. It's fickle.