I am a bit disappointed that nobody dives into this, so I'll post again in a separate thread.
My head is spinnning around it, but it needs a clear view perhaps. The forum used to be so experimental and challenging, so here's your chance, wise guys! Figure out how to translate the white radiant lines onto a flat surface.
It would be a very decent way to enhance planet/crater landscapes. The same dots to be used in reduced size for the craters (and with a rim, using a merge shader), and the radiants as mask for (non-)rock distribution.
Here is a version of my test file (probably easier then the one with multiple craters, it's a mess). I too really am interested in other peoples take on this. Pretty sure I asked people to take a shot at it.
The displacement along the y forces the noise to spread out so when flattening it the noise contracts because it is no long pushed upward to the stretching effect. I wonder if the stretched noise can be used as a color mask and connected to a Surface layer then it would be flat for that layer.
Quote from: Chris on November 12, 2015, 03:44:12 AM
The displacement along the y forces the noise to spread out so when flattening it the noise contracts because it is no long pushed upward to the stretching effect. I wonder if the stretched noise can be used as a color mask and connected to a Surface layer then it would be flat for that layer.
That's what I did, used it as a colour b/w image for displacement/masking.
In your recent file you posted? Because i have not looked at it yet.
What you're saying in reply #2, Chris, is the point that needs to be solved. How to translate the visible (stretched) lines to a flat surface. Anyone?
I'm not really sure what the end effect is Ulco. I've missed any earlier threads about the problem.
I've had a go at a crater shader which is a bit rough at the moment. Looks like the Clangers live there. I'll play a bit more with it and post something.
I was kinda hoping/expecting you would chime in, Jon. Typically a blue node challenge for you ;) Your idea looks good. if you'd mix that with rougher ground. My idea was to use the white radial lines from the distorted normal setup, but they only exist on the cones, and I can't seem to translate them to flat ground. On the cones they don't even work for displacement, very weird.
Ulco - your setup is indeed very weird as it seems that the stretched lines are
not accepted by anything.Maybe they need to be computed first or maybe it's
another weirdness of the "Distort normal" feature.Whatever it is it's beyond me.
And I'm busy with other time consuming stuff.
Jon's approach looks promising,hope you guys can figure out something cool!
Not finished with this but I'll post it up so others can contribute.
[attachimg=1]
Quote from: Dune on November 12, 2015, 10:50:37 AM
....from the distorted normal setup, but they only exist on the cones, and I can't seem to translate them to flat ground. On the cones they don't even work for displacement, very weird.
The problem with that method, as you've found out, is that it needs direction to spread the ejecta. The direction is coming from the displaced terrain via it's normal. If the ground is flat, there is no normal. I think you could get this to work if you subtract the displacement after the mask is made. A bit messy though, but the effect could be better than my attempt.
That may be an idea. I was thinking of nodes like visualize normals, etc. but couldn't get any further.
EDIT: Your file is really great to investigate! I've broken up the hard rims by a PF instead of the constants, that helps to decrease their hardness.
The procedural craters is nice, and work fairly well with actual terrain. One issue though is the ejecta being globa (and not based on a craters position) makes it a little too hard to actually have ejecta displacement. For example, you have the ejecta fully masked, but when you convert it to scalar and multiply to add displacement, and then pipe it into the surface layers displacement, you get weird lines all over the body of the planet from displacement that should be masked and not having to do with a crater. See straight vertical lines in preview below.
(http://i.imgur.com/gV2uUGD.png)
I actually haven't tried to figure out what ejecta really are; should they displace or are they just barren of rock debris, as former deposited debris is blown away by new larger rocks ejecting from the latest hit? The latter I've tried with Jon's setup, and that worked very nicely. As soon as I have time I will try to soften the rims with another partial blue node setup. But that may take some time :P
I suppose there are "types". From images I've inspected there are streaks cut into the terrain, other seems like scorched rock. For example, Phobo's largest impact Stickney has some very noticeable ejecta displacement.
(https://upload.wikimedia.org/wikipedia/commons/5/59/Stickney_mro.jpg)
http://www.wisegeek.com/what-is-ejecta.htm
Here's a nice crater with directional ejecta you could visit in Canada. :D (I wish I could at least)
https://en.wikipedia.org/wiki/Pingualuit_crater
If you put a colour adjust shader after the warp shader on the ejecta mask in my previous set up, you should be able to raise the black input level which will 'pull back' the edges you're seeing. You may want to lower the white level to compensate the 'pull back' on the ejecta shape, maybe a little gamma adjustment too.
You may have to reduce the crater size and increase the scale to get this all to marry up.
The same technique could be used just before the crater displacement shader if your displacement is forming these cell edges.
Quote from: Hetzen on November 23, 2015, 07:30:22 AM
If you put a colour adjust shader after the warp shader on the ejecta mask in my previous set up, you should be able to raise the black input level which will 'pull back' the edges you're seeing. You may want to lower the white level to compensate the 'pull back' on the ejecta shape, maybe a little gamma adjustment too.
You may have to reduce the crater size and increase the scale to get this all to marry up.
The same technique could be used just before the crater displacement shader if your displacement is forming these cell edges.
Yeah, the adjustment is now grey-scale and through a colour adjust shader to create a mask for ejecta displacement. The visual preview shows the mask should only be around craters, however it is global despite there being no visiable detail to create it anywhere else but the craters. So something else is interfering with the setup. It's acting as though there are parts of the mask there that aren't there. I even tried using the Colour Adjust to create a absolute black and absolute white on the craters, and it was still global. Its as though despite being converted to greyscale, the scalar from the original texture is still passing through and creating displacement where it has been removed.
I'm sill going to tackle this but it should be easy now using inverted erosion along a soft outer mask for the crater.
Quote from: Chris on December 05, 2015, 02:41:46 AM
I'm sill going to tackle this but it should be easy now using inverted erosion along a soft outer mask for the crater.
I think it may work better when the erosion shader has a update. Currently it's floor model may make erosion way to soft.