I have absolutely no idea what I am going to do. So far only some new rock structures tested.
You're a winner anyway, so pick any ;) I love these structures, but liggend 10 is the most realistic in my eyes. Great work, René!
It will probably come down to it again that I have so many tests that I can't make a choice. :P
Gee these are awesome! Perhaps after the contest you can make a tutorial how to do these or similar rocks?
Great start.
I really like jgfj.jj and liggend.
You got some really good square, angular shapes there.
Looking forward to more :)
Wow, first I thought there are some reference images as well... :)
I like especially the second one. Great rocks already :)
Spectacular! There are too many aspects of the various images to single any one out as best, but certainly whatever combination of your excellent technique that is submitted should yield a prize. Thankfully you can only enter one! ;)
Very nice Rene, I expected nothing less than this level of quality
8) 10 liggend and jgfj.jj are most promising.
Thank you all for your comments.
I think I'm going to explore this scene a little more.
Quote from: René on January 18, 2018, 05:34:16 AM
I think I'm going to explore this scene a little more.
Goood idea !
You know, looking at these structures, especially the ones from "10 liggend" makes me think it's best to drop the towel right now. ;) There's a core of truth to that, there's a little voice yelling "Give it up! You stand no chance!"... But on the other hand it makes me think "Ha, René may have the best rocks in town, but I still think I have the best ideas!" ;D
I'm really looking forward to see how this develops!
Good ideas look better with good rocks :P ;)
Go for it! :)
WIP
WIP
These last two are both excellent. I prefer the atmospheric qualities of the second one. The waterfall has promise though.
Thanks! The waterfall still requires a lot of work. I will only use it if it is credible.
WIP
I like them. Waterfall needs work indeed, or I would dump it. There's a strange artifact in the other 2. And beware of floating grasses.
I like especially the last one.
WIP
Very atmospheric. Where's the Orc fortress? :D
Quote from: cyphyr on January 23, 2018, 03:19:23 AM
Where's the Orc fortress? :D
Valid question for sure, looking good Rene !
It reminded me of a location/scene from Skyrim ...
Great, clouds and trees really do the trick to make it huge!
Wow some of these cliff WIPs you lot are knocking out are truly epic in scale.
Quote from: Dune on January 22, 2018, 11:51:37 AM
I like them. Waterfall needs work indeed, or I would dump it. There's a strange artifact in the other 2. And beware of floating grasses.
There is some exploding geometry here and there, and if I have decided what I am going to make, I will deal with that. If I can't make a realistic waterfall (probably) I won't use it. If I only could attach clouds to geometry, it would be a no-brainer. :)
WIP
Very good!
Great lighting.
This will be a contest for yourself; one about choosing ;)
Quote from: Dune on January 28, 2018, 02:11:28 AM
This will be a contest for yourself; one about choosing ;)
Yep! A few more weeks and then they can lock me up in a mental institution.
Look what I found on Streetview! [attach=1]
Spanish coast.
You're always driving random Spanish roads when the weather is crappy in The Netherlands, Andy? :P
QuoteYou're always driving random Spanish roads when the weather is crappy in The Netherlands, Andy?
So that you know it: They're not random but selected for their proximity to a coast!! :P And it's what I often do while waiting for a render to finish... ;D
Quote from: René on January 22, 2018, 09:43:28 AM
Thanks! The waterfall still requires a lot of work. I will only use it if it is credible.
very good work rene, BTW I hope you still go forward with the waterfall. I see a win entrance in that one...
Cliffs just beg for a waterfall. If I can pull it off there certainly will be one or more waterfalls.
WIP
Another set-up with tons of flaws.
Another attempt at creating a waterfall. :-\
And quite a succesful one! You probably also used a plane conforming to the side of the rock, with some fractal displacements, and perhaps a no-RT reflection. Though reflection is probably not even needed, just some translucency. Top and bottom are always the hardest, but this looks really good.
I have used a card object with three displacement maps, one for the horizontal rounding, and one for the lateral rounding so that the card gets a bend. On top of that, there is a displacement for the water structure. And of course a transparency map.
The problem is that I constantly have to go back and forth between Photoshop and Terragen and that the smallest differences in gray value have major consequences.
What I don't manage to achieve is the softness of the edges of the waterfall because the transparency has only two values, transparent or not, so no soft transition. I have tried to solve this by building up a grid of very small dots in the transition.
It is therefore basically a proof of concept
One of the best I've seen. It really is time that we had grey scale transparency!
That would be useful right now. :)
INdeed!!! The waterfall looks fantastic, René!
Excellent work on the waterfall so far, as well as the rest of your scene!
So what do you do in Photoshop, paint a mask?
For soft edges I've used a glass shader as a base, with soft edges. That would be the only way for now. My main problem with waterfalls was to have sidestreams rushing over stones and 'getting back' to stream further down, not ending in nothing.
I painted a mask for the contours of the waterfall. For the bending (lateral and horizontal) of the card object, I used the gradient tool in Photoshop corrected with the curves tool to create a displacement map. The water structure is also a displacement map created using a combination of photos and painting, but I think that I can achieve better results with procedural textures.,
Rocks sticking through the waterfall are indeed problematic, so you have to find a place suitable for the placement of a waterfall, or prepare a part of the terrain. I intend to investigate whether the waterfall can be inflected naturally around the rock by using the rock displacement as a warper.
I hope that this gibberish makes sense. :)
It does.
I once (or twice) used the main terrain without the finer displacements, and fed that into a plane. Then dented the main terrain where the fall is supposed to be (but hard to find the perfect spot), so only the rather smooth plane displacements are visible there. Then applied fall displacements and colors, and if the sides were off the rock and supposed to be gradual used a water shader (and later a glass shader). Something like that.
I'm going to try the glass shader, but I'm afraid that the render times will skyrocket, mainly because I also want to use some clouds for splashing, foam and fog.
Suppose that I get this working then it must also be possible to animate the waterfall. However, this will be something for after the contest. :)
You needn't make the glass shader reflective, and that makes a big difference!
Cleaver construction, great mood. This is certainly going to be a top 3 contender
Quote from: mhaze on February 01, 2018, 05:18:47 AM
One of the best I've seen. It really is time that we had grey scale transparency!
hear hear
An amazing start René and thanks again for revealing the "Easter Egg" of displacing cards..displacement any where and at any angle...I'll be forgetting about the Plane entirely now.
And thanks for talking about method as us 'not so sharp' ones really benefit from it.
Exploring a new kind of rock.
IMHO much better.
Brilliant, I've seen cliffs like that.
8) Nicely shaped rocks these are.
Nice sharp edged displacements!
Oh, that's beautiful, René!
Indeed! Great shapes and colors!
That's quite impressive!
Outstanding rock shapes and foliage placement. Fine work!
Ground breaking stuff Rene
8 hrs and going. With such long render times, a resume rendering capability is indispensable. The render has just started with the lake, so that means I can't work with Terragen for the rest of the day.
Looks good though doesn't it! There's some really nice cracks and subtle mix of scales. Doesn't look too stretched either. The only critique I have is that I think the trees could be blended better.
The distribution of the trees is bad and I used only one type of tree. The colors are also wrong. But I will take care of that later when I am sure which tree species I am going to use. For now, I concentrate only on the rocks. :)
I don't think its bad, just randomly regular.
I wonder how long that water is going to take? If it becomes too much it might be worth investigating using a dumbed down version of your mountains for a separate water render.
Does your render need the water shader's transparency? If not, set it to 0, which will decrease render times. I guess it would be similar to Ulco's advice to just use a slightly displaced reflective shader instead of the water shader.
Transparency has already been switched off. It's not only the water that slows down rendering but especially the textures that I used; two types of Voronoi merged together plus an alpine shader. In the end I will have to render in parts and composite them together.
Oh, the alpine shader... That's indeed increasing rendertimes.
Quote from: René on February 14, 2018, 03:21:10 AM
8 hrs and going. With such long render times, a resume rendering capability is indispensable. The render has just started with the lake, so that means I can't work with Terragen for the rest of the day.
In future, maybe set a limit to the amount of threads available and run 2 Terragens? That way you can render on one and set up something else to use on the other. Sure your render will be slower but you can still get on with things. I often have a few open. It's possible on MacOS too. ;)
Great potential in this one. I tend to turn atmosphere off when test rendering, the new clouds , though nice, tend to be slow to render especially when defer atmosphere is turned on!
Looks like it will be another Rene special!
I have resorted to sometimes setting up a scene on the old kermudgeon, transferring it to render on the new one while I rest, because I have the TG bug too. Going blind watching black squares turn color in anticipation.....haha :)
I temporarily changed the water shader with a default/power fractal to speed up rendering. Still quite a lot to be done.
It's a beauty already...
Excellent cliff and light!
Excellent, great atmosphere and composition - looking forward to see were this goes.
Absolutely loving the tree waterfall pop where it looks like a waterfall of trees, and well the whole darn thing is amazing.
Looking promising indeed. The rays coming down from the cliff on the right are fantastic. Don't lose those! :D
- Oshyan
Another idea to explore. I will have to take a decision very soon.
Rene the lighting is the pure magic :o ::) ;)
And another beauty!
Exceptional piece
Oh yes!! Fantastic!
Lovely! Great isolation of accent elements as always. :)
Well I will enter this comp, but have no illusions about who's gonna be top o the heap...well done once more. Must be nice having the luxury of choice when both are winners heh heh
Last test (I hope) :o
I hope not, I like your tests ;) This is incredible too. I sometimes wonder if we use kind of the same method by (happy) accidence.... how many compute terrains do you need?
Dune: None. I hardly ever need them.
Looks great Rene. No displacement 'fins'. Really well controlled.
Quote from: Dune on February 27, 2018, 02:01:56 AM
how many compute terrains do you need?
If you start with large displacements with no roughness and work from there to smaller ones with more roughness, you don't really need any compute terrains. I do almost all displacements with shaders and keep the terrain is as simple as possible.
Good insight into your technique, thanks. Pretty much the same proven process as sculpting or painting, start with the large areas and work down to the finer details.
Quote from: zaxxon on March 01, 2018, 12:26:26 PM
Good insight into your technique, thanks. Pretty much the same proven process as sculpting or painting, start with the large areas and work down to the finer details.
Exactly! :D
No compute normals either? For 'regular' lateral displacements you need one at least, unless you work with redirects, vector displacements, or fake stones, which I do mostly too. It's a bit harder if you need to use a minimum slope for something like textures, but that too can be overcome by overlaying textures with maximum slope restraint.
And too many vector displacements also take up render time, so one compute normal or compute terrain and using lateral/normal displacement might yield the same speed. It's an interesting playfield anyway, trying to get maximum result with minimum nodes and speed.
I have two compute terrains, endless displacements and complex masking! takes about 40 mins to render add in veg another half hour add in two small trickles of water which have to have transparency to work and render time rockets to 13.33 hrs!
I also didn't use any compute normal. As you can see in the picture the terrain is very steep. The only displacement shader is a strata shader. I did use 'use world space (final position') for all shaders to avoid stretching.
You can also use a surface shader to straighten(even after texturing) the terrain even more: limit by altitude and lateral displacement offset
As far as I know the lateral displacement offset is not working as a mask (lateral displacement needs a compute normal anyway), but I have to check out what you exactly did here (maybe you found some secret I didn't know yet ;)). I do know that you can stretch by altitude offset and/or slope limit. You can also use smoothing effect with a minus number, btw.
I'm learning a lot here! Thank you, René & Ulco!
As far as I know both lateral displacement only -and normalized work fine without a compute normal. It may be that things are lost in translation, :-\ so I will get back to this when this contest is over. If I forget, don't hesitate to remind me. :)
Yes, we have to sort this out, because why otherwise does it say; compute normal needed? Fake stones have an inbuilt 'normal calculation' as far as I can tell, hence the possibility to use those for the lateral displacements without compute normal/terrain.
This is interesting!! Looks great!
Hmm, I find that the compute normal is needed for true lateral displacements although not for Vector displacements. Often a normal displacement can look just as effective as a lateral and render cheaper!
Sometimes adding in a texture coordinates XYZ can help too.
Good canyon you got going on there.
Rendering cheaper? Never noticed that actually, but if that's so, it's good to know. Yes, I tend to use one compute normal at least, as it's easier then to get nice laterals. VDisp is always slanting one way, unless you make a mix, but that takes time too. And my latest experiments didn't quite yield what I wanted. And XYZ's I use quite a lot.
Epic stuff Rene