Planetside Software Forums

General => Image Sharing => Topic started by: Dune on May 21, 2014, 02:51:57 AM

Title: Niagara Falls
Post by: Dune on May 21, 2014, 02:51:57 AM
Testing something again, and ran into something I can't figure out (yet). I used a simple shape as a basin, warped by a perlin with minimum size of 5m. OK. I wanted to warp the top 10m of the falls by the same fractal, but with smallest size 1m (to soften the water falling from the top down), so I added them, masking the small sized fractal by a distribution shader with height restraint. Nice theory, I thought, but it doesn't work. Tried different setups, but still...  I'll make it work though, but it's harder than I anticipated.
The water is just a reflective shader, by the way, there's no lake or water shader used. 
Title: Re: Niagara Falls
Post by: bigben on May 21, 2014, 05:02:45 AM
Very interesting effect. 
Rather than change the scale of the second PF, maybe increase roughness and/or contrast?
Title: Re: Niagara Falls
Post by: Kadri on May 21, 2014, 05:20:22 AM

Looks interesting Ulco.
I have an ongoing project in that i will run into this problem too probably.
It is not a waterfall but the shape is kinda similar but rocky.
Curious what you will do.
Title: Re: Niagara Falls
Post by: N-drju on May 21, 2014, 05:35:31 AM
Not too bad I guess. It looks a little bit uniform but only after you take a closer look at it. It DOES look like waterfall tough.

Damnation, do you guys know where I could find that .pdf tutorial on making small, vertical waterfalls? The first page of the document has a picture of the waterfall with some mist on the water surface. I grabbed it once but probably deleted it. >:( You know what I'm referring to?
Title: Re: Niagara Falls
Post by: choronr on May 21, 2014, 09:26:00 AM
The beginnings of a very interesting evolution ...looking forward.
Title: Re: Niagara Falls
Post by: Dune on May 21, 2014, 11:23:36 AM
@Ben: it just doesn't work, I can't get the height to work right for the warp, neither with final position, nor position in terrain. So I just added another vector displace at a later level with height restricted additional small perlin. That works fine.
Title: Re: Niagara Falls
Post by: kaedorg on May 21, 2014, 11:32:41 AM
second is much better. It has more volume with the variation of colours.
still a problem with the white parts i would say

Title: Re: Niagara Falls
Post by: choronr on May 21, 2014, 12:06:59 PM
That is a big improvement.
Title: Re: Niagara Falls
Post by: oldm4n on May 21, 2014, 01:31:41 PM
very good image :)
Title: Re: Niagara Falls
Post by: Upon Infinity on May 21, 2014, 04:41:58 PM
Nice idea.  Good work.
Title: Re: Niagara Falls
Post by: DannyG on May 21, 2014, 04:50:56 PM
Thats excellent as always Ulco, very convincing
Title: Re: Niagara Falls
Post by: bigben on May 21, 2014, 05:13:22 PM
If you have a function doing the displacement then you don't really need an altitude mask. Your displacement function is one. Plug that into a colour adjust shader and you have very precise control over the mask. Gamma adjustment might be useful in this case as well.
Title: Re: Niagara Falls
Post by: TheBadger on May 21, 2014, 06:01:48 PM
Hi Ulco. looks very good so far. I downloaded sat data of this location a while ago. Hope to use realflow for the water.

Thanks for the good inspiration! Im sure you will get it the way you want!
Title: Re: Niagara Falls
Post by: Dune on May 22, 2014, 02:32:41 AM
It won't work from the top, see attached simple setup.  You can't get the distribution shader to work at that level. But like I wrote, the remedie is simple, just add finer VDISP after compute terrain.
Title: Re: Niagara Falls
Post by: Dune on May 23, 2014, 02:40:02 AM
Redone, but now with the falls attached to a plane for more control over the terrain.
Title: Re: Niagara Falls
Post by: bigben on May 23, 2014, 05:26:55 AM
Looks good.

This was the mask I was thinking of.
Title: Re: Niagara Falls
Post by: archonforest on May 23, 2014, 06:54:02 AM
Last picture is very nice!! ;)
Title: Re: Niagara Falls
Post by: Dune on May 23, 2014, 10:14:28 AM
@ Ben; I'm afraid it won't work that way as you need smaller displacements, not just a hardening or softening of the first fractal. In the same surface layer it is easier to just add a vector displaced small fractal with min height and min slope.
Title: Re: Niagara Falls
Post by: choronr on May 23, 2014, 10:55:19 AM
Very good. I like the results with the added low fog in the distance and the large, reddish rock outcrop and the change you've made with the clouds. I would expect you will back up the camera and show more of the foreground as you did with the first image?
Title: Re: Niagara Falls
Post by: j meyer on May 23, 2014, 12:12:23 PM
Like the latest version.
I wonder,if the problem might be caused by the stretching of the noise.
What do you think is the reason?
Title: Re: Niagara Falls
Post by: bigben on May 23, 2014, 04:52:46 PM
I was only suggesting that as a replacement to the altitude restriction. It automatic starts at the range of your larger displacement and can. Be easily tweaked with the colour adjust node without having to determine the the actual elevation.
Title: Re: Niagara Falls
Post by: Dune on May 24, 2014, 03:19:28 AM
@ Ben; I don't understand that, but it's early morning, so I'm not fully awake yet.
@ Jochen; I don't have a clue. It can't be the stretching, I suppose, as it's still restricted to a certain height.  I may find out, although it's not really important.

EDIT: I know what it is (quite logical actually); the downward displacement is done áfter the distribution shader limits the height of the small fractal, so it doesn't yet have a clue about what height it's referring to. No matter if you use final position, it just doesn't 'know' yet before it starts displacing.
Title: Re: Niagara Falls
Post by: Dune on May 24, 2014, 03:36:48 AM
I did have another play and came up with an even simpler solution, without the warp shader. Turns out it doesn't accept heights somehow  ???
Title: Re: Niagara Falls
Post by: oldm4n on May 24, 2014, 03:58:38 AM
the last image (on the previous page) is very good and especially the island which is simply superb (what kind of tree did you use for it).
Title: Re: Niagara Falls
Post by: j meyer on May 24, 2014, 12:32:39 PM
Agreed,it's not really important,but interesting.
If you are right then that might be the explanation for the effect I noticed
and that made me think of the stretching as being the possible culprit,too.
Don't have that much time for TG at the moment,so I just had a really quick
look.The aforementioned effect is that the displaced fine ripples start to show
with a min altitude of minus 2.At zero or minus 1 nothing shows (as expected),
but then it starts to show and its running all the way down to the ground,which
was surprising to me.That made me wonder,maybe a lack of pre-compute terrain
experience on my side,though.
Anyway,curious to take a look at the latest file tonight. :)