Kirk Andrews wrote on 14/02/2017 20:01:
> Hey everyone! Been a long time!
>
> I found myself in need of a creative outlet, and so I dug up my old POV files.
> When last I left off years ago, I had been working on trying to get more
> interesting water formations in my landscapes. I had apparently stopped mid-way
> through a little algorithm for tracing the flow of streams through the
> landscape, forming lakes in the valleys. I decided to try and finish it out a
> little more.
>
I like it, especially the details like rivers and clouds.
Paolo

Post a reply to this message

Thomas de Groot <tho### [at] degrootorg> wrote:

> Oh, I do! I do! This is coming at a time when I have been wondering how
> to fill those valleys with streams. Lately, I have been working with a
> combination and alternation of Wilbur, Bryce, Terrabrush, GeoControl,
> and World Machine on the same height_field, to generate landscapes where
> river valleys are eroded down. Problem is the water and I have not got
> that far indeed. So, I am quite interested in this development. It is
> looking very good!

> --
> Thomas

That erosion element is the missing piece here, of course. This scene looks like
it's never rained here before, and there are no actual streams or lake beds,
just water pouring over the surface.

I've made some improvements and have been trying to get another render done, but
unfortunately, I seem to be running into a bug with my build of POV running on
this Mac (the last of my PCs died a couple months ago). It'll parse fine, and
start rendering just fine, but it keeps crashing and closing mid render.

You're more than welcome to see what you can do with it though!

https://www.dropbox.com/s/1j0kes0x79u9zz9/Water3.pov?dl=0

Post a reply to this message

Stephen <mca### [at] aolcom> wrote:

> No, to get the water down you just need gravity. :-)
> Seriously, Would you intend to make close ups or an animation? These
> landscapes are generally quite static over short time spans. (I am sure
> Thomas might have an opinion.)
>
> It actually gives me an impression of the Scottish Highlands with
> interconnected lochs. It can't be because the sun is shining and it
> doesn't look cold. ;-)
> I like the pattern of vegetation as well.
> How long did it take to parse and render?
>
>
> --
>
> Regards
>      Stephen

I haven't played too much with animation, but suppose you could. It's not going
to look too fantastic if you do a close up of those streams, since they're made
of cones and spheres. You'd definitely be able to tell that up close.

I've thought that maybe I could switch to building a mesh for the streams
instead; that would probably yield better close-up potential.

At the moment, it takes about half a second per stream to parse. There are about
100 streams in this render. If I don't have other settings cranked up too high,
render times are just a few minutes right now.

Post a reply to this message

Ok, well, I cut way back on the number streams and finally got a finished
render. There's only 10 streams in this one.

Post a reply to this message

Attachments:
Download 'water3_03.jpg' (324 KB)

Preview of image 'water3_03.jpg'

On 16-2-2017 5:45, Kirk Andrews wrote:
> Ok, well, I cut way back on the number streams and finally got a finished
> render. There's only 10 streams in this one.
>

Like you said, with no provision for erosion, the streams here seem more 
artificial then from above, running over the (lowest point) surface. 
Nonetheless, it is impressive in my eyes! I am eager to experiment with 
your code. I just downloaded it, thanks.

-- 
Thomas

Post a reply to this message

On 2/16/2017 4:36 AM, Kirk Andrews wrote:
> Stephen <mca### [at] aolcom> wrote:
>
>> No, to get the water down you just need gravity. :-)
>> Seriously, Would you intend to make close ups or an animation? These
>> landscapes are generally quite static over short time spans. (I am sure
>> Thomas might have an opinion.)
>>

>
> I haven't played too much with animation, but suppose you could. It's not going
> to look too fantastic if you do a close up of those streams, since they're made
> of cones and spheres. You'd definitely be able to tell that up close.
>
> I've thought that maybe I could switch to building a mesh for the streams
> instead; that would probably yield better close-up potential.
>
> At the moment, it takes about half a second per stream to parse. There are about
> 100 streams in this render. If I don't have other settings cranked up too high,
> render times are just a few minutes right now.
>

They are good parsing times for something that looks complicated.

I only mentioned animation and close ups because you mentioned inertia. 
Which made me think of the lakes draining. But for that I think you 
would need to create the fluid dynamics in SDL or in a third party program.



-- 

Regards
     Stephen

Post a reply to this message

Thomas de Groot <tho### [at] degrootorg> wrote:
> On 16-2-2017 5:45, Kirk Andrews wrote:
> > Ok, well, I cut way back on the number streams and finally got a finished
> > render. There's only 10 streams in this one.
> >
>
> Like you said, with no provision for erosion, the streams here seem more
> artificial then from above, running over the (lowest point) surface.
> Nonetheless, it is impressive in my eyes! I am eager to experiment with
> your code. I just downloaded it, thanks.
>
> --
> Thomas

Looking forward to seeing what you can do with it!

Maybe you can help me figure out the bugs in my code. The lakes don't seem to be
filling up as much as they should, and I can't seem to figure out why.

Post a reply to this message

"Kirk Andrews" <kir### [at] tektonartcom> wrote:
> Thomas de Groot <tho### [at] degrootorg> wrote:
> > On 16-2-2017 5:45, Kirk Andrews wrote:
> > > Ok, well, I cut way back on the number streams and finally got a finished
> > > render. There's only 10 streams in this one.
> > >
> >
> > Like you said, with no provision for erosion, the streams here seem more
> > artificial then from above, running over the (lowest point) surface.
> > Nonetheless, it is impressive in my eyes! I am eager to experiment with
> > your code. I just downloaded it, thanks.
> >
> > --
> > Thomas
>
> Looking forward to seeing what you can do with it!
>
> Maybe you can help me figure out the bugs in my code. The lakes don't seem to be
> filling up as much as they should, and I can't seem to figure out why.

I'm impressed. Downloaded the code. More impressed that you did this in 426
lines. Rendered and got to 81% in a couple of minutes. The big lake has slowed
things down and it was 22 minutes to finish. 3.0 GHz 4 core I5.

I see what you mean about the lakes not filling up enough to overflow and form a
follow on stream. I'm guessing Thomas will crack the problem before I do.

This is very nice work.

JimT

Post a reply to this message

So, I'm starting to get my head around what's happening with the lakes not
filling up. As it searches for a point lower than the water level in an
increasing radius, it doesn't take into account whether or not the water can
actually flow that way.

So, for example, maybe the lake extends away to the north, beyond the current
testing radius, and there is a ridge to south, but then a lower point. Since the
lake extends away to the north, the algorithm detects that as the lowest point
(equal to the current water level), and therefore does not raise the water
level, but does keep expanding the radius. Eventually, it does detect a lower
point, but it's on the other side of the ridge. Right now, the algorithm has no
method to detect that the water can't get there from here, so it goes ahead and
jumps the stream through the rock to the other side.

Post a reply to this message

On 16-2-2017 18:28, JimT wrote:
> "Kirk Andrews" <kir### [at] tektonartcom> wrote:
>> Thomas de Groot <tho### [at] degrootorg> wrote:
>>> On 16-2-2017 5:45, Kirk Andrews wrote:
>>>> Ok, well, I cut way back on the number streams and finally got a finished
>>>> render. There's only 10 streams in this one.
>>>>
>>>
>>> Like you said, with no provision for erosion, the streams here seem more
>>> artificial then from above, running over the (lowest point) surface.
>>> Nonetheless, it is impressive in my eyes! I am eager to experiment with
>>> your code. I just downloaded it, thanks.
>>>
>>> --
>>> Thomas
>>
>> Looking forward to seeing what you can do with it!
>>
>> Maybe you can help me figure out the bugs in my code. The lakes don't seem to be
>> filling up as much as they should, and I can't seem to figure out why.
>
> I'm impressed. Downloaded the code. More impressed that you did this in 426
> lines. Rendered and got to 81% in a couple of minutes. The big lake has slowed
> things down and it was 22 minutes to finish. 3.0 GHz 4 core I5.

My initial render was not finished after 4 hours, using 6 cores and i7. 
Just switch to the defined 'WaterTex' texture and forget the one at the 
end of the scene. The culprit probably is the finish. The whole image 
now renders in a couple of minutes.

>
> I see what you mean about the lakes not filling up enough to overflow and form a
> follow on stream. I'm guessing Thomas will crack the problem before I do.

I am not that far yet... ;-)

>
> This is very nice work.
>
> JimT
>
>
>


-- 
Thomas

Post a reply to this message