Author Topic: Water fall test (9 sec)  (Read 2671 times)

Offline cyphyr

  • Member
  • *
  • Posts: 3583
    • richardfraser
Water fall test (9 sec)
« on: May 25, 2012, 11:23:04 AM »
I could do with some feedback on this, I've become monitor blind as I've been looking at this for the past week!

waterFall test

Comments very welcome, what do you actually see here.
I'm not too concerned about the water spray at the foot of the falls, I'll work on that later.

thanks in advance

Richard
www.richardfraser.co.uk
/|\

i7 5930 @3.5Ghz, 32Gb
i7 2600k @3.4Ghz, 16Gb
i7 980 @3.33Ghz, 16Gb

Offline Kadri

  • Member
  • *
  • Posts: 7858
Re: Water fall test (9 sec)
« Reply #1 on: May 25, 2012, 12:09:22 PM »

You could kinda get it to work in TG2 but the nuke example on Facebook looks very nice.
I would try to use it. If you want to try if it can be done in TG2 is another thing of course  :)

Offline j meyer

  • Member
  • *
  • Posts: 2022
    • orkwarts
Re: Water fall test (9 sec)
« Reply #2 on: May 25, 2012, 02:43:03 PM »
There are gaps in the streaks of the fall.And these falling gaps spoil the
illusion,I'd say,at least for me.
Keep going,J.

Offline TheBadger

  • Member
  • *
  • Posts: 7416
  • what's the difference?
Re: Water fall test (9 sec)
« Reply #3 on: May 25, 2012, 11:40:29 PM »
It is very close!

I think you should add a stream like the spray you have at the bottom. That is, make a cloud flow down with the water fall, it will help to break up the 'chunkynes" of the falls. Blend it in as good as you can.

This is great so far! You can make this work!
It has been eaten.

Offline Dune

  • Member
  • *
  • Posts: 12804
  • Corkscrew Bird
    • www.ulco-art.nl
Re: Water fall test (9 sec)
« Reply #4 on: May 26, 2012, 07:37:46 AM »
I think you need to slow down the upward movement of the spray, or just animate the warp, and you need to fill the fractal spaces in the fall, or add another fall with a different seed in the same space. But it looks good.
From this distance you won't have the problem of hard shadows at the sides  ;)
In case you still haven't seen enough of my work: www.ulco-art.nl

Offline cyphyr

  • Member
  • *
  • Posts: 3583
    • richardfraser
Re: Water fall test (9 sec)
« Reply #5 on: May 26, 2012, 09:03:55 AM »
Thanks for the feedback guys.

Interested that several have picked up on the spaces in the fractal falls. They were put in quite deliberately since the falls I was using as reference did have similar features (Angel Falls).

I shared this on FB as well and Matt Bell posted a link to this. Really great technique and a very effective waterfall. Notice how the water is falling in "chunks" that are in turn breaking up. It was this that I was trying to replicate with the broken fractals. I think it's a particle system rendered in Maya and baked onto planes and re-rendered in Nuke with an alpha.

@Dune: You want more!! :) It's already 10 cloud nodes plus three for the bottom spray! lol. Your right about the Spray at the bottom though :)

Cheers

Richard
www.richardfraser.co.uk
/|\

i7 5930 @3.5Ghz, 32Gb
i7 2600k @3.4Ghz, 16Gb
i7 980 @3.33Ghz, 16Gb

Offline Dune

  • Member
  • *
  • Posts: 12804
  • Corkscrew Bird
    • www.ulco-art.nl
Re: Water fall test (9 sec)
« Reply #6 on: May 27, 2012, 06:16:20 AM »
Clouds? I thought they were foam PF's on a water shader. That's what I did anyway. The positive with clouds is that you probably only need very little 'quality' and will thus render faster than a water shader. Have to try that as well.
I just did some animated lava lumps being thrown up from a crater. They were clouds (particles) as well, with quality 0.05 or so. And they still were (too) soft, actually.
In case you still haven't seen enough of my work: www.ulco-art.nl

Offline cyphyr

  • Member
  • *
  • Posts: 3583
    • richardfraser
Re: Water fall test (9 sec)
« Reply #7 on: May 27, 2012, 07:05:18 PM »
I'm actually finding the reverse, for a still image, your right, low sample rate is fine and even introduces some useful noise.
But for an animated sequence that noise is, er, noisy and needs to be removed so the sampling is set high: quality=2, samples=312, but the cloud depth is 3000m and a radius of 50m (which accounts for the high sample rate).
I'm up to 11 cloud layers for the waterfall, including the foam at the bottom. It's starting to look quite believable I think. Quite fiddly though.
Richard
« Last Edit: May 27, 2012, 08:31:04 PM by cyphyr »
www.richardfraser.co.uk
/|\

i7 5930 @3.5Ghz, 32Gb
i7 2600k @3.4Ghz, 16Gb
i7 980 @3.33Ghz, 16Gb

Offline Dune

  • Member
  • *
  • Posts: 12804
  • Corkscrew Bird
    • www.ulco-art.nl
Re: Water fall test (9 sec)
« Reply #8 on: May 28, 2012, 09:12:20 AM »
I know a nice trick for you, but it's fiddly as well: place your clouds on another planet with its center on the main planet's ground, so that it's cloud layer is perpendicular to the ground on planet 1. Then you only need a cloud layer of a few meters thick, instead of 3000m. I won't give away more, but it works well.
In case you still haven't seen enough of my work: www.ulco-art.nl

Offline jo

  • Planetside Staff
  • *
  • Posts: 1729
Re: Water fall test (9 sec)
« Reply #9 on: May 28, 2012, 11:05:21 PM »
Hi Richard,

I've have seen a few tall waterfalls in real life and they did have similar gaps to yours. I think the difference would be, as you mentioned, that the flow often will break up and/or feather out the edges of the gap as the water continues to fall.

Regards,

Jo

Offline Matt

  • Planetside Staff
  • *
  • Posts: 2850
  • I'm the crazy one
Re: Water fall test (9 sec)
« Reply #10 on: May 29, 2012, 01:47:07 AM »
I know a nice trick for you, but it's fiddly as well: place your clouds on another planet with its center on the main planet's ground, so that it's cloud layer is perpendicular to the ground on planet 1. Then you only need a cloud layer of a few meters thick, instead of 3000m. I won't give away more, but it works well.

Ah, so you put the centre of the planet far away from where you want the waterfall, but on the landscape somewhere (maybe a few kilometres away), so that you see the cloud layer turned on its side? Very clever :)
Just because milk is white doesn't mean that clouds are made of milk.

Offline Dune

  • Member
  • *
  • Posts: 12804
  • Corkscrew Bird
    • www.ulco-art.nl
Re: Water fall test (9 sec)
« Reply #11 on: May 29, 2012, 07:38:37 AM »
 ;D
In case you still haven't seen enough of my work: www.ulco-art.nl

Offline cyphyr

  • Member
  • *
  • Posts: 3583
    • richardfraser
Re: Water fall test (9 sec)
« Reply #12 on: May 29, 2012, 08:37:23 AM »
I will give this a go, thankyou :)
rICHARD
www.richardfraser.co.uk
/|\

i7 5930 @3.5Ghz, 32Gb
i7 2600k @3.4Ghz, 16Gb
i7 980 @3.33Ghz, 16Gb

Offline rcallicotte

  • Member
  • *
  • Posts: 9542
Re: Water fall test (9 sec)
« Reply #13 on: May 29, 2012, 11:21:37 PM »
More blur... (latecomer to the party)
So this is Disney World.  Can we live here?

Offline Simius Strabus

  • Member
  • *
  • Posts: 262
    • Terragen 2 Renders
Re: Water fall test (9 sec)
« Reply #14 on: May 30, 2012, 12:54:48 PM »
Looks great!
Why stop dreaming when you wake up?

iMac i7 2.8GHz 8Gb