Planetside Software Forums

General => Terragen Animation => Topic started by: Tangled-Universe on January 13, 2021, 06:16:36 pm

Title: Path traced water animation, simple test
Post by: Tangled-Universe on January 13, 2021, 06:16:36 pm
Simple test for a different purpose...Nothing really special, but some may appreciate it perhaps.
Rendered with PT AA4 and 25pps. Took about 18 hours.
Title: Re: Path traced water animation, simple test
Post by: gao_jian11 on January 13, 2021, 07:04:16 pm
Yes, I like it very much. Beautiful lighting, clean rendering.
Title: Re: Path traced water animation, simple test
Post by: WAS on January 13, 2021, 08:03:18 pm
Looks very nice! Ive wanted to try rendering my river/stream idea but just takes too long. I have it so the center is a little more turbulent and flows faster, and the outside near shore flow a tad slower sorta deal.
Title: Re: Path traced water animation, simple test
Post by: Hannes on January 14, 2021, 06:11:14 am
Cool! I'd say, the stones should be a lot more glossy and darker near the waterline. They look quite dry although they were just covered with water.
Nevertheless especially in the shade you can see, that PT makes a big difference.
Title: Re: Path traced water animation, simple test
Post by: Tangled-Universe on January 14, 2021, 11:16:46 am
Thanks Gao and others,

Quote from: WAS on January 13, 2021, 08:03:18 pmLooks very nice! Ive wanted to try rendering my river/stream idea but just takes too long. I have it so the center is a little more turbulent and flows faster, and the outside near shore flow a tad slower sorta deal.

I can imagine rendering such setups with varying flow speeds etc. is slow. Otherwise I'd offer to help a hand with rendering. Does that setup really require PT rendering?
For this scene I chose it specifically, because I was just curious to see how PT volume scattering looks with animated water. The scattering is towards normal, btw.

Quote from: Hannes on January 14, 2021, 06:11:14 amCool! I'd say, the stones should be a lot more glossy and darker near the waterline. They look quite dry although they were just covered with water.
Nevertheless especially in the shade you can see, that PT makes a big difference.

Cheers Hannes! You're absolutely right about the lack of glossy reflections and darkness near the shore. It crossed my mind to try it, but the idea of trying to find a solution to masking the stone reflections from the underwater surface quickly made me think I'd rather spend my time on something else :P
As you can read above the primary reason to render this with PT is not the shadow detail, but the volumetric scattering of the water. I think water looks much nicer in PT, pity it's slow.
Title: Re: Path traced water animation, simple test
Post by: WAS on January 14, 2021, 03:54:18 pm
Quote from: Tangled-Universe on January 14, 2021, 11:16:46 amDoes that setup really require PT rendering?
Actually, probably not, as I'd imagine if you'd really want to use it, you'd want to highlight it so you'd want a more reflective angle to see the turbulence and such easier.
Title: Re: Path traced water animation, simple test
Post by: Hannes on January 14, 2021, 04:58:44 pm
Quote from: Tangled-Universe on January 14, 2021, 11:16:46 amreason to render this with PT is not the shadow detail, but the volumetric scattering of the water
I see. I think I never tried it.
Title: Re: Path traced water animation, simple test
Post by: DocCharly65 on January 15, 2021, 02:22:25 am
Like it. Looks already quite convincing :)
Title: Re: Path traced water animation, simple test
Post by: Dune on January 15, 2021, 02:26:01 am
Quote from: Tangled-Universe on January 14, 2021, 11:16:46 amthe idea of trying to find a solution to masking the stone reflections from the underwater surface quickly made me think I'd rather spend my time on something else
It looks wonderful. Regarding the underwater reflections; it must be possible to convert the water altitude in every frame to an altitude mask for the 'planet line' of  nodes.
Title: Re: Path traced water animation, simple test
Post by: Tangled-Universe on January 15, 2021, 01:05:03 pm
I'm not so sure whether if that's possible Ulco. A long time ago I tried, but failed.
Just tried again for 15 minutes or so and I don't see how I can circumvent the inherent obstacle that TG uses one principle 'planet line' of nodes for these kind of things.
Then we didn't have access to conversions to world space, but I didn't get it to work.
(I think it has to do with that TG actually parses the network from bottom up and thus it can't be agnostic for multiple object's coordinates).
Kudos to you if you know how, I'd be happy to learn!
Title: Re: Path traced water animation, simple test
Post by: WAS on January 15, 2021, 09:00:55 pm
Perhaps this is something that Terragen should have an option for for volumetrics. A disable reflection rays for shaders within volume sort of feature.
Title: Re: Path traced water animation, simple test
Post by: WAS on January 15, 2021, 09:57:46 pm
Also you'd think you'd be able to limit alt the same alt as the plane/sphere with a distribution node, and then warp it back up by the water displacement at the same amplitude, but it doesn't align right, and none of the transform settings will align it right that I can find.

Here is an exmaple of the issue. The displacement is inverted to follow the water level, but it seems like some of the secondary mixed noise is not read correctly or is inverted because of using -1 disp (or complement scalar) which then makes the fine details of the water out of sync. Or so it looks. Maybe @Matt  can weigh in on why exactly it can't line up perfectly on the micro disp level. Edit: it aligns much better with PF noise based disp (not inverted), but still not perfect down to fine details of the disp. Follows the larger shapes easy tho.
Title: Re: Path traced water animation, simple test
Post by: Dune on January 16, 2021, 01:39:46 am
I'll have a look at it. It's not really volumetrics, though, but just a thin sheet with a certain displacement. But I do agree a built in 'reflections off' for under 'water' would be cool, but maybe hard to implement.
Title: Re: Path traced water animation, simple test
Post by: WAS on January 16, 2021, 01:47:08 am
Quote from: Dune on January 16, 2021, 01:39:46 amI'll have a look at it. It's not really volumetrics, though, but just a thin sheet with a certain displacement. But I do agree a built in 'reflections off' for under 'water' would be cool, but maybe hard to implement.
Pretty sure PT adds volumetrics to water doesnt it? Approximated for speed, but there, refractions in shadows too if on a object. I dont know how its all setup but i imagine using the shader as a key you could be able to flag and cull any reflection rays from other stuff. Also probably best to have it selectable from a reflection shader itself so if you wanted you could still have reflections like specularless mirror surfaces.
Title: Re: Path traced water animation, simple test
Post by: Dune on January 16, 2021, 02:25:52 am
Water is not like clouds, with particles, but just an infinitely thin sheet, like the planet. IMO the 'depth' is just a calculation, not 'real'.
Title: Re: Path traced water animation, simple test
Post by: Dune on January 16, 2021, 12:23:13 pm
Sorted. I'll post a file in file sharing tomorrow.
Title: Re: Path traced water animation, simple test
Post by: WAS on January 16, 2021, 01:55:42 pm
Quote from: Dune on January 16, 2021, 02:25:52 amWater is not like clouds, with particles, but just an infinitely thin sheet, like the planet. IMO the 'depth' is just a calculation, not 'real'.
Volumetrics are just calculations though, and don't rely on particles or anything. Matt introduced this as proper subsurface scattering which is a type of volumetrics. Only Standard is pseudo-volumetric (says so too). Having a type of refraction in shadows alone says there is volume calculations to give you that calculated effect.
Title: Re: Path traced water animation, simple test
Post by: WAS on January 17, 2021, 12:20:13 am
@Tangled-Universe  here is a project for your study. I got something to work for shore masking after a day of fiddling: https://planetside.co.uk/forums/index.php/topic,28790.new.html
Title: Re: Path traced water animation, simple test
Post by: Dune on January 17, 2021, 02:27:18 am
Maybe my English is not sufficient to truly understand the word/concept of volumetrics. Always believed it was particle related, like the clouds in TG.

Masking solved, great.
Title: Re: Path traced water animation, simple test
Post by: WAS on January 17, 2021, 02:24:35 pm
Quote from: Dune on January 17, 2021, 02:27:18 amMaybe my English is not sufficient to truly understand the word/concept of volumetrics. Always believed it was particle related, like the clouds in TG.


I do think volumetrics fall under different categories in 3D, but in general they're just simulation of volumes. Volumes of glass/crystal. Volumes of clouds. Volume of skin, etc, etc. Anything that has volume where light transmission, light refraction, and other properties take place. Particle simulation is probably the most common, and even TG's water may use something similar by how noisy it is at low samples?
Title: Re: Path traced water animation, simple test
Post by: Tangled-Universe on January 18, 2021, 01:00:40 pm
Thanks WAS for chiming in on the masking, going to check it out!

Clouds can be considered "participating media", as they often refer to it in technical literature and is what I think WAS refers to with 'volumes'.
TG's water object and shader is indeed an infinitely thin plane with an approximation for volumetrics, but with different fidelities available.
A very crude non-volumetric one for standard rendering, a volumetric basic one for scatter towards normal and a fully volumetric one for scatter in all directions.

I think particles are actually not so common, except when simulating smoke/fire/water for animation. For still image work a volumetric shader suffices, because the 3d wireframe or volume shader defines the boundaries. In sumulations those boundaries are calculated from particles. The exception, of course, is when an artist chooses to simulate something for a still image, but then the shader itself is still a volumetric one.
Title: Re: Path traced water animation, simple test
Post by: WAS on January 18, 2021, 02:23:50 pm
Quote from: Tangled-Universe on January 18, 2021, 01:00:40 pmThanks WAS for chiming in on the masking, going to check it out!

Clouds can be considered "participating media", as they often refer to it in technical literature and is what I think WAS refers to with 'volumes'.
TG's water object and shader is indeed an infinitely thin plane with an approximation for volumetrics, but with different fidelities available.
A very crude non-volumetric one for standard rendering, a volumetric basic one for scatter towards normal and a fully volumetric one for scatter in all directions.

I think particles are actually not so common, except when simulating smoke/fire/water for animation. For still image work a volumetric shader suffices, because the 3d wireframe or volume shader defines the boundaries. In sumulations those boundaries are calculated from particles. The exception, of course, is when an artist chooses to simulate something for a still image, but then the shader itself is still a volumetric one.
Yeah, and I think when water has defined volume with intersection of underlying terrain, as you can see the effects changing with the depth of the water with decay and surface refraction. It looks more realistic where SR kinda just looks like XYZ warping of what's visible at surface with a haze kinda effect..
Title: Re: Path traced water animation, simple test
Post by: Dune on January 19, 2021, 02:17:41 am
Thanks, Martin. Clear.
Title: Re: Path traced water animation, simple test
Post by: james adamson on April 30, 2021, 03:30:01 pm
Wow! 
How do you go about getting the water to have milky and then clear areas?