Underwater Effects

[nickej]

This is a frame of an in-progress underwater scene I'm working on. It's more of a proof-of-concept at this stage, to test out whether the pass elements I was rendering would composite well. To save time on the test, I didn't do any depth-of-field or antialiasing , so the texture's a little pronounced. I'm interested in getting any feedback from people who've done underwater stuff about compositing, texture motions...anything that can help "sell" the look. I've been watching all the "Blue Planet" DVDs for inspiration and have watched my own clip so many times I'm sea-sick. I just need a couple new pairs of eyes....
http://www.nickjainschigg.org/Images/Trilobite/Comp%201_3.jpg

Or, if you'd like to check out the motion (about 2MB, QT Sorenson3):Animated Underwater (http://www.nickjainschigg.org/Images/Trilobite/Water_Comp_1.html)

[prospector]

From looking at the video,
the smaller bumps look like upside down smoke.
I think they should be resized MUCH larger,min 50X and less strenth.

[nickej]

I see what you mean about the texture looking like smoke....I used the "Crumple" texture inversed to get the "wavelets", and it's just too sharp. When I enlarge it, though, it looks really awful, even weakened. I've spent the last few days trying different noises, bu without much success. The ones that look watery look like still water, ie just ripply. The ones like Crumple, that have more definition, are just too sharp. So far, the best has been Hybrid Multi-Fractal, with one horizontal dimension considerably larger than the other....

[prospector]

If you look at water closely then you'll see that even on the roughest days the water disturbence is rythmic, that is to say, there is always a pattern to the waves. They are formed BY something, weather a rock (circular) or a moving object (boat waves) or wind (waves in lined up fasion), waves never just form out of nothing, and even if boat waves hit wind wave at different directions, there is still a pattern to be seen. So you may have large wind waves, and the wind is blowing whitecaps so that the little droplets of water are making small circular waves, but the pattern is there.

Decide what is CAUSING your waves and go from there.
.02 cents worth

[Murphy]

a couple of years ago, me and another guy I work with did a tv ad with lots of underwater stuff. it looks allot better (depending on the depth of water you're trying to make) if get some green reflected from the sandy yellow ocean floor.

I can post some pics of the one we did if you want.

murphy ( ^.^)/~~~

[nickej]

Thanks for all the input--I've done my best to take advantage of it where I can...and any posts of reference (especially with explanations!) are greatly appreciated.
The scene is about one meter below the surface of a shallow sea (approximately 60-80 meters) during the Ordovician period, 450 million years ago. The local weather is calm, with a scattering of light clouds. The climate and latitude was tropical, and the world's landmasses seem to have been largely gathered into an interrupted ring around the equator. I've thought of this as producing a large periodic wave pattern, sort of like the "roaring forties" nowadays, except lower, because the sea was shallower, and of course tropical. The setting is many hundreds of kilometers offshore. As a tropical sea, I've imagined it as more transparent than current northern seas, because of the relatively lower oxygen levels in warm water. The particles are imagined as small and sharply defined because they were quite possibly mostly the small crustacea whose skeletons made up the White Cliffs of Dover, rather than membranous or gelatinous things. There was evidently a great need to bind carbon into shells at the time (calcium carbonate). As you can see, I have been trying to use the available data to rough in the look, but I'm limited in that I have a great deal of difficulty in imagining what sort of circumstances lead to sharper or softer chop pattern, and whether the larger wave patterns ought to be related to the smaller in any direct proportion....
As usual, any criticism probably acted upon, and thanks again!
http://www.nickjainschigg.org/Images/Trilobite/Waves_n_Particles.jpg


And for the 3.5MB revised movie, click here. (http://www.nickjainschigg.org/Images/Trilobite/Waves%26Particles.html)

[prospector]

MUCH MUCH better!!! :)

[Rory_L]

That`s gorgeous! Would you be a saint and post the scene? The entire LW community would thank you.

I notice that the particles themselves are showing up and not just the hypervoxel shading, giving the plankton a black spot in the middle. Was this intentional? When I made such a scene I hid the particles to avoid the black spots, but then I wanted fuzzier plankton than you do...

Cheers,

R

[nickej]

I don't know about everybody being grateful, but I have no problem posting a basic file, since I took a lot of the main idea from Peter D. Hunt's article "Underwater Exploits" from Keyframe issue # 30.
You can grab the file here (http://www.nickjainschigg.org/WaterScene.zip)

I have to warn you, though, that what you're seeing above is a composite from After Effects. I couldn't integrate the sky and the fog satisfactorily, so I used a displacement channel, as well as a transparency channel to create the mix. The particles are hypervoxels applied to a box of points created by "Make Random Points" or whatever it's called. I used about 10,000 points, although few of them showed up on the screen, I wanted to be able to add or subtract by expansion/contraction of the point field. They were also rendered separately with an Alpha and composited in. I used "surface" setting, about 5 mm size, with edge transparency activated, I also had a pre-rendered series of images with a caustic pattern projected onto them from a spotlight, to give a little sparkle, and that, I think, is what caused the black spots. Additionally, they were assigned another texture displacement. The basic process is explained at my site detailing the project--a pelagic trilobite (http://www.nickjainschigg.org/Trilobite3.html)
Finally, keep in mind that depending on the depth and effect you're going for, water takes a lot of tweaking to get it to "feel" right for depth, light, haze, etc. What worked at this scale probably won't work right out of the box for any other scenes. That's definitely something I've learned from the practice and the other posts on this thread...Thanks guys!

[Rory_L]

Thankyou very much. I`m sure it`ll be useful!

Comping the clouds in After Effects explains something about them that had me wondering. They did seem to be a little static, when perhaps they would really have been swirled around more by the swell and refraction.

Thanks again,

Cheers,

R

[nickej]

Thankyou very much. I`m sure it`ll be useful!

Comping the clouds in After Effects explains something about them that had me wondering. They did seem to be a little static, when perhaps they would really have been swirled around more by the swell and refraction.

Thanks again,

Cheers,

R

I agree about the clouds. The whole thing is more an exercise in seeing if it can be done before putting in the hours and hours (and hours) to get it "just so" on a large-scale render. What I'm probably going to do is change the intensity of the bump-maps so that they carry more into the smooth areas, and use that output to distort the cloud layer. I'm also going to have a few trilobites swimming around, so hopefully the clouds won't be hogging all the attention, too, although my aim is tha t every element should stand on its own.

[prospector]

A try for your clouds and fog.

What if you mapped the clouds or cloud animation to a plate that is above the water?
And what if you gave it the same bump as the water?

Well you'd get the clouds moving exactly as water doing a 'fake' refraction and not taking a hit in rendering time.
and you could offset the frames for the clouds by 1 frame for an even more realistic refraction motion.
No compositing and now you could get the fog to work better.

Just a thought.

Oh yea Rhode Island :D
From Appanaug..where the bad boys are from :D