PDA

View Full Version : geet me a boat & ah piece o' cloth then I might get started



fabmedia
02-28-2006, 10:24 PM
I know this is starting to become redundant, but my question really isn't the same. I have an animation of a "boat" moving through some "water" and the hull needs to interact with the water with - ta-da - via clothe fx. Now if was to take a base setup like eki's seagen where it makes a phenominal mesh for the ocean (which is good), but i'm thinking that just in calculation time it would be most disapointing. SO, I'm looking for answers on how do I create the effect that I want, keep calculation times down, and make the whole ocean work as one. BTW, I've noticed that when you apply a displacement texture and animate it to an object that you wish to behave like water, it falls to the wayside. Any ideas on that as well?

IF was to do this in stages, what would be the best way to do it? The trouble is that I have a distant shot coming in, then a beauty shot that will last about 10 seconds or so.

toby
03-01-2006, 12:06 AM
Hi Arlen

I'd probably try painting a displacement map for this. I haven't done a wake for a boat but I think you're right that calculating dynamics would take forever. You would render the boats' hull from underneath it, then paint ripples ahead of it on your map.

Where I work, fluid effects are done in Houdini, and when they render the dynamics it can take more than a day to calculate - so actually, you could try it in LW and just let it go for a while :)

fabmedia
03-04-2006, 12:13 PM
Okay, I've done 3 test renders with Cloth Simulation and need some basic feedback on what looks the best or what I need to change.

You need Quicktime 7 as they are encoded H264.

wave test 1 (http://fabmedia.no-ip.com/OceanFX/Movies/seatest.mov)

wave test 2 (http://fabmedia.no-ip.com/OceanFX/Movies/seatest01.mov)

wave test 3 (http://fabmedia.no-ip.com/OceanFX/Movies/seatest02.mov)

1 and 3 may look the same, but they have different depths for the waves. Of course, the simulation will take place over a distance of 500 metres.

And just for fun, the night time scene with the SH-60...
Hover Test (http://fabmedia.no-ip.com/OceanFX/Movies/hovertest.mov)

jahn
03-04-2006, 02:44 PM
it will be great if you are able to use naturefx.

these were my renders i did 2 years ago. i wish it might help.

http://www.antrepo.org/temp/07.jpg

fabmedia
03-04-2006, 09:05 PM
What the heck is "naturefx"? I'm on a mac here. Those renders look awesome!

toby
03-04-2006, 09:14 PM
yea that does look great.

So does the cloth fx; How long did your calculations take, Arlen?

fabmedia
03-04-2006, 10:01 PM
yea that does look great.

So does the cloth fx; How long did your calculations take, Arlen?

Thanks. Which one did you like 1,2 or 3? AND did you like the night time shot? The grid is 50 x 50 meters w/ 11,800 polys for the calculations, and 1000 polys for rendering with a x10 object mesh at render time. This render only took 2.3 seconds on my dual 867 g4.

Now for the final I have over 183,500 polys to calculate (100 m x 500 m) and a base mesh of 7000 polys to render with a x20 object mesh. HOWEVER, i have to also calculate a huge ocean that uses the same deformations as the calculated mesh, then blend the 7000 polymesh with a transparency setting to composite that in Combustion.

This is all for a competition at another site. I also have to figure out how to implement an explosion into the works. The LCS is going to be bombed, and the helicopter also has to affect the water too. Actually 2 helicopters. But the thing is I'm not calculating the interaction of the boat with the mesh. I'm using a b/w normal displacement to move the mesh for calculations. It's been a b*tch to figure out which is the best method to use, but with the boat mesh interacting with the base mesh, it was taking too long to calculate. NOW, I have to find a way to get particles to "splash" over the bow, and get develop a good texture for the water. BUT the best thing is that I'm going to make it a cloudy day with fog & mist. I'm going to use the lighting render cheat with a PLD 7 pass w/gaussian AA with a dithered motion blur (with vector blur, the normal blur option) @ HD 720p. This will work for a good demo reel peice (hopefully) & maybe get me a job with a nice company. I figure that I can render out the helicopter wings @ pld 9 as a separate pass (w/vector blur) to increase their smoothness. Of course there will be separate passes for the sky, each helicopter, the LCS, missles, gun fire, and the rain/ mist, while using Combustion to colour correct, gunfire, missle smoke, explosion, heat wave at the back of the seahawks where the exhaust is, as well as DoF blur where necessary. NOW if you think that's alot, I'm going to create a mattle painting of mountains if I have time.

Whew! I have to get this all done within the next 20 days. The rendering is what is going to do me in. I might have to render out only 5-6 seconds rather than the 15-20 seconds that I want.

toby
03-05-2006, 03:35 AM
I like the second one the best.

The night shot is pretty good, but I have great idea to help the rotors' motion blur - render the rotor blades all by themselves and comp them back in. Then you can render as many passes as you need to get a smooth blur, and it'll render really fast.

This only took 5 seconds to render at enhanced extreme with dithered motion blur - that's 66 motion blur/AA passes!

pooby
03-05-2006, 04:22 AM
http://www.newtek.com/forums/showthread.php?t=29664&highlight=water+wake

this is a water wake I did a while back.. It uses a 3D cloth fx sim..

fabmedia
03-05-2006, 11:55 AM
@ Toby

I think I had mentioned the separate pass in my last post, but that is great!!! I love the look.

@ Pooby

Man, you were my inspiration!!! I had sent you a PM asking whether you had tried your technique with a hight polygon count and such. You know, the final animation is going to involve a 500 m x 100 m stretch of water simulation. I wouldn't think that it would take long to calculate with some 183,500 polys, but then again, I could be "–bleep–" out of luck too. BUT I want you to know, that i had a **** of a time with setting up everything. I have to figure out how to set up calculations based on weight maps. So that when the wake nears the edges (about 10%) it completes the fall off. That will change the number of polys used for the simulation.

After the simulation, I'm going to use deformation to implement the traditional water effect buy affecting the "clothe" with a local setting. That's something I just learned about a couple of days ago.

So I have another test here without the boat… wave test 04 (http://fabmedia.no-ip.com/OceanFX/Movies/seatest04.mov)

It's in a Qt MP4 format. Some people had trouble with the H264 crashing their browsers.

pooby
03-05-2006, 12:38 PM
Oh Sorry.. I'm not very good with keeping up to track of PM's

To be honest, I haven't access to LW at the mo and I did that so long ago that I can't remember offhand what I did.
My studio has split up and I'm starting a new business so I've been a bit distracted lately. I hope you work it all out.

ps.. that last test looks good

fabmedia
03-05-2006, 12:40 PM
Sorry to hear about your studio. It can be really hard to keep things going properly. Thanks for the very positive comment. I'm hoping I can pull this all off before the 24th.

toby
03-05-2006, 01:40 PM
@ Toby

I think I had mentioned the separate pass in my last post, but that is great!!! I love the look.

:D so you did - I must have skimmed too fast

cgbloke2004
03-05-2006, 09:33 PM
(bleh, cant get my img tags to work anymore :o )


this is an old test but i've done a few ships this way.

iirc the sea was mostly crumple and some fractal bumps,
the wake was separate image files rendered separately and then reapplied onto the sea mesh.
add some particles to break it up.
the sea mesh is broken into such a way that only the local area around the ship [and bit that gets deformed the most, via earlier mentioned image maps] has a heavy poly mesh, and it gets lighter the further away from the camera you are.
i've done the same for subs and torps, the generally same idea.

its messy as you can see, but with some sorting out it does sort of work.

you can also add random wakes across the sea by applying another layer that is white and is only affected [and therefore visible] when the sea mesh [the few bumps] intersect with it.
but i've never had a satisfying far off shot with it as it eventually creates an odd pattern.

hope that helps.

in the end, if its too hard or wrong in certain areas, and/or either map on real world footage [i believe they did for the wakes in Titanic], or change the camera work so it avoids the dodgy areas..

fabmedia
03-05-2006, 10:12 PM
Thanks, the renders look good. I'm currently adjusting the b/w deformation map that I'm using for the clothe simulation. Once I have it proper like, I will post another test.

fabmedia
03-06-2006, 02:45 PM
These are the last 2 tests for the bow wake. They are 2 slightly different items, but I think that this is it. Give your opinions.

Sea Test 05 (http://fabmedia.no-ip.com/OceanFX/Movies/seatest05.mov)
Seat Test 06 (http://fabmedia.no-ip.com/OceanFX/Movies/seatest06.mov)

fabmedia
03-06-2006, 02:46 PM
These are the last 2 tests for the bow wake. They are 2 slightly different items, but I think that this is it. Give your opinions.

Sea Test 05 (http://fabmedia.no-ip.com/OceanFX/Movies/seatest05.mov)
Seat Test 06 (http://fabmedia.no-ip.com/OceanFX/Movies/seatest06.mov)

fabmedia
03-06-2006, 05:11 PM
Quick question, does anybody know of some mpg resources of a helicopter flying over water? I want to see how the air and water is affected.

toby
03-06-2006, 10:31 PM
uh, the only ones I can think of are in really bad movies :D Jaws II where the shark comes up out of the water and bites a helicopter, and Rambo II where Rambo comes up out of the water and... bites a helicopter

those tests are looking really good! The boat's not pushing much at the start but the rest looks perfect. It looks really hard, but if you can get a big ripple in the front I'd say you're done. Maybe try a ship that's a little bit blunt at the front? You can always use an invisible hull to get the dynamics you want -

You must be having a blast! I'm jealous - but I may be starting a really cool project too :) There's cool ones where I work, but personal projects are SO much more satisfying.

fabmedia
03-07-2006, 02:37 PM
Thanks! I love the Rambo thing. Did you know that where the 3 movies were filmed is only about 30 minutes from where I live? Funny, I only saw the 1st movie once and recognize all the area where it was filmed. INFACT, there is a small restaurant that has a permanent reservation for Sly anytime he comes back. I can't remember the name of the restaurant though.

I spent some 5 hours last night trying to find stuff. I found out that I have a softbody calculation of 339, 352 polygons. All for 500 metres of mesh. Thank god I'm smart enough to know that I can run the boat a couple of times through the same area without the viewer knowing that it's the same water.

I wonder how long it will take to calculate on the Dual G4 867 MHz system... BTW I'm not getting paid for this... I need to make some money. I'm hoping that someone is watching this and will give me some work... hint, hint.

toby
03-07-2006, 10:15 PM
I'm away from my computer about 20 hours a day from bedtime til after work the next evening, (probably the opposite of you right now!) if you want email it to me I'll give it a go -

firstsingle
03-08-2006, 04:55 PM
Hey Fella's. Excuse me for asking but, what are wakes?

cgbloke2004
03-08-2006, 06:03 PM
Hey Fella's. Excuse me for asking but, what are wakes?
the white stuff on the surface of water left behind by an object moving through it.

1. The path left behind a ship on the surface of the water.
2. The turbulent air left behind a flying aircraft.

http://www.steelnavy.com/images/Simon1200/Figure%208.jpg
http://www.steelnavy.com/images/Simon1200/Figure%208.jpg
http://www.math.fsu.edu/~xxu/test/ship.jpg
http://www.math.fsu.edu/~xxu/test/ship.jpg
http://w3.physics.uiuc.edu/~m-stone5/mma/shipwakes_klein.jpg
http://w3.physics.uiuc.edu/~m-stone5/mma/shipwakes_klein.jpg

firstsingle
03-09-2006, 09:52 AM
the white stuff on the surface of water left behind by an object moving through it.

1. The path left behind a ship on the surface of the water.
2. The turbulent air left behind a flying aircraft.

http://www.steelnavy.com/images/Simon1200/Figure%208.jpg
http://www.steelnavy.com/images/Simon1200/Figure%208.jpg
http://www.math.fsu.edu/~xxu/test/ship.jpg
http://www.math.fsu.edu/~xxu/test/ship.jpg
http://w3.physics.uiuc.edu/~m-stone5/mma/shipwakes_klein.jpg
http://w3.physics.uiuc.edu/~m-stone5/mma/shipwakes_klein.jpg

Thanks cgbloke. And thanks for the links. I haven't started getting into any of lightwaves sims yet. This helps out alot now that i'm ready to take the plunge.