View Full Version : ocean surface
olivier MICHON
01-14-2008, 05:45 AM
I am working on an ocean surface, and most of my search results lead to a tutorial at www.simplylightwave.com, but unfortunately I got an error when trying to connect to the website.
All other tutorials I have found are really old now.
Any other idea or more recent development on this topic ?
Could nodes help to put some foam regarding the angle between poyls at crests, for example ?
Thanks for any help or suggestion.
serge
01-14-2008, 08:10 AM
I'd also very much like to find a solution. Because of the different sizes (heights) of the ocean waves a simple Y-gradient won't work. (Big waves have foam at the crest, but so do small waves.)
Okay, just brainstorming here: The definition of a crest of an oceanwave is that it's the highest point within a certain area, or, that the nearby surrounding area is lower. Hmmm, Dponts "Nearest Point node" comes to mind, but I'm not sure if I'm on the right track here (I'm pretty bad with math and nodes and stuff). If we would be able to not only evaluate the nearest point but a certain number of nearest points, let's say the 20 nearest points, then we could say that if those 20 nearest points are a certain amount lower (Y distance) than our point, then our point is a crest.
But then, the crest of a wave isn't just a peak (like a mountain), it's a stretched peak... I'm gonna need some more coffee...
Did all this make any sense? :) :D
Like I said, I'm just brainstorming here. Hopefully somebody can come up with something.
Maybe this would be a good chance to try out William Vaughans tutorial about distance controlled localized morphing. In his video he makes something that looks similar to an ocean wave, I'm sure it could be adapted.
Here's the link to his video:
ftp://ftp.newtek.com/multimedia/movies/w3dw/NormalD.mov
serge
01-14-2008, 08:20 AM
Etch, yes, for breaking waves something like this could work. Or Dpont's mdd pointer node with which we can layer the displacements might work better. (Or a combination of the two.)
But Olivier's question is specifically about creating foam at crests. So, once the crest or breaking wave is created how do you generate foam at that point (preferably without the manual placement of emitters :))?
Lightwolf
01-14-2008, 08:30 AM
Depending on how you approach it, here is something I'd try:
Use the same procedural you used to create the displacement, but increase the contrast. This can be layered onto the surface as a white layer to show where the foam/splashes will start.
You could also use the same procedural to control where an emitter will emit particles (use it as a texture, and make the actual ocean surface, or a low-res proxy mesh, emit).
As long as the main parameters for the procedural are the same, you can change the contrast and still have the same pattern (i.e. some of the included procedurals have an offset parameter).
Cheers,
Mike
Oh sorry totally misunderstood the question.
The only thing that comes to mind is using hypervoxels tied to the geometry of the waves, that only become visible once they pass a certain point. If it works like it does in my mind then it would kind of be a reverse to what you were saying (the whole certain number of points closest to point A). You could cheat a little bit with the emitter and only allow the desired amount of particles to be generated at the point of null.
Would that work? From my very limited experience in layout that's pretty much all I can think to contribute, sorry I can't be of more help.
olivier MICHON
01-14-2008, 08:31 AM
I think the localized morphing idea is suitable for algae or any other subaquatic vegetation, but not for an ocean where waves are not at a predefined location. I can't imagine how to define such surface with morphs.
I also tried to find out some nodes combinations from Denis that could help (he always has some) but I am not great at that game 8~
Brainstorming in progress....
doh ! sorry, I am bit slow to reply.
olivier MICHON
01-14-2008, 08:36 AM
Depending on how you approach it, here is something I'd try:
Use the same procedural you used to create the displacement, but increase the contrast. This can be layered onto the surface as a white layer to show where the foam/splashes will start.
You could also use the same procedural to control where an emitter will emit particles (use it as a texture, and make the actual ocean surface, or a low-res proxy mesh, emit).
As long as the main parameters for the procedural are the same, you can change the contrast and still have the same pattern (i.e. some of the included procedurals have an offset parameter).
Cheers,
Mike
Good suggestion, Mike !
I have to try this way. Thanks !
serge
01-14-2008, 08:58 AM
Depending on how you approach it, here is something I'd try:
Use the same procedural you used to create the displacement, but increase the contrast. This can be layered onto the surface as a white layer to show where the foam/splashes will start.
You could also use the same procedural to control where an emitter will emit particles (use it as a texture, and make the actual ocean surface, or a low-res proxy mesh, emit).
As long as the main parameters for the procedural are the same, you can change the contrast and still have the same pattern (i.e. some of the included procedurals have an offset parameter).
Cheers,
Mike
Mike, could you please explain? Wouldn't this work more or less the same as a simple Y-gradient? It would create foam at the high places of the mesh, but also the flat high places, no? You only want foam at all the crest parts of the mesh (whether they are at a high place or low place). (A gradient based on the slope also won't work because that will create foam at the valleys as well as the peaks of the mesh.)
olivier MICHON
01-14-2008, 09:04 AM
I don't know if it's possible, but if there is a way to get informations about the displacement (nodal stress map ?) it would be easier to select crests, no ?
Lightwolf
01-14-2008, 09:33 AM
Mike, could you please explain? Wouldn't this work more or less the same as a simple Y-gradient? It would create foam at the high places of the mesh, but also the flat high places, no? You only want foam at all the crest parts of the mesh (whether they are at a high place or low place). (A gradient based on the slope also won't work because that will create foam at the valleys as well as the peaks of the mesh.)
More or less, yes.
However, you can still layer/subtract the same procedural with different settings to get an outline for example. This gives you a whole new range of possibilities.
Cheers,
Mike
dpont
01-14-2008, 09:36 AM
I remember a very convincing ocean and foam
mainly concentrated around a submarine in LW 8.0 Content,
may be a slope gradient in layer texture will work
for crest.
Denis.
serge
01-14-2008, 09:55 AM
More or less, yes.
However, you can still layer/subtract the same procedural with different settings to get an outline for example. This gives you a whole new range of possibilities.
Have to try that. Thanks Mike.
I remember a very convincing ocean and foam
mainly concentrated around a submarine in LW 8.0 Content,
may be a slope gradient in layer texture will work
for crest.
Denis.
Hi Denis,
But the problem with slope is that it would generate foam at the valleys as well as the peaks of the mesh.
MUCUS
01-14-2008, 09:56 AM
Have you try to use two planes? one for the water, the other one for foam?
The first and the second will use the same displacement map to simulate the waves but the second will be a very very little bit up the first and got transparency or clip map, so that it will be only visible at the top of the wave...it coud be done with a gradient texture I think
dpont
01-14-2008, 10:14 AM
I remember a very convincing ocean and foam
mainly concentrated around a submarine in LW 8.0 Content...
53782
Denis.
olivier MICHON
01-14-2008, 10:38 AM
Have you try to use two planes? one for the water, the other one for foam?
The first and the second will use the same displacement map to simulate the waves but the second will be a very very little bit up the first and got transparency or clip map, so that it will be only visible at the top of the wave...it coud be done with a gradient texture I think
Interesting suggestion, Mucus, but I am afraid it will have a high cost in polys count (the first layer is already 2 millions polygons) and render time for a resut that will be more or less Mike's idea (higher contrast texture).
olivier MICHON
01-14-2008, 10:42 AM
53782
Denis.
Thanks for the archive image, Denis ;)
But it looks a bit "old fashioned" regarding what we could expect for this kind of effect today.
serge
01-14-2008, 11:05 AM
Interesting suggestion, Mucus,
I agree, it's very interesting. (Let's just forget about your polycount for awhile, Olivier :p ) If I may brainstorm further on this: what if you would make the diplacement map of the second plane less contrasty (I think this is more or less what Mike was hinting about in his second post), then the peaks of the first plane would penetrate the second plane, wouldn't they? This should give us something to work with I think. Thanks Mucus! :thumbsup:
serge
01-14-2008, 11:08 AM
Denis, I do agree it's pretty good looking, I've checked the scene in LW8. But you're not going to get away with closeup shots of the waves with this approach.
dpont
01-14-2008, 11:21 AM
Except for foam near the submarine,
the approach is the same than Mike's suggestion,
a Dented procedural Displacement is also applied with same
Scale & Position in 3 Texture Color Layers
to contrast it as a mask for foam.
Denis.
munky
01-14-2008, 01:47 PM
Hi there,
Here is my sea, a mix from the 3d world and the simply lightwave tuts, not fantastic but I hope it might be of some use.
regards
Paul
vbk!!!
01-14-2008, 06:46 PM
the 2 planes trick works :
http://spinquad.com/forums/showthread.php?t=10661&highlight=foam
hope it helps
olivier MICHON
01-15-2008, 01:25 AM
the 2 planes trick works :
http://spinquad.com/forums/showthread.php?t=10661&highlight=foam
hope it helps
Thanks VBK, I admit it seems to work great :thumbsup:
But the thread is, one more time, refers to the simply lightwave tutorial, and as I said in my first post I CAN'T download it :(
Would someone be kind enough to send me the tut files ?
Thanks.
olivier MICHON
01-15-2008, 01:33 AM
the 2 planes trick works :
http://spinquad.com/forums/showthread.php?t=10661&highlight=foam
hope it helps
BTW, did you never have found the time to explain more about your technique ?
olivier MICHON
01-16-2008, 06:28 AM
Hi there,
Here is my sea, a mix from the 3d world and the simply lightwave tuts, not fantastic but I hope it might be of some use.
regards
Paul
Thanks for sharing, Paul :thumbsup:
That's a good starting point for what I want.
mrpapabeis
01-19-2008, 11:13 AM
53782
Denis.
What's the file called?
TIA,
Gp
serge
01-19-2008, 11:37 AM
What's the file called?
TIA,
Gp
typ21
RonGC
01-19-2008, 12:58 PM
Hi there,
Here is my sea, a mix from the 3d world and the simply lightwave tuts, not fantastic but I hope it might be of some use.
regards
Paul
The free Ocean tutorial video on the Simply Lightwave site is very informative, but be warned it is big and takes a while to download.
3d world issue 98 has a good article on creating waves and LW scene files
Ron
Elmar Moelzer
01-19-2008, 01:50 PM
I did that ocean for the LightWave 8.0 content. Its a little old fashioned by now, thats true. You might want to have the ocean become an emitter that emitts particles controlled by a texturemap that you can use the procedural for that was used for the white foam caps.
I havent really tried that yet, but you should get some convincing effects.
CU
Elmar
Glendalough
01-19-2008, 03:49 PM
the 2 planes trick works :
http://spinquad.com/forums/showthread.php?t=10661&highlight=foam
hope it helps
As olivier MICHON says here, you have never told us HOW YOU DID IT??
It is now 2 years ago!
MUCUS
01-19-2008, 05:11 PM
Here's a recent work posted on CGSociety for those who don't see it yet:
animation: http://www.maxon.net/pages/dyn_files/dyn_htx/htx/2238/02238_02239_6_l.mov
thread:
http://forums.cgsociety.org/showthread.php?threadid=584305&utm_medium=plugblock&utm_source=cgtalk
work is done with C4D but the making of ocean is adaptable on LW I think :) (compositing work helps a lot realism!)
olivier MICHON
01-20-2008, 04:34 AM
As olivier MICHON says here, you have never told us HOW YOU DID IT??
It is now 2 years ago!
Maybe VBK has changed his mind about it. I also miss it since it's one of the more convincing foam effect I've seen so far done with LW.
Moreover It's quite close to the effect I need in my current project.
I bet it will be done with Maya fluid at the end...once more time.
olivier MICHON
01-20-2008, 04:41 AM
I did that ocean for the LightWave 8.0 content. Its a little old fashioned by now, thats true. You might want to have the ocean become an emitter that emitts particles controlled by a texturemap that you can use the procedural for that was used for the white foam caps.
I havent really tried that yet, but you should get some convincing effects.
CU
Elmar
Elmar, I Hope I haven't harmed you with my comment about this scene.
But we all know that 3D is looking outdated so quickly ;)
Thanks anyway for providing us this useful content.
olivier MICHON
01-20-2008, 04:47 AM
Here's a recent work posted on CGSociety for those who don't see it yet:
animation: http://www.maxon.net/pages/dyn_files/dyn_htx/htx/2238/02238_02239_6_l.mov
thread:
http://forums.cgsociety.org/showthread.php?threadid=584305&utm_medium=plugblock&utm_source=cgtalk
work is done with C4D but the making of ocean is adaptable on LW I think :) (compositing work helps a lot realism!)
Interesting link, Mucus. Thanks.
Not sure it could be adapted to LW so easily since they are talking about some specific dev for the waves.
Anyway that's a good reference to work with.
MUCUS
01-20-2008, 06:44 AM
Yes, the work will be to find the corresponding procedurals in Lightwave, the good movements for the nulls that will control displacement and bump map....but for the foam, I'm asking myself if a tillable bitmap with a displacement layer under it cannot do the thing?
Anyways, hope you will be able to do the production with LW!:)
olivier MICHON
01-20-2008, 07:01 AM
Yes, the work will be to find the corresponding procedurals in Lightwave, the good movements for the nulls that will control displacement and bump map....
Well, that's the game I am playing at right now.
I think a second layer with a procedural or bitmap displaced according the first layer is the way to go... I just hope to have enough time for experimentations.
Thanks again for the ideas.
Glendalough
01-20-2008, 07:19 AM
Maybe VBK has changed his mind about it. I also miss it since it's one of the more convincing foam effect I've seen so far done with LW.
Moreover It's quite close to the effect I need in my current project.
I bet it will be done with Maya fluid at the end...once more time.
Well I'm glad others are interested in this!
I think the 2 plane solution, (something as VBK is doing and not telling us) is the best solution to an open sea (large area). I think the second mesh should be pretty much identical to the main one and it should just be the foam, foam texture for surface.
Think the trick here is going to be to perforate this mesh in some manner (clip map or other ) that resembles the shape of foam on the sea. It should be noted that the foam does not always appear on the crest and also can be seen floating down the sides of waves. Perhaps also the clip map (or weight map??) should be animated.
To reproduce the complexity of the ocean beyond a certain degree would be impossible, we all know, but I think a plausable solution is really about a texture floating and reacting (on the surface) with the main water plane.
That is, the displacement job on these animations is done, finished (though perhaps the motion could be refined or improved slightly through manipulation of the graph editor etc and other parameters). The whitecaps need a texture solution IMHO.
The CD4 clip that MUCUS has linked is great but it is nothing LW couldn't do as far as I can see, looks very similar to the stormy ocean toot.
Don't think Maya fluid or any other fluid will work on this large scale, but for a more local view, as around the direct local, area, of a ship or vessel it could be great. Am attempting something like this myself and will be posting soon.
But there has just got to be an answer to the open ocean with whitecaps. It's there somewhere in LW, just needs to be unlocked!
Elmar Moelzer
01-22-2008, 01:59 PM
Hello Olivier
Dont worry, I know it is quite old now ;)
Yeah CG gets old really quick...
CU
Elmar
Elmar Moelzer
01-22-2008, 02:18 PM
Oh in regards to the double plane thingy...
Back when I did the Typ21 scene(s) I tried something like this:
I made a highly subdivided version of the Ocean plane (noone says you have to subdivide it at rendertime anyway) in Modeler. I killed all the polys and made single point polys with the points to polys tool. Then I applied the same displacement and shading to that plane, but also put in a clip map to only show the foam caps (basically the textures that color the foam caps on the main ocean). This plane was then rendered just as partigons would be.
That worked... almost. However back then (and I dont know about nowadays, since I havent tried it since version 7.5 of LW) the particles would cause some really ugly artefacts when they went further away from the camera (I am not 100% sure anymore now, but I think the fog had something to do with it too). I am not quite sure anymore what else I tried, since I abandoned that idea due to the artefacts, but I think I also applied the foam caps texture to the transparency channel. I could be wrong though. In any case you need to have fade with distance on...
Now, IMHO VBK is either doing that, or this:
Have a second ocean plane with clipmap for the foam. This is then moved very rapidly up and down in the Y axis to add some fake thickness to it and to enhance the foamy effect (needs motionblur). With the new PRM blur it might look even better.
I could also imagine that he might have some other texture applied to it that modifies the foam clipmap depending on the position of the ocean plane object (some kind of gradient, or mabe even a simple envelope that is in sync with the y motion).
I think I tried something like that too, when I experimenting arround. I cant remember anymore why I did not do it. I was either not too pleased with the effect, or it was to rendertime intensive or something like that (our workstations here were actually quite a bit slower back then).
CU
Elmar
olivier MICHON
01-23-2008, 03:34 AM
Thanks for sharing your experimentations, Elmar. That's interesting tracks to explore :)
I was figuring out something like that about VBK's stuff. For sure, the photoreal blur will improve the result a lot.
-EsHrA-
01-23-2008, 04:08 AM
the foam in the planetx fx c4d example is largely real footage of foam composited onto the cg water since it is still hard to do realisticly in 3d.
mlon
vBulletin® v3.8.2, Copyright ©2000-2012, Jelsoft Enterprises Ltd.