I'm writing a macro to create animated oceans, so I post this very first
animation (I know the texture is not good for ocean, but very good to debug
and anyway the macro will not provide the texture!)

Comments, advices and criticisms warmly welcome

Philippe

PS : run it in a loop

Post a reply to this message

Attachments:
Download 'wave.m1v.mpg' (620 KB)

Wew, this looks very promising.
After that sorry "FIRST POVRAY OCEAN" business I was planning to have a try
myself, since the maths may be less complicated than I thought (see
http://www.imagewks.com/imageworks/docs/research/research_nate2.html).
I was wondering what was the best way to proceed :
- creation of textures to be used as height fields
- mesh generation
- isosurface (with a parameteric definition of the waves) in the Superpatch.
I tried the last two approaches and I went as far as creating a wave macro that
makes a simple wavy surface mesh but I stopped there. The problems I have to
solve are 1) the mesh /isosurface size and parsing and 2) how to introduce
randomness in the waves. If you want the source it's available.

Gille Tran

Ph Gibone wrote:

> I'm writing a macro to create animated oceans, so I post this very first
> animation (I know the texture is not good for ocean, but very good to debug
> and anyway the macro will not provide the texture!)
>
> Comments, advices and criticisms warmly welcome
>
> Philippe
>
> PS : run it in a loop
>
>                Name: wave.m1v
>    wave.m1v    Type: Vmpegwin File (video/mpeg)
>            Encoding: x-uuencode

Post a reply to this message

>I was wondering what was the best way to proceed :
>- creation of textures to be used as height fields
>- mesh generation
>- isosurface (with a parameteric definition of the waves) in the
Superpatch.
>I tried the last two approaches and I went as far as creating a wave macro
that
>makes a simple wavy surface mesh but I stopped there. The problems I have
to
>solve are 1) the mesh /isosurface size and parsing and 2) how to introduce
>randomness in the waves. If you want the source it's available.


The solution I'm currently investigating is :
1) I start with a still image (multifractal or fractionnal Brownian motion)
and then
2) I add 2 sine waves of different Ampitude, Frequency and Direction and
then
3) I add a Three (x, y (or z) and time) dimensionnal correlated noise (don't
know yet what will be the final form)

The sample I've posted is made of Part 2 only (the smoothest ocean possible
!),
The code for part one is written and working (!!!?)
The final idea for part three is in my head I just have to squeezzzzzz my
brains to take it out !

Of course I'm interested in your Code, I hope I will come with an idea for
your point 2, because it is my current concern

Philippe

Gilles : J'adore ton site (J'aime aussi Dali, Magritte et Tanguy, ce n'est
donc pas une surprise)

Post a reply to this message

Gilles Tran wrote:
> 
> I was wondering what was the best way to proceed :
> - creation of textures to be used as height fields
> - mesh generation

The problem with these two techniques is that they are both finite.
With the mesh, you could potentially use a higher resolution mesh near
the camera that fades out to a very low resolution towards the horizon.

> - isosurface (with a parameteric definition of the waves) in the Superpatch.

I think you could probably make an infinite (or at lease very big) ocean
surface which doesn't repeat using this technique.

> I tried the last two approaches and I went as far as creating a wave macro that
> makes a simple wavy surface mesh but I stopped there. The problems I have to
> solve are 1) the mesh /isosurface size and parsing and 2) how to introduce
> randomness in the waves. If you want the source it's available.

I'd love to see it.  :-)

-Nathan

Post a reply to this message

Ph Gibone wrote:
> 
> >I was wondering what was the best way to proceed :
> >- creation of textures to be used as height fields
> >- mesh generation
> >- isosurface (with a parameteric definition of the waves) in the
> Superpatch.
> >I tried the last two approaches and I went as far as creating a wave macro
> that
> >makes a simple wavy surface mesh but I stopped there. The problems I have
> to
> >solve are 1) the mesh /isosurface size and parsing and 2) how to introduce
> >randomness in the waves. If you want the source it's available.
> 
> The solution I'm currently investigating is :
> 1) I start with a still image (multifractal or fractionnal Brownian motion)
> and then
> 2) I add 2 sine waves of different Ampitude, Frequency and Direction and
> then
> 3) I add a Three (x, y (or z) and time) dimensionnal correlated noise (don't
> know yet what will be the final form)
> 
> The sample I've posted is made of Part 2 only (the smoothest ocean possible
> !),
> The code for part one is written and working (!!!?)
> The final idea for part three is in my head I just have to squeezzzzzz my
> brains to take it out !
> 
> Of course I'm interested in your Code, I hope I will come with an idea for
> your point 2, because it is my current concern
> 
> Philippe
> 
> Gilles : J'adore ton site (J'aime aussi Dali, Magritte et Tanguy, ce n'est
> donc pas une surprise)

I don't know how much this differs from the approaches mentioned here, but I've
also seen a sea/ocean-surface with heightfields generated with an algorithm
similar to the famous fire-algorithm you'll find all over the demo-scene. The
effect is actually quite nice.

While I'm at it: it's not really a topic in this thread but the idea of making
the most realistic ocean is a bit pointless because the ocean (sea) will look
different all over the world. If you compare the Mediterranean to the Northsea
for instance, one is clear-blue the other muddy-greyish (which apparently means
the water is cleaner...)

Regards,

Remco

PS if you like I can dig up an example of the animated seasurface I mentioned.

Post a reply to this message

>While I'm at it: it's not really a topic in this thread but the idea of
making
>the most realistic ocean is a bit pointless because the ocean (sea) will
look
>different all over the world. If you compare the Mediterranean to the
Northsea
>for instance, one is clear-blue the other muddy-greyish (which apparently
means
>the water is cleaner...)


I'm interested in the movement, not the "material", as you've seen on my
sample, and, anyway I'm not running after photo-realism but after
credible-realism where the ocean can only be part of a scene.
>
>PS if you like I can dig up an example of the animated seasurface I
mentioned.

I'd love too, thanks a lot

Philippe

Post a reply to this message

Ph Gibone wrote:
> 
> >While I'm at it: it's not really a topic in this thread but the idea of
> making
> >the most realistic ocean is a bit pointless because the ocean (sea) will
> look
> >different all over the world. If you compare the Mediterranean to the
> Northsea
> >for instance, one is clear-blue the other muddy-greyish (which apparently
> means
> >the water is cleaner...)
> 
> I'm interested in the movement, not the "material", as you've seen on my
> sample, and, anyway I'm not running after photo-realism but after
> credible-realism where the ocean can only be part of a scene.

That's why I said this was "off-thread"...
> >
> >PS if you like I can dig up an example of the animated seasurface I
> mentioned.
> 
> I'd love too, thanks a lot
> 
> Philippe

Here it is...
I'm sorry but I had to ZIP the file.
I had to shrink it to a miniature and even then I was afraid it'd be too large.
You can view it at double the size, it won't matter much. It's also supposed to
be a loop.

Regards,

Remco

Post a reply to this message

Attachments:
Download 'zee.zip' (612 KB)