Paq
Is this grayscale image any different?

Producing a normal map with the ,free, Nividia filter plug in for Photoshop is a piece of cake anyway--though I also have a normal plugin in Carrara as well.

In most instances, I have no problems with any of the three file types in the Carrara Displacement tag as far as I can tell. Maybe problems with different apps?

I thought that the Carrara image showed that the screen grab from MoI worked--despite the "theory"?

Brian

PaQ

As a follow up. (Maybe my apps are different)

I can get good 3D displacement either with colour images or b&w or normal maps.

Or NO images at all! --see attached

Brian

Hi Brian,

Just to let you know that you have an overlapping situation in your displacement map (circled) and if you use it over a large area it will show up like dogs you know what :)

Ha! You picked it!

There were two, but the other one was ouside the screen save.
Old blokes, when it's late, take shortcuts rather than go back--OFTEN!

But it was good learning all the same.

Brian

>> I thought that the Carrara image showed that the screen grab from MoI worked--despite the "theory"?

Well at that rendersize, you can use any kind of greyscale pattern and it will look right.
But if you're goal is to learn how create good displace/normal map, the method is wrong.

You can probably come to something close in this example if, when you grab the screen, you use a front light (actually you dont), and you should also hide the
wireframe. But if you do any closup or -mid closup, you'll see that the sphere shape will not look like sphere anymore, but more like some kind of spike.

Anyway you'll never be able to grab any subtile shape into a displacement map (or normal map) with that method, using 8bits value image from any 3d viewport lighting value.

As for the photoshop normal plug, it's a great helper to extract normal information based the grayscale value of an image (there are different method in the plugin), and can work quite well for high frequency details ... but if the actual grayscale is actually wrong, the normal will not look beter.

It's quite different that baking the normal from an highres geometry like x-normal can do.

Personally I would just write a shader program that handled the displacement analytically.... No resolution concerns at all! :)

- Michael

Hi Brian, as images are better than my english

The test object


Different result


The last row show a real displace/normal baking from the 3d geometry (using modo) ... the ugly banding effect in the modo displacement
comes from the 8bits image format.

Hi Michael,

>> Personally I would just write a shader program that handled the displacement analytically.... No resolution concerns at all! :)

Well renderman is really great with all that stuffs, there is no render time difference between a bump or a real displacement map ... it's really super fast.
The wrong side is a super slow raytracing :'(

Btw do you have any plan in mind for rendering solution inside MoI for the future (MoI 8.0 ? :))

Does it possible to apply an displacement image totally different of the model?
like this for example

I think, seeing people want to make it so hard, that I am better off not using an image at all--like these.

Brian

Hi Pilou,

> Does it possible to apply an displacement image totally
> different of the model?
> like this for example

Generally no - that kind of displacement is typically oriented more towards moving the shape along the surface normal to create things like bumps and indentations, not for creating a completely different form.

- Michael

> Well renderman is really great with all that stuffs, there is no
> render time difference between a bump or a real displacement
> map ... it's really super fast.
> The wrong side is a super slow raytracing :'(

Back a while ago some guys had an interesting project where they captured the polygon output from a RenderMan renderer (I think maybe AIR?) to get the kind of baked version of the geometry and then used it for STL / Rapid prototyping.

So it kind of allowed for shader-driven details but with mesh generation. For stuff like Knurling or dimples on shoe soles, etc... I think it was called GeoDump but they don't seem to have the page for it anymore. :(


> Btw do you have any plan in mind for rendering solution
> inside MoI for the future (MoI 8.0 ? :))

No real finalized plans yet, but it will probably start to get some attention in v3 I think.

My current thinking is that I may write a very simple and quick renderer that will be built in to do very basic stuff, but rely more on plug-ins / 3rd party type renderers for more advanced type things.

- Michael

>Back a while ago some guys had an interesting project where they captured the polygon >output from a RenderMan renderer (I think maybe AIR?)

That's correct Michael. AIR engine can do this. Also Rhino can do this (not as advanced as AIR yet). And Modo as well. Both AIR and Rhino can displace Nurbs objects directly and output a displaced mesh...

@ Brian - The idea to use a screenshot as displacement map only works if you can grab an image of the screenbuffer (see attached image) looking from the Top viewport. This is great for making quick alphas from 3D models that you can apply to fake "hard modeling" details in your sculpting app, such as gears, wall sockets, etc... Or you can just make a single knurl feature and repeat it as many times as you need in U/V directions in your render engine. This is just simple pyramid made in Photoshop... A tiny bit of blurring may be necessary to get the best displacement effect...

jonah

Hi jonah, one of the interesting parts about the AIR mechanism is that it allows for execution of renderman-style shader script programs.

That allows for writing completely custom shaders that actually execute a kind of script code to control the output. It's pretty much the ultimate in flexibility, but of course you have to write code to control it.

- Michael

jonah
Thats what I used of the MoI screen shot idea--and the top view.

Hey!, it worked ok for me! (Gosh, I use colour image files regularly for displacement also--no probs!)

I am thinking more now of just going down the Carrara potentials I showed--not using any image at all. Seems far easier, totally simple and low file size. Why not?

Brian

>My current thinking is that I may write a very simple and quick renderer that will be built in to do very basic stuff, but rely more on >plug-ins / 3rd party type renderers for more advanced type things.

That makes a lot of sense to me as well Michael. I've always liked that idea for some reason.

For example, there is an easy sketchup to kerkythea plug in, and Cycas has an inbuilt portal to povray. Though the two systems are a little different.

In both systems, it is possible to set up cameras, lights, materials inside the modeling program. In the first system, the scene is opened in KT and the rendering operation is controlled from there.

In the Cycas system you don't have to talk to povray at all. A very workable range of render settings can be selected from within the modeler. Which may be very good for some users who don't have time to get into the deep subleties of rendering.

Just some thoughts.