Attached Link: http://julien.chaplier.free.fr/Main.html

Hey all, Julien chaplier just posted on the Carrara list about a new plugin he has , looks interesting.Here is his post Hi all I added a new plug-in on my website (http://julien.chaplier.free.fr/Main.html ). Its name is Deeper and it handles normal maps in Carrara. Normal maps create the amazing real time bump mapping displayed in modern video games such as Doom3. Along with some information on normal maps, I posted 2 tutorials to show how they can be used to create realistic billboards or bas-relief style bump mapping. Deeper works with Carrara 3. The Carrara 4 version should be ready shortly after the SDK is available. Upgrade will be free of charge. The Mac version is not available yet, but I'll post it as soon as I can build it. If you've got questions regarding Deeper, feel free to ask me on this list or directly at julien_chaplier@yahoo.com. Regards, Julien

Normal mapping? Yow. I was just reading about that in Game Developer. This could come in handy for larger animations. Apropro of which, has anyone attempted to fit Carrara into their pipeline on a game development? I know it lacks the specialized toolsets avail to, say, Max...

That is very interesting. Thanks Mike & Julien! ArchiTools looks as though it will be a great plugin too.

He has some interesting looking plug ins, would love to see more of the ArchiTools but his site needs some major work as most of the graphics and images don't show up or are missing...

Yeah, I have the same problem when using Mozilla. IE works fine though.

You should try getting though there on iCab! Must be another one of those IE-tailored sites...

I think this plugin does look quite nice. When combined with Anything Groves and/or Zbrush there is some serious power avaliable.

As long as the whole pipeline is there. I mean, you gotta be able to generate, export, then re-apply that normal map to get the savings in polygons. But I guess this DOES do the re-import, because Carrara will render the normals data already (I know, because I rendered out ALL the alpha channels while I was generating a picture and bump for the texture on something else.) If it can be as simple as generate, save, import material, apply, it could be a saver every time a scene gets a little large to handle.

I understand the difference between a normal map and a regular bump map, and I can appreciate why a normal map is faster to render, because the renderer doesn't have to sample the surrounding pixels to determine the correct normal... but couldn't Eovia (or some other kind of plugin) just do some kind of pre-processing on a regular bump map to turn it into a normal map? Then the first time you render an image it would be slow, but after that Carrara could use the cached normal map and render more quickly. Then you wouldn't have to go through the extra step of making a normal map from geometry.

Thanks for reporting the problem with the web site. I don't have much time to spend on it, so I try to keep it as simple as possible to avoid incompatibility like that, but I made the mistake to build the website with Word. It works fine on IE, but not on Netscape. And when I retouched it with Composer, and it works on Netscape but not on IE. I ended up doing it in the text editor. Does somebody knows some a simple html editor I could use instead of these 2, they're driving me crazy ? As for the plug-in, it makes and reads normal maps as they are described on this site (but it's much easier to make them with Deeper since its a renderer that does the job, no special shader or lighting to do): http://members.shaw.ca/jimht03/normal.html but not the one described here: http://amber.rc.arizona.edu/lw/normalmaps.html (I'm currently working on something that could do that)

I've run into similar problems with using Word to create web pages. I finally gave up on Word and got Macromedia Studio. Great plug-in, and I know I'll be buying it soon to use with Carrara. Chris

Yeah, I would use either Adobe Go Live, or Macromedia Dreamweaver... Word, IMO, creates far too much code and is not very cross-platform compliant. If you have a Unix or Linux hosting account, or an account with PHP, look into PHP nuke as it is an entire content management system that you can install, customize and use the built in editors to create articles, galleries, etc... Regards, Paul

As I understand it a normal map gives tighter control than a bump map. A bump map wraps a displaced surface around the original poys. A normal map takes every little texel of surface and treats it invidually. As I understand it, though, you can't paint a normal map. Perhaps as tools get more advanced there will be ways of adding detail automatically -- perhaps starting with something as simple as a Perlin noise generator, or perhaps a way to fold in an ordinary height map. In Doom3 they built real geometry to generate the normal map, then applied that to the low-poly runtime models. I could see this as a help even for those of us doing stills; it would let you model lavishly, even wastefully, then just before you move to the next object in the scene pull normals and swap out to a low-poly model.

" As I understand it a normal map gives tighter control than a bump map. A bump map wraps a displaced surface around the original poys." This might help: "Normal mapping is a method where by the normals of a mesh are given more detailed lighting information by means of a bitmap image. This give the mesh the appearance of more realistic lighting and more surface detail. Normal maps are encoded to a bitmap image, much like a greyscale bump map. The way in which they differ however is that in traditional bump mapping only distinct changes in elevation are registered. A gradual change in height is almost impossible to perceive. Normal maps not only describe height, but accurately depict the angle of faces and are able to describe curves and complex angles much more accurately. In short, they create a much better illusion of height. Normal maps are encoded in RGB values which are used to indicate which direction the normal faces relative to the face of the polygon it is mapped to. The red channel describes normal facing in the X direction. 100% red means it is facing right 0% red means it is facing left. A 50% value means it is facing straight out in the direction of the polygon." http://www.ionization.net/tutsnorm1.htm Just to be clear, bump mapping and normal mapping don't actually displace the mesh like, for instance, sub-poly displacement does.

That's a better explanation that the one I gave, maxx. Blame my lack of coffee. And that's it in a nutshell; a bump map can only say "this pixel is supposed to be higher than that pixel" and the software has to generate a slope from one to the other. A normals map can go right down to the texel level to specify not height but exactly which way that texel is pointing -- without having to slop an imposed slope over its neighbors. Say I had a bump map that was a tight checkerboard; one pixel white, one black. The result would be, at best, a kinda lumpy surface. Bump mapping can't derive a hundred little towers. In a normal map, though, I could make one of those diffraction grating patterns in which adjacent rows of pixels had crisp, distinct angles. That would be an interesting effect, by the way -- one of the reasons I look forward to a way of mathematically superimposing outside data on a normal map.

It wouldn't be too difficult to create simple normal maps, that will tile, within Carrara. As nomuse points out, the output is already there in the G-Buffers. Julien's plugin presumably saves you the trouble of dropping the output into the RGB channels of a blank image in Photoshop - as I did here - and of course the plugin makes the end result usable. More complex stuff would be more of a challenge but not impossible.

yeah,exactly what i need.:)

I just re-read what is said on the site about Deeper. It adds Normal Map as a shader function (in the bump map channel), and since it behaves like an ordinary texture map it can be tiled. It also appears to add some new way of rendering the normal map (Carrara already does this -- not sure what is being added here). Also read up on Architools, which if properly customizable could be a great aid in some projects I've been considering.

Here are a few precisions about what Deeper and Carrara exactly do. Carrara can render normal map but in global space, the one rendered with Deeper are in camera space, the z axis pointing toward the camera. And when the camera is a perspective camera, Deeper modify the orientation of the vectors to compensate the camera deformation. So its possible to create normal maps with Carrara, but the camera need to be an isometric camera, and to be place in the top view. Then a few more operations outside Carrara are necessary to extract the map from the g-buffers and place them into the color channels. Deeper renderer just removes all these constraints. Then Deeper provide a way to read the map in Carrara shaders, which wasnt possible before. I hope this helps. And thanks for your advices for the website construction tools.

Well, I got the plugin. :-)

Here is a quick test. Both images use a plane primitive lit with two lights, a blue on the left and a red on the right. The top image uses a bump map (made in C3 using distance G-Buffer) tiled across the surface, the bottom image a normal map (made in C3 using Deeper) and applied to the bump channel.

No vertices are displaced so the plane edges are still absolutely square for both methods but the advantages of normal maps over bump maps are clear, I think.

Nice!! Can you run one with the full geometry, and one with the face on a plane, and tell us what the render times are like? There should be a significant advantage. . . Markd

Falconproject's example of normal maps versus bump maps is the best comparison I have seen.Quite a clear difference.

Hoo-ya. I knew normal mapping was impressive, but boy this just encapsulates it in one striking image. Good work, Falcon!

Glad you like it. :-)

I think maybe the bump map could be improved. But normal maps still provide a very good way of getting detail without increasing the polys and slowing everything down.

Here's another test. On the left Mike cruelly truncated but rendered as geometry: Full Raytracing;Indirect Lighting: Lighting quality - Good (two distants); Accuracy: 4 pixels.
Render time: 1.02 min.
On the right, the normal map version on a plane, same setup. Render time 4 sec.

As you would expect the color balance in the normal map affects how it renders. This is the same map as above, beefed up a bit using curves in Photoshop. The actual map is below.

Cool. Definately looks like you could render out some heavy surface detail (what do the movie space-ship modellers call it? Cruft?) and then cut-and-paste a bunch on to a new document, and apply that to your less-than-detailed hull. I imagine after a while you could even start to think in 3-color space. Painting the stuff by hand would be very difficult, though! The thing that really gives it away is the lack of cast shadows. And edges, of course -- same as bump maps.

Thanks for the numbers. They are about as striking as I would expect! Markd

Well. When the Mac version is available, pleas let us know! I am sure this will open up new dimensions!

here is a plugin to create normal maps with photoshop,if you need to make some normal maps for this carrara plugin. http://developer.nvidia.com/object/nv_texture_tools.html

Thanks, that's a cool plugin. What it does is convert bump or height maps to normal maps. You can adjust the depth and preview the 3D look, complete with movable light.

no problem.:)

So is Deeper an alternative to "Anything Grooves"?

So is Deeper an alternative to "Anything Grooves"?

So is Deeper an alternative to "Anything Grooves"?

Sorta. Sorta. Sorta. More like an alternative to traditional bump mapping. Trouble with displacement mapping (aka Anything Grooves) is that you are introducing more geometry into the scene, making for longer renders. The cool thing about normal mapping is it does everything bump mapping does, only better, and faster, too. In fact, you can think of normal mapping as being Anything Grooves in reverse; taking geometry and reducing it to a map, leaving you with a simpler model.

Sounds great, but I guess the edges (outline) of the model will remain smooth with Deeper...right? As opposed to Anything grooves which will add detail...right?

Right. Edges remain whatever the actual geometry is. Plus I'm not so sure about the ability of normal-mapped geom to self-shadow.

Nomuse, thanks for clearing this up for me. I think I'm starting to understand and appreciate the difference.

From what I've seen here the normal mapping gives a real feeling of depth which makes the lack of shadowing even more apparent and noticable. It just doesn't look "right". IMHO Ed

Ed, I'd like to stress that the tests I've posted here (and the image I put in the gallery) are not intended to show how to use normal maps; rather I've tried to show how much better they are than bump maps by pushing the detail possible. Normal maps will never be as good as actual geometry where only actual geometry will do. What they can do is add detail when you need to keep down the poly count. This probably applies more to animation than stills but even in still renders they could be useful, particularly in fine detailing where the lack of shadows is less noticeable.

Message edited on: 10/07/2004 04:15

flaconperigot, You're right of course. I tend to look at everything through my "still life" glasses. Ed