THanks a bunch Andy, will try that in the next few days. Will also
check out the Nvidia solution.
Cheers!
Bernard
On 7/18/06, Andy Jones <andy(at)thefront.com> wrote:
> Or, I made a preset that will hopefully let you rendermap the
> tangentspace normals directly... I just threw this together and haven't
> done extensive testing to make sure it's working right, but it appears
> to give consistent results with the least rigorous of tests -- a noIcon
> bumpmap on a sphere. Basically, it just puts the tangent in world
> space, then does a series of dot products with the tangent, the normal
> and the binormal to derive the tanget space normal map. This will be
> helpful if, say, you have a procedural bumpmap shader. Otherwise, using
> the NVidia tools may be a little easier. Here's the link:
>
> http://rapidsharing.com/download.php?id=AA6F76A0
>
> If you try it, let me know if it works out okay.
>
> -Andy
>
> Ben Barker wrote:
>
> > If you are only using the normal map as a bump and not the difference
> > between two meshes perhaps you can use a utility like this:
> >
> > http://developer.nvidia.com/object/photoshop_dds_plugins.html
> >
> > to just convert the bump map to a normal map. It will probably be faster.
> >
> > On 7/18/06, *Bernard Lebel* <3dbernard(at)gmail.com
> > <mailto:3dbernard(at)gmail.com>> wrote:
> >
> > Hi Andy,
> >
> > We are using same res meshes. The goal is to texture objects with
> > bumpmaps, and convert to normal maps. I find normal maps faster to
> > render and less prone to flicker and such typical bumpmap problems.
> >
> > Our tests were done with objects of polygon count between 840
> > triangles (geoapprox step 1).
> >
> >
> >
> > Thanks
> > Bernard
> >
> >
> > On 7/17/06, Andy Jones <andy(at)thefront.com
> > <mailto:andy(at)thefront.com>> wrote:
> > > I can't think of any reason why it *should* take that
> > long. Computing
> > > the normal should be at least as fast as normal bump-mapped
> > rendering,
> > > and putting it in tangent space should be as fast as a little vector
> > > math and a matrix multiplication. This should be way faster
> > than, say,
> > > rendermapping occlusion, which has always gone at a reasonable
> > speed as
> > > far as I know. I tried ultimapping a sphere to a higher-res
> > sphere. It
> > > seems to go pretty slow. It's certainly slower than
> > rendermapping the
> > > same sphere. With identical geometry, it seems to go somewhat
> > faster.
> > > So, maybe the computation of the corresponding projected surface
> > point
> > > on the high res mesh is what's slowing it down. Could you do a
> > > rendermap directly on the object itself, and render the normal
> > converted
> > > to tangent space, or does that totally defeat the purpose? I'm not
> > > exactly sure what the best way is to do that. There was a D3
> > normalmap
> > > generator shader. Seems like there should be a better way now,
> > like a
> > > built-in tangent-space normal-map generator, but I'm not
> > actually aware
> > > of it...
> > >
> > > -Andy
> > >
> > > Bernard Lebel wrote:
> > >
> > > > Hello,
> > > >
> > > > We are doing some tests with Ultimapper in order to generate
> > tangeant
> > > > normal maps. The goal is to replace bump maps.
> > > >
> > > > Now, when generating normal maps from objects that have bump
> > maps, it
> > > > just takes eras to generate! A few statistics:
> > > >
> > > > ( Render region and pass options' anti-aliasing:
> > > > Sampling: -1/2
> > > > Treshold: 0.200
> > > > Filter: box 1/1 )
> > > >
> > > > Test 1:
> > > > Rez: 1024
> > > > Quality: Highest
> > > > Generation time:1h30
> > > >
> > > > Test 2:
> > > > Rez: 1024
> > > > Quality: Medium
> > > > Generation time: 30 min.
> > > >
> > > >
> > > > Those tests were done on a reasonably performant machine (dual
> > proc
> > > > Athlon MP 2200 1.8 Ghz with 2Gb of ram)
> > > >
> > > >
> > > >
> > > > So the question is, are those map generation time actually
> > standard?
> > > >
> > > >
> > > >
> > > > Thanks
> > > > Bernard
> > > > ---
> > > > Unsubscribe? Mail Majordomo(at)Softimage.COM
> > <mailto:Majordomo(at)Softimage.COM> with the following text in
> > > > body:
> > > > unsubscribe xsi
> > > >
> > >
> > > ---
> > > Unsubscribe? Mail Majordomo(at)Softimage.COM
> > <mailto:Majordomo(at)Softimage.COM> with the following text in body:
> > > unsubscribe xsi
> > >
> > ---
> > Unsubscribe? Mail Majordomo(at)Softimage.COM
> > <mailto:Majordomo(at)Softimage.COM> with the following text in body:
> > unsubscribe xsi
> >
> >
> >
> >
> > --
> > Ben Barker
> > Hair/Fur
> > Omation
>
>
> ---
> Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
> unsubscribe xsi
>
