Thank you for the response once again, this gives me something to work
on over the Christmas break ;-)

Just to be clear, when I said 'nothing displays' what I meant was the
Moon renders the surface bitmap, but the normalmap does not seem to be
applied - the craters are not shaded along the terminator. When the
movie is displayed in CS4 the normal map is applied correctly.

I am using the [Embed(source="LALT_normalmapV2_2048x1024.png")]
construct to load the files into the swf file.

I have a very nice 5760 pixel bump map (and the same size tangent
nomal map), though I am having to scale down the bumps in Photoshop to
make them more realistic.

Mark


On Dec 22, 3:20 pm, Fabrice3D <[email protected]> wrote:
> > I'll see if I can get my head around creating a normal map from a hi-
> > res model to lo-res in Prefab. Though as we know where the surface of
> > sphere should be, is it not possible to calculate a 'perfect' normal
> > map for any specific resolution model?
>
> For a sphere and many other geometrical shapes yes, in theory. That's what 
> Stuart class is doing in case of the sphere.
> In the practice, depending on size of the map, you get artefacts because you 
> can't draw smaller than a pixel
> and flash tends to use nearby pixel info when the map is blowed up.
>
> Prefab has 3 weapons against this:
> - you have a blur option, in case of a sphere, it would not be a problem use 
> 2 or 4 on a bigger map
> - you can render at bigger size and export at desired size, plus you have a 
> little preview where you can inspect the results right away
> - thx to Mario, we have a bleeding of colors filling up the regions arround 
> the traced area's.
>
> note that for your moon, you can also load a bumpmap, it will alter the map 
> as well.
> in this case, no need to load the highres file if you have a good bumpmap 
> available.
>
> > Now I have a problem that the normal map shading shows up fine in the
> > CS4 IDE preview, but when the movie is viewed in a HTML container (Win
> > 7 Gadget), it does not appear at all :-(
>
> I'm not using this system (and never will) but are you sure its related to 
> it? One blind shot at the cause might be
> the url if you use external sources. Are you sure the to be loaded sources 
> are at the right location?
> If you see nothing, it often points at a fatal error like this.
>
> Fabrice
>
> On Dec 22, 2009, at 3:53 PM, MarkC wrote:
>
>

Reply via email to