> I cobbled together a scene similar to the one posted, and it's really hard to
> get that SEM 'look.' Even more irritating is that the indents almost always look
> like bumps. I tried lights, radiosity with no lights, a radiosity 'ring' around
> the object, no_shadow, etc. This is the closest I've gotten--and it still looks
> somewhat ambiguous.
>
> There's something intrinsic to electron microscope and SEM images that might
> make recreating the look with a 'standard' lighting setup difficult.
>
> Most such images appear to have 'lighting' that *surrounds* the object--except
> in the front. And no shadows, in the usual meaning. They almost look like
> photographic negatives. (I found one image--a really cool one of an ant--and
> 'inverted' it in Photoshop; the result looks like it's lit with a SINGLE light
> source, shining directly from the camera position...and with a rapid fall-off of
> light on all the side/slanting surfaces.) How to get the reverse of that with
> 'positive' lighting may not be so easy. For example, it's hard to get regular
> light sources to 'fall off' in brightness in just such a way as those SEM
> indents show. If the object was like an ant--in other words, a detailed 3D model
> in space--then maybe a radiosity set-up might work, using a white HEMISPHERE
> behind the model as the only 'light source.' But a simple planar surface--like
> the OP's model--can't take advantage of that.
>
> So my final idea was kind of a cheat: to color the planar surface with a
> gradient y pigment, receding into the surface--somewhat dark at the very
> surface, then immediately lighter, then falling to darkness again to the bottom
> of the pits. And using two area lights to either side.
>
One way could be to use a negative light, maybe with a fade_power larger 
than 2, and a bright surrounding providing radiosity illumination.
You could also use a ring of actual lights.


Alain

Post a reply to this message