Glad you found a solution. It makes sense, the lighting command msDepthBias option (default 0.01) is specified as a fraction of the scene diameter. The default value of 0.01 means that crevices with depth is less than 1% of the scene diameter won't be shadowed. Since your scene diameter is large relative to the crevices you want to shadow you need a smaller depth bias value. The default setting is trying to give best appearance in typical cases -- shadowing bigger features instead of very small ones.

On Oct 14, 2019, at 1:18 AM, Gabor PAPAI wrote:

Dear Tom,

The second model was a very long DNA (around 1000 Å). I guess the scene size changed how the shadows are cast. Cropping the DNA helps, but I try to set up a large scene. I tried to increase the msMapSize this does not help. However when I lowered the msDepthBias from the default 0.01 to 0.001, the shadowing became close to the appearance of the first model alone.

Best regards,

Gabor

---
Gabor Papai
IGBMC
Integrated Structural Biology
1, rue Laurent Fries, Bp 10142
67404 Illkirch, France
Tel: +33-3-69485288

On 11 Oct 2019, at 19:12 , Tom Goddard <goddard@sonic.net> wrote:

Hi Gabor,

The soft lighting is supposed to look as if shadows are cast from all directions, so you would not expect it to change when a second model is opened (except to the extent that the second model shadows the first). But that is the ideal situation and approximations have to be made to compute the lighting fast enough and I am pretty sure those approximations are causing your problem. Take a look at the lighting command and try increasing the msMapSize (multi-shadow depth map size) from 1024 to a larger value like 2048 and see if that makes shadows darker. This parameter will effect shadow appearance as the size of the scene changes. Basically it effects the granularity of the shadows -- how large a crevice in your molecule needs to be before it appears darker due to shadows. Other multi-shadow lighting parameters "multiShadow" (number of directions to cast shadows from (default 64) and msDepthBias (avoids self-shadowing, default 0.01 times scene size).

I have a meeting now so I cannot experiment with this. Tell us if you find something that works.

Tom

On Oct 11, 2019, at 1:57 AM, Gabor PAPAI <papai@igbmc.fr> wrote:

Dear ChimeraX team,

If I open a second model in ChimeraX (0.91) the ambient occlusion in soft lighting mode on the first model disappears or becomes diminished. How can I revert this effect? I attach an image to illustrate this phenomenon. The only change between the two displays that I opened a model of DNA and placed close to the first model.

Best regards,

Gabor

<lighting_change.png>

---
Gabor Papai
IGBMC
Integrated Structural Biology
1, rue Laurent Fries, Bp 10142
67404 Illkirch, France
Tel: +33-3-69485288

_______________________________________________
ChimeraX-users mailing list
ChimeraX-users@cgl.ucsf.edu
Manage subscription:
http://www.rbvi.ucsf.edu/mailman/listinfo/chimerax-users