DICOM: was "airways" preset always so dim?

[Elaine Meng]

I wonder if something changed with the airways preset (from clicking the lungs icon). I don't recall it being so much more dim than the "Airways II" preset. I have an image in the "DICOM in VR" tutorial from after using the airways preset by clicking the icon, and it seems reasonably bright:
<​https://rbvi.ucsf.edu/chimerax/docs/user/tutorials/dicom-vr5-crop.png>

Currently the airways preset sets intensity at higher levels than Airways & Airways II, but unlike them has dimtransparentvoxels true, so maybe it is all intentional.

I only made a ticket because I became concerned that it looks extremely dim now, and might been changed by accident since I created the VR tutorial. However, I could be wrong, and ticket can just be closed if current behavior is intentional.

[goddard@…]

Worth looking at whether some inadvertent code change broke this.  We can check in Github.  I'd more suspect that the dimness depends on the thickness.  I recall some presets working well on single slices and not full volumes.  The ChimeraX volume rendering in general is not able to show a single plane and a thick volume using the same brightness/transparency curve settings where both 2d and 3d look good.  Maybe I rejiggered the airways preset to look good in 2D but not 3D.

[Tom Goddard]

The VR dicom lung image you link to appears to be maybe 5 or more years old. The volume viewer histogram is clearly visible and has a very low peak of the yellow curve. That same appearance is obtained now by using "volume dimtransparentvoxels false", as in the attached dimvoxelsfalse.png image with brightness curve peak height reduced to about 0.3. While dimtransparentvoxels = false works fine for the full volume, it is way too bright if the brightness curve height is 1.0. On the other hand, for a single plane view that setting has no effect and the height 0.3 brightness curve makes a single plane nearly invisible, while height 1.0 is reasonable.

So I think the dimtransparentvoxels true setting is desirable to make brightness less sensitive to how many planes are shown.

Ultimately the limited brightness parameters in our image rendering can't make the image always the right brightness independent of thickness using the same yellow brightness curve.

Might be worth making a separate ticket for that.

I don't think anything has gone wrong with the airways preset. Probably in the distant past it was horribly dim or horribly bright depending on thickness or brightness curve and now the compromise may be different. In the 2022 ChimeraX R01 there is a section for better 3D image lighting. Maybe 2 to 6 months of effort could go into exploring that to find more flexible and effective lighting models.

[Tom Goddard]

Too bright, dim voxels false.

[Tom Goddard]

Good brightness, dim voxels false, brightness curve reduced to 0.3 height.

[Tom Goddard]

Dim appearance, current airways on full volume, dim voxels true.