Replies

You persist in ignoring the fact that since my original posting I have posted three series where only the angle changes. . . .

Images lit this way do not show weird artifacts when processed by the emboss filter using arbitrary angles. Images lit by incident light are sensitive to angle. For an ordinary photograph, the nost natural looking effect is produced when the angle is close to the angle of incident light in the original scene.

So you have... and that raises the suspicious question of why did you change the offsets on the Shroud images?
Here is your original Xray image:

Now, let's extrude it using the same settings YOU used on your bogus 3D Shroud image: Angle: 18º, Height 60 Pixels, amount 48%

What do we get? Oh, wow. We get garbage—just like you produced in your bogus Shroud picture.
Where is your claimed lack of angle dependence in back lit Xrays? Let's look at one of your examples extrusions:

Now, here is the same Xray as I extruded above using the 60 pixel (or more) offset you used in your bogus Shroud picture—which produced garbage—this time with an offset of only 9 pixels, using the exact same angle. (I did change the opacity ratio percentage from 48% to 223% to more closely match your opacity settings):

Again, Wow, it looks very similar to your examples that you claim are not dependent on angle. They aren't and the Shroud isn't. Both are, however, extremely dependent on the offset, as I have been telling you, because that is inherent in the technique.
Where is your claimed angle dependence on the Shroud? It simply does not exist. I have demonstrated that merely changing the amount of offset back to a more reasonable level restores the pseudo 3D that you claim is angle dependent. Your blurry, bogus picture is an artifact of your manipulation and/or lack of understanding of what you are doing. There is no light source angle on the Shroud.
By the way, here is your Xray, extruded to the settings used in your Shroud picture with the opacity setting you used in the Xrays:

Not very useful, is it? Unless you want a starting point for a Rorschach inkblot.

When I was asked for the settings I used I posted them. You persist in using settings that I did not use.

Now to prevent you from making a complete idiot of yourself in public, try this: take an image such as the x-ray, find settings that give a reasonable 3D effect, and then vary only the angle. That is my methodology in all of the examples I have posted. Anything else is just stupidity.

What you refuse to accept is that despite all the technobabble, all of these renderings are false 3D. They are just interpreting density gradients as depth. There is no guarantee that the interpretation of depth represents anything real.

There is a clue to be found, however, when you discover that a good looking interpretation goes kablooey due to changing the angle control, that the gradients in the original image were the result of incident light.

If you are so stupid to start with an interpretation that doesn’t look 3D, there is no point in playing with the angle.

Now, I have done what grey whiskers asked me to do. I published a series of images showing the effect of varying the angle. How about if you do something for me. Select an image, fiddle with it to make a plausible interpretation, then rotate the angle and show us the results.

May I point out that you have failed to show a series of renderings starting with one good one and demonstrating what happens whin you change the angle and nothing else.

You have also failed to explain how the pupil of the eye shows up in an image that is a graph of the distance from body to cloth.