Exposure mapping

[polka]

Vignettage instead of vignetting ?

This would be the opposite of "parler franglais" : "englench speaking" ? 8)

Paul

[durian]

Thanks for the birthday wishes, 44 years old...good that life starts at 40 anyway

I implemented the "blur" formulas as well: http://durian.se/pinhole/blur.html

Polka, thanks for the scan of the notes!

-peter

[polka]

Another Bravo ! Now I would be pleased to see the same tool for anamorph cams. the model is not so different, it should not be so difficult ?

Here another paperwork scan

Paul Attached files

[polka]

And two other scans to complete...

Paul

To comment the two vignetting formulas :
The numerator (?) of the fraction is the shape reduction due to the thickness of the pinhole plate (e=0 always gives 1)
The denominator (?) is the ovalisation of this shape due to grazing incidence

To comment on the blurring formulas :
Geometric means circles (actually shapes) "of confusion"
Diffractive means first (central) diffraction zone
Radial means in the direction following a line from the center of image to relevant point
Tangential means perpendicular to this direction
When optimized pinhole diameter (with usual tools), the 4 values are the same (in flat but not in anamorph cams) for x=0 Attached files

[toniox]

Great pinhole scientists here !

[durian]

Another Bravo ! Now I would be pleased to see the same tool for anamorph cams. the model is not so different, it should not be so difficult ?

Here another paperwork scan

Paul

Et voilÃ*: http://www.durian.se/pinhole/anamorphevdrop.html

I'm not sure I implemented it correctly though, it's at least a totally different graph. I plotted f(x) = 2/pi * (acos(ef/xd)... &c

-peter

[polka]

Hi Peter,

I guess the only thing missing is to take log base 2 of this expression ? Then you should get this (with your default values)

red : classic flat plane with 0.0 pinhole plane thickness
blue : classic flat plane with 0.01 pinhole plane thickness
green : anamorph with 0.0 pinhole plane thickness
yellow : anamorph with 0.01 pinhole plane thickness

Paul Attached files

[durian]

Hi Peter,

I guess the only thing missing is to take log base 2 of this expression ? Then you should get this (with your default values)

red : classic flat plane with 0.0 pinhole plane thickness
blue : classic flat plane with 0.01 pinhole plane thickness
green : anamorph with 0.0 pinhole plane thickness
yellow : anamorph with 0.01 pinhole plane thickness

Paul

Of course, fixed! Now it looks like your graph!

-peter

[polka]

To be noted that for very extreme angles : on one side, the distance to film is minimal but the ovalisation is maximal ans on the other side it is the opposite, so that : as these two effect tend to counteract, light falloff is less severe and the maximum EV is for 60° (when pinhole plate thickness is 0, approximately so when it is non negligible). Thus, in order for that maximum not to be too near the edge of the film, one should keep the pinhole plane at shortest distance possible from the upper edge of the film (in my POLKa for the picture window, x lies between 8mm and 64mm, for cylinder radius f=36mm).

To be noted also that the actual aperture at maximum may not be simply taken as d/f, it is approx. 1.5 EV under that, because of ovalisation etc.

Very nice work Peter !

Paul

P.S. To answer one of your questions, I tried to invert the models, and found that I cannot solve the equations that I get analytically, except for a regular (flat film) pinhole camera with zero pinhole plate thickness. In this case, I get :

x(EV) = f * SQRT( (SQRT(2)) ** (-EV) - 1 )

And one can note that x(-2) = f => it is at f from the center of the frame that you loose 2 stops

[durian]

P.S. To answer one of your questions, I tried to invert the models, and found that I cannot solve the equations that I get analytically, except for a regular (flat film) pinhole camera with zero pinhole plate thickness. In this case, I get :

x(EV) = f * SQRT( (SQRT(2)) ** (-EV) - 1 )

And one can note that x(-2) = f => it is at f from the center of the frame that you loose 2 stops

Yes, indeed.

By the way, I tried to do the anamorphic blur circles as well: http://www.durian.se/pinhole/anamorphblur.html

Does it look right ?:-)

-peter