The impact of tilt on color cast

[dchew]

In the IQ3 100 thread I made the statement to Bernard that tilt does not impact color cast issues because the angle of light rays hitting the sensor doesn't change. Several people chimed in and said it does, just not dramatically. I am confused by this. In my very crude drawing below where the top plane is the rear nodal point of the lens and the bottom plane is the sensor, I don't see how the angle changes when you tilt. Lens design obviously is more complicated, and they won't tilt around the nodal point, but I can't see how that is a big deal. I am also assuming tilt occurs along the centerline of the lens, but that is how Alpa works and I think most lens boards (?)

It seems a much bigger factor is that when doing the classic tilt where the photographer wants a horizontal plane in focus from close to the tripod out to infinity, we dial in tilt at infinity focus. Infinity focus is the worst case since the lens is closer to the sensor and the angle is at its widest. Isn't that really the driver to why you might see more cast when using tilt?

[Paul2660]

I don't know the answer, but love the drawings!! 

On the Arca, rm3di, tilt is made by moving the lens from the front controls.  Shift is done by moving the digital back, side to side or up and down.  I believe Cambo has the same setup, not sure on Alpa.  Shift definitely brings in color cast, but I have never noticed it with tilt on wides.  28mm, Rodie, 32mm Rodie, 35mm SK (now for sale)  or 40mm Rodie.  On the 28mm I rarely needed more than 1/2 a degree of tilt to improve foreground DOF.  The rm3di has 5 degrees up or down.  On the 35mm SK, 1.5 degrees was max and on the 40mm 1 degree.  With all of these shots, the lens was still centered and the LCC showed no color cast like on a shift where the edge will be surely showing a red cast to some degree.

Paul C

[Bart_van_der_Wolf]

In the IQ3 100 thread I made the statement to Bernard that tilt does not impact color cast issues because the angle of light rays hitting the sensor doesn't change.

Hi Dave,

It depends on where the tilt axis is positioned in the lens design. As soon as the entrance pupil of the lens shifts as a result of tilt, i.e. the rotation axis is not in the entrance pupil plane, then the image will shift a bit (the perspective changes). That small shift amount will cause a slight change in the ray's angles. So if the rotation axis is aligned with the entrance pupil of the lens, then nothing much happens other than the principle plane and focus plane rotation.

It's usually easy to see if the image moves when the lens is tilted, and one might want to compensate with an opposite shift to keep a given composition.

Cheers,
Bart

[dchew]

I don't know the answer, but love the drawings!! 

Yeah we are just verifying why I pick up a camera instead of a brush.

Alpa can actually tilt, shift or rise/fall from the lens side or the back side depending on where you place the adapter/lens combo. But yes, normally you place the tilt adapter on the front of the camera and the shift assembly is to the rear where the back is mounted. One of the reasons I like the Alpa design is you can flip the camera around and shift the lens / tilt the back if you want to. Although I must admit I rarely do that.

Hi Dave,

It depends on where the tilt axis is positioned in the lens design. As soon as the entrance pupil of the lens shifts as a result of tilt, i.e. the rotation axis is not in the entrance pupil plane, then the image will shift a bit (the perspective changes). That small shift amount will cause a slight change in the ray's angles. So if the rotation axis is aligned with the entrance pupil of the lens, then nothing much happens other than the principle plane and focus plane rotation.

It's usually easy to see if the image moves when the lens is tilted, and one might want to compensate with an opposite shift to keep a given composition.

Cheers,
Bart

Thanks Bart. I suppose my error is in my simplification of things (as usual). Alpa tilts at the front of the tilt adapter, which I would think in most cases is closer to the entrance pupil plane. Perhaps that is why I've never seen any effect. Also my widest lens is the 40hr. Paul reports not seeing any either, and he has much wider lenses than I do. That makes me think whatever effect there is must be quite small.

Dave

[JoeKitchen]

Hi Dave,

It depends on where the tilt axis is positioned in the lens design. As soon as the entrance pupil of the lens shifts as a result of tilt, i.e. the rotation axis is not in the entrance pupil plane, then the image will shift a bit (the perspective changes). That small shift amount will cause a slight change in the ray's angles. So if the rotation axis is aligned with the entrance pupil of the lens, then nothing much happens other than the principle plane and focus plane rotation.

It's usually easy to see if the image moves when the lens is tilted, and one might want to compensate with an opposite shift to keep a given composition.

Cheers,
Bart

Yes, if you are performing axil tilt, the angle at which the light hits the sensor will not change.  However, if you are performing base tilt, which is pretty much what you are doing with longer lenses on technical plate cameras, the angle will change since the position of the nodal point changes as well. 

For instance with my RM3Di, my SK 35mm is mounted so the nodal point is very close to, if not right at, the axis that the helicoid tilts on.  So tilt is really not going to change the angle of light hitting the sensor.  However with my Rodie 90mm, the nodal point (and the lens itself) is a good distance away from the tilting axis.  So tilting here will move the lens, changing the angle that the light comes in at.   

Wow, I really need to proof read these posts in the morning. 

[torger]

I hadn't really thought about that nodal point tilt doesn't change the angle, but you're right that's true.

However I would guess that no tech camera system + lens is that nodal that you have zero change to color cast if you tilt. I've always assumed that there are small changes to color cast even with small "nodal" tilts, but I haven't actually done a test.

[yaya]

I hadn't really thought about that nodal point tilt doesn't change the angle, but you're right that's true.

However I would guess that no tech camera system + lens is that nodal that you have zero change to color cast if you tilt. I've always assumed that there are small changes to color cast even with small "nodal" tilts, but I haven't actually done a test.

If you are moving/ tilting the focus plane then you will get changes in the colour cast. It's the same as changing the focus point...However like you say with wide angle lenses the typical amount of tilt is very small and therefore doesn't change the amount of cast, just the shape (which also means there's normally a need for another LCC when tilting the lens).