Lateral CA partially explained (1 Viewer)

For you optics buffs, here is an interesting concept, due to the great physicist R. A. Stokes. First proposed as a method for blatantly exposing CA in well corrected achromatic telescopes, I think it helps us understand some of the color fringing effects seen in binoculars. What follows is my rephrasing from H. Dennis Taylor's classic, "Adustment and Testing of Telescope Objectives".

The achromatic objective produces a longitudinal secondary spectrum, with yellow green focused closer to, and red and blue farther from the lens. The user will choose a focus setting somewhere between the two, as the best compromise for multicolored light. At this chosen point, consider the unfocused light that lies below the optical axis. It contains yellow green from the upper half of the objective lens, which has already crossed the optical axis, and also red and blue from the lower half of the objective, which has not yet crossed. Although lacking blue-green and orange, these colors mix together to give the general impression of white light. So, there is fringing, but it is not very noticeable, because the light is not strongly colored.

Now imagine what happens if the lower half of the objective is covered. The unfocused light below the chosen focus point is now all yellow green, and as a result, stands out glaringly, as color fringes. The fringe above the focal point is purely made of red and blue, so looks purple.

Covering half the objective and observing the effect is "Stokes' Test", and is an extremely sensitive test for color error. It is said to reveal errors even in the finest APO objectives.

Now begins my binocentric interpretation. Blocking half the exit pupil of a binocular, which is an image of the objective, is equivalent to blocking half of the objective. So, if the eyes are not well positioned, and the edge of the iris cuts off part of the exit pupil, color fringes will be seen. Stokes test shows this form of lateral CA and longitudinal CA to have the same origin, so a reduction in longitudinal CA, by means of Fluoride glass objectives, should also reduce the view's sensitivity to eye position.

The way in which the part of the objective viewed through becomes skewed off center is a little different when the eye opening is smaller than the binocular's exit pupil (7x50 by daylight) or vice versa (8x20 at dusk). But in either case, if the eye is not centered, one is looking through an off-center part of the objective, and the principle of Stokes' Test applies.

There is another source of color fringing, near the edge of the field, that is due purely to a color effect in the eyepiece. I say this, because I have seen it in reflecting telescopes, whose objectives have no color error. Jupiter at the edge of view in fast Newtonian with a Nagler eyepiece, for example, takes on a decidedly putrid appearance, chartruse on one side, and magenta on the other.

I don't know how to separate these two sources of lateral CA, they must mix it up somewhat. But understanding Stokes' Test gives an explanation for the common appearance of near-centerfield color fringes, if the eyes are not carefully positioned, and the reduction in this effect with ED objectives.

So, is this old hat for you guys? Did I explain it OK? Am I missing something? Did I get something hopelessly mixed up?
Ron

Thanks Ron, that was a really interesting explanation. I´d often wondered about lateral CA, and having read your post, well...my brain hurts, I may have to go and lie down for a while, but there were a few brief flashes where I almost understood what was going on (a bit like Real Life, really...)

Thank you, Sancho. If my post has made even one brain hurt, it was not in vain.
ron

Thanks for the information, Ron.

Now I can add another possibility to a growing list of candidates for what causes lateral color in binoculars. A quick check of my homemade lateral color target indicates that covering the lower half of the objective largely eliminates lateral color in the lower half of the field in any binocular, while the upper half is unchanged. It remains something of a mystery to me why binoculars with similar exit pupils and field widths have such different amounts of lateral color, especially near the center of the field. The worst performers can be quite expensive and seem to have a few design traits in common, such as air spaced achromatic objectives, internal focusing elements and possibly roof prisms. Using ED glass in the objective appears to improve things, but does not seem to be a complete fix for binoculars like that, while even very cheap Porros with simple cemented doublet objectives and no focusing element can have very little lateral color except toward the edge where you expect it in the eyepiece.

Henry

...So, is this old hat for you guys? Did I explain it OK? Am I missing something? Did I get something hopelessly mixed up?
Ron

Click to expand...

Ron,

It seems to me from what you said the Stokes test reveals the extent of underlying longitudinal (axial) CA, assuming that lateral CA has been corrected by the achromatic design of the objective. Apparent color fringing is therefore enhanced by covering the lower half of the objective because the "... unfocused light below the chosen focus point is now all yellow green, and as a result, stands out glaringly, as color fringes. The fringe above the focal point is purely made of red and blue, so looks purple..." Correct me if I'm not getting this right. Covering half of an achromatic objective should produce more apparent fringing, not less. A related implication is that whether seen or not as color fringing, axial CA still degrades the image.

Regarding your interesting comments about eye-ocular interface effects, masking by the iris might contribute an increase in "apparent" CA (i.e., due to the Stokes effect), but the aggregate aberrations of off-axis or cross-axis viewing have never been clear, at least to me. It's not an unknowable subject, but it would no doubt take extensive computer modeling for a particular binocular design. Simplifying equations are needed, but so far I've not seen even a rough approximation to how CA (both types) produced by the objective, prism, eyepiece, and eye combine in the final retinal image. The nature of CA suggests that net effect involves cancellation as well enhancement.

Thanks for the provocation to think.

Ed

Henry and Ed,
Thanks for reading, and for your comments, which make me think too, although probably to little avail. I hardly thought I'd put this complicated matter to rest, only wanted to share one small thing I discovered that makes a little bit of sense to me.

Henry, thanks for the experiment. It didn't turn out like I would have expected from my reading of Stokes' effect. I tried it with my achromat telescope and it worked as I imagined it would. I guess an astronomical telescope viewing a point source is a lot simpler situation than a complex binocular viewing a scene. I believe that we may infer, however, if you covered the objective entirely, all the CA would disappear!
Ron