inksandpaper
Offline
•
Upload & Sell: Off
|
 p.1 #20 · Sensor and lens tilt and how to DIY adjust your equipment | |
Fred Miranda wrote:
I've determined those bodies to be out of spec without using a micrometer.
I prefer doing a quick trial and error experiment. If I have 3 or 4 well centered lenses performing equally great on 2 or 3 bodies, I assume the lenses and bodies are optimal. If I add a new body to the mix and the results are different, it's easy to spot the issue.
For example: I have a well centered Sigma 14-24 @14mm@@ performing great on my A7R IV and A7R III bodies but when I tested it on a new A7R II, it was noticeably softer on one side of the image field. I knew that something was off with sensor/mount parallelism....Show more →
Interesting, thanks. That was also how Joseph Holmes first suspected his camera to be crooked. You can't use a micrometer to measure sensor skew BTW. When you try many combinations, you can spot trends. Two bodies are not so good for a deterministic judgement, three increases the probability you're right by quite a bit.
For what it's worth, my Voigtlander 65mm A-L performs identically on my a7R II and then-new a7R IV (no longer with me), but the a7R IV was actually crooked by about 30 microns. The bayonet machining differences accounted for that. I was very surprised to discover this, not expecting the bayonets to vary by so much. I wish I have several hundred cameras on hand and an assistant to gather more representative data. It's very, very, very complicated - the errors can come from anywhere and negate each other.
When you say 'well-centered', I take it that you mean both optically centered (all four corners/edges perform equally when they are in focus, which you will know by using magnified live view to refocus on only that part of the image field), but I assume you also mean the focus plane is also not tilted (you are not refocusing in the corners during your testing to compensate, which I know you do to check for field curvature in your tests). I feel it's important to distinguish the three issues, centering, tilt and curvature.
It's interesting that you found that shimming the lens at the mount for tilt is effective. I've never tried that before because I always assumed that if one element is slightly tilted, that could not be fixed at the mount.
What process do you use to shim the lens at the mount?
Yes, it is very effective for focus plane tilt. If a lens is centered but field tilted (it doesnt matter at this point to know if one or more elements inside are tilted since it's centered), it is worth saving, because centered lenses are already hard enough to come by, while every lens I've ever seen is tilted (Roger Cicala also says the same about tilted lenses, and his words carry way more weight than mine).
The right kind of test subjects and knowing how to look are important to reveal if the lens is merely tilted and not decentered. And to also reveal to what extent is the focus field curvature contributing to the problem. Field tilt needs to be separated from curvature. You're finding the line of best fit through the curvature and trying to balance that to be parallel to the image place.
https://cacaoeditions.com/your-cameras-and-lenses-are-crooked/#adjusting-your-lenses
This section would be what you are looking for. You need a 'giant ruler' test subject to see and evaluate tilt by finding the focus plane visually, compute it in terms of microns-worth at the image sensor, translate that amount to the screw distances on the lens' bayonet, and then put little ring shims in the correct orientation around the screw holes under the bayonet mount. It takes a little trial and error to get it right, several rounds of adjustments are necessary to get a good correction. This is normal, the folks at Lensrentals/service centres often say the same of any optical adjustment.
I gather the fussiest photographers, usually the astro-folks, have been doing this for decades. What I am proposing is not new, rather it is a refinement of the technique. Marianne Oelund in particular, mentioned in the first post, who is well respected on the DPR forums explained not long ago how she shims her lenses for just the same reasons as us.
|
