p.5 #1 · Sensor and lens tilt and how to DIY adjust your equipment
inksandpaper wrote:
I'm not 100% sure (because I dont have the same parts to try) but I'm pretty sure it should work well. Assuming the adapter indeed behaves as described. I trust the Contax is sharp enough wide open for you to visually identify the true center of the plane of focus very precisely. Observing the blur gradient of the near edge of the depth of field is necessary to be certain. If its sharp in the corners doesn't mean the camera is straight. Offhand without calculations, the depth of focus at f/4.5 to 5.6 is over 30 microns, possibly over 50. This is way harder to see than one could ever expect, for those attempting this the first time....Show more →
This tripod lens adapter attaches to the zoom barrel and restricts the 100-130mm FLs. It's possible to use the lens starting at 135mm and it's very sharp at center even wide open. Not quite sure what the aperture would be at 135mm but perhaps f/5.0.
It does matter in that you have to pan the camera over if you are flipping the lens to put the test subject on the opposite side of the frame. Otherwise you're not eliminating the error introduced by the lens's optics.
p.5 #2 · Sensor and lens tilt and how to DIY adjust your equipment
I’ve got the Fotodiox adapter at home. It isn’t any good for shifting in my opinion because it uses some non-stick material to make the shift mechanism slide, the only problem is it’s reflective which causes internal flaring. So I took it apart (don’t do that) and later found out it is impossible to get back together. haha!
It will be fine for your intended use.
Next I bought a Kipon Emount to EOS shift adapter. It had a tripod mount, but I found out it was for Sony Crop cameras, not full frame. But rather than sending it back I decided to modify it (don’t do that). I’ve made it accept full frame now, but needed to make a bellows for it to avoid light leaks (which I did), which turned out to be very impractical in the field.
Last year I finally purchased a Mirex adapter. Sadly they are gone now you say. They are the best.
All these adapters will be fine for your intended use, but Mirex is best for shifting and tilting (still flares though into the sun).
I use mine Mirex mostly on my Canon 200 2.8 II for tilt for landscapes.
I also use it Mamyia 80 2.8 for shift or tilt. I got a Canon TSE90 II (the sharpest lens I’ve ever used). Sharper, but much heavier the the Mamyia. But I’ve tended to go with my VC 110/2.4 over these options, has macro and a lot smaller than the Canon. Can’t bring myself to downsize my kit.
p.5 #3 · Sensor and lens tilt and how to DIY adjust your equipment
Fred Miranda wrote:
This tripod lens adapter attaches to the zoom barrel and restricts the 100-130mm FLs. It's possible to use the lens starting at 135mm and it's very sharp at center even wide open. Not quite sure what the aperture would be at 135mm but perhaps f/5.0.
p.5 #4 · Sensor and lens tilt and how to DIY adjust your equipment
TakenWild wrote:
I’ve got the Fotodiox adapter at home. It isn’t any good for shifting in my opinion because it uses some non-stick material to make the shift mechanism slide, the only problem is it’s reflective which causes internal flaring. So I took it apart (don’t do that) and later found out it is impossible to get back together. haha!
It will be fine for your intended use.
Next I bought a Kipon Emount to EOS shift adapter. It had a tripod mount, but I found out it was for Sony Crop cameras, not full frame. But rather than sending it back I decided to modify it (don’t do that). I’ve made it accept full frame now, but needed to make a bellows for it to avoid light leaks (which I did), which turned out to be very impractical in the field.
Last year I finally purchased a Mirex adapter. Sadly they are gone now you say. They are the best.
All these adapters will be fine for your intended use, but Mirex is best for shifting and tilting (still flares though into the sun).
I use mine Mirex mostly on my Canon 200 2.8 II for tilt for landscapes.
I also use it Mamyia 80 2.8 for shift or tilt. I got a Canon TSE90 II (the sharpest lens I’ve ever used). Sharper, but much heavier the the Mamyia. But I’ve tended to go with my VC 110/2.4 over these options, has macro and a lot smaller than the Canon. Can’t bring myself to downsize my kit....Show more →
Thanks for the rundown on the Kipon and Fotodiox adapters. I've not held either in my hands before. Light leak complaints I found were worrisome indeed. And the fit issues too.
I had to add flocking to the insides of my Mirex adapter to eliminate flaring, as well as a light baffle. I can share some pictures when I'm back home on my main computer. My Mirex is a different one than yours so unfortunately the carefully made cutting template for the flocking I have wont suit yours. I've tried to find used Mirex adapters for sale. Nothing! If anyone doesn't want theirs anymore, please let me know. I'd be very interested to acquire it from you.
p.5 #5 · Sensor and lens tilt and how to DIY adjust your equipment
Fantastic write-up and a very entertaining read!
I don't have much to offer to the effort except a minor tip that might help better visualize the sharpness differences when comparing the two edges of the frame/sensor - I suggest overlaying the two find-edges images on top of each other and generating an animated GIF of the overlay. It makes sharpness differences much easier to visualize. For example:
As compared to the natural image pair:
I'd also suggest playing with different edge-detection algorithms, particularly Sobel Gx vs Gy, which will isolate the two axis. I posted samples of this here.
p.5 #7 · Sensor and lens tilt and how to DIY adjust your equipment
snapsy wrote:
Fantastic write-up and a very entertaining read!
I don't have much to offer to the effort except a minor tip that might help better visualize the sharpness differences when comparing the two edges of the frame/sensor - I suggest overlaying the two find-edges images on top of each other and generating an animated GIF of the overlay. It makes sharpness differences much easier to visualize. For example:
I'd also suggest playing with different edge-detection algorithms, particularly Sobel Gx vs Gy, which will isolate the two axis. I posted samples of this here.
Thanks @snapsy! You're much more advanced than I am in many technical fields, so it's very humbling to know you enjoyed the article. Thanks also for the suggestion of the find-edges visualisation technique. I've used it many times before but found it of limited use when I'm making the final shimming corrections to within 5 microns of perfectly straight. The outlines are too thick and don't let me see sufficiently well very subtle focus shifts.
Maybe the Sobel Gx vs Gy would be more revealing, though I have no idea how to implement that in Photoshop yet. In the end, after 40 months of unintentional and unavoidable visual exercise wrestling with these issues, I eventually learned the technique to see it for what it is, which I felt was better than using visual aids. But doubtless, they are a worthwhile technique and can highlight something that may accidentally go unnoticed. I shall explore it some more.
p.5 #8 · Sensor and lens tilt and how to DIY adjust your equipment
inksandpaper wrote:
Thanks @snapsy@! You're much more advanced than I am in many technical fields, so it's very humbling to know you enjoyed the article. Thanks also for the suggestion of the find-edges visualisation technique. I've used it many times before but found it of limited use when I'm making the final shimming corrections to within 5 microns of perfectly straight. The outlines are too thick and don't let me see sufficiently well very subtle focus shifts.
Maybe the Sobel Gx vs Gy would be more revealing, though I have no idea how to implement that in Photoshop yet. In the end, after 40 months of unintentional and unavoidable visual exercise wrestling with these issues, I eventually learned the technique to see it for what it is, which I felt was better than using visual aids. But doubtless, they are a worthwhile technique and can highlight something that may accidentally go unnoticed. I shall explore it some more....Show more →
You can apply the Sobel kernel in Photoshop by going to Filter -> Other -> Custom and entering the Gx or Gy kernel (source). If you want to apply both you can do each on a separate layer.
p.5 #9 · Sensor and lens tilt and how to DIY adjust your equipment
snapsy wrote:
You can apply the Sobel kernel in Photoshop by going to Filter -> Other -> Custom and entering the Gx or Gy kernel (source). If you want to apply both you can do each on a separate layer.
Thanks so much for the tip, I was just playing around with it and the Sobel Gy kernel suits my track test subject superbly well and was very effective at revealing subtle focus shifts! It's wonderful, feels better than the Find Edges filter.
p.5 #10 · Sensor and lens tilt and how to DIY adjust your equipment
I’ve got some micrometers now and tested most of my Canon to Sony adapters.
Astonishingly, my Mirex Canon to Sony shift-tilt adapter had no Symmetry issues at all. Amazingly well made. Sad they have gone out of business.
The next best is my Sigma mc-11 adapter 40 micrometers out from side to side at worst. Then my dumb Kipon adapter with 90 micrometers out at worst. My metabones 4 adapter is 80 micrometers out at worst and my metabones 3 adapter is 90 micrometers at worst.
I leant some interesting information about why my Canon TSE 17 and 24 lenses have trouble with metabones and my Mirex adapters. They are too short. My sigma mc-11 is 150 micrometers more than Mirex and 100 micrometers more than my Metabones 4.
I think the 17TSE the MC-11 is about right and the 24TSE needs an extra 50 micrometers on top of that to get the best out of it.
p.5 #11 · Sensor and lens tilt and how to DIY adjust your equipment
TakenWild wrote:
I’ve got some micrometers now and tested most of my Canon to Sony adapters.
Astonishingly, my Mirex Canon to Sony shift-tilt adapter had no Symmetry issues at all. Amazingly well made. Sad they have gone out of business.
The next best is my Sigma mc-11 adapter 40 micrometers out from side to side at worst. Then my dumb Kipon adapter with 90 micrometers out at worst. My metabones 4 adapter is 80 micrometers out at worst and my metabones 3 adapter is 90 micrometers at worst.
I leant some interesting information about why my Canon TSE 17 and 24 lenses have trouble with metabones and my Mirex adapters. They are too short. My sigma mc-11 is 150 micrometers more than Mirex and 100 micrometers more than my Metabones 4.
I think the 17TSE the MC-11 is about right and the 24TSE needs an extra 50 micrometers on top of that to get the best out of it....Show more →
Thanks for the update. Remarkable about the Mirex. As I recall, mine is really good too. The 360-degree rotating rear-end wobble was also well under 10 microns.
Too short isn't a problem, rather too long would be. I would be hesitant to shim the adapters to an exact thickness, since the flange distance of Sony cameras seem to vary by up to 100 microns (not counting extreme ouliers), and thermal expansion in hotter weather could add to the problem. Overall too-thin adapters don't cause optical problems with TS-E lenses, shifted or not. It's one side thicker than the other that is the problem. 80-90 microns of tilt is quite substantial. My Metabones was something like 20+ microns tilted, and my Novoflex is 10+. Those are serious tilts you're looking at, by my standards.
p.5 #12 · Sensor and lens tilt and how to DIY adjust your equipment
I just upgraded to a Sony a7rIV from an rII and the rIV has sensor tilt and is visible on a few of my lenses. I am not sure how bad the tilt is, but using a sigma 65mm f2 lens, at f2 if I focus on a tree on the left of frame at 18m, on the right the lens needs to be focussed at 26m to be sharp. Testing like this seems to be a simple way to see if the sensor is off.
I have sent the rIV back for replacement as I discovered it within 2 days of buying so I hope I will get one that is better!
p.5 #13 · Sensor and lens tilt and how to DIY adjust your equipment
That's a 70 micron tilt. How did the return go? I have an a7rV with the same amount of tilt and wonder if I send it to Sony under warranty whether they'll just send it back to me unrepaired "within spec".
p.5 #14 · Sensor and lens tilt and how to DIY adjust your equipment
For anyone in the future who's the same trouble, here's what I did when I received an apparently decentered lens:
I very recently ordered a Sony 40mm f/2.5G to match my a7Riv, it's my first native and first AF lens, all I used before were adapted lenses since 98% of the time I'm shooting on film but I wanted to make use of digital more for the convenience and reliability. I chose the 40mm because it's the closest to my Mamiya 7 80mm.
As soon as I got it I tested it for centering using the 4 corners at infinity and shooting some detailed paper on a wall at around 3 or 4 meters. Results were mixed, a corner was basically as sharp as the center, 2 corners were good but not great and the last one was noticeably weaker. I was disappointed, but I started doing more and more testing.
I found out one side was always focused closer, which raised some questions. I redid the corners test but this time refocusing on each corner, performance was equally great this time, so the lens was capable of producing sharp corners but not at the same time. I tried the zeiss Siemens star and the defocusing test didn't show any decentering...
Doing more research I found this thread, very eye opening! But... What now? I don't have the rotating adapter or a micrometer, and I'm pretty broke right now plus I live in Costa Rica, not the most technically resourceful country.
I remember the find edges filter test in Photoshop and ran it through all my lenses: the Sony, 3 adapted Nikkors and 3 adapted FD lenses. All of them showed a similar tendency: some swing and tilt in the same direction, so I assumed my camera body was the culprit. How do I shim that thing without importing tiny shims from the US? Used acetate sheets and it worked surprisingly well! I had bought them for wet film scanning, it's a cheap 25 sheets package.
Now the problem is, how do I measure the sheets and shape them? I didn't, I cut them to size and hoped for the best and simply assumed my lens was ok because all showed similar tendency. To adjust the thickness of the acetate shims I used sandpaper, it works surprisingly well. With some trial and error on both X and Y orientations I managed to get a very acceptable result, always judging by the find edges test on Photoshop
