p.3 #1 · Should M-to-E adapters be 9.80mm or 9.74mm?
Makten wrote:
Ummm, I think all of my adapters are like 0.5 mm too thin (not measured, but checking where infinity focus really is with a lens attached). 0.05 mm is not gonna make any difference at all.
My understanding of a shorter-than-necessary adapter:
* non-optimal performance of lenses with floating elements
* greater than necessary MFD
* inability to rotate to infinity and be focused
If lenses and camera bodies are made to +/- 0.01mm, then the adapter should be as well -- and if you want to leave it short, it should be 0.01-0.02mm short. I don't see why it should be any shorter than that.
p.3 #2 · Should M-to-E adapters be 9.80mm or 9.74mm?
turnstyle wrote:
If lenses and camera bodies are made to +/- 0.01mm, then the adapter should be as well -- and if you want to leave it short, it should be 0.01-0.02mm short. I don't see why it should be any shorter than that.
Nor do I, but I think the lenses are nowhere close to those tolerances, and that's why ”more slack” is needed for generic adapters to work with every single lens. And then there is the different sensor filter stacks on cameras with the same mount.
If you get a lens with floating elements, you could of course shim the adapter to make it dead on at infinity (if the lens is perfectly collimated, that is), and let the rest of your lenses do what they do. There is sheets of brass suitable for shimming that you can buy and cut with scissors at ~0.025 mm and up, that should do the trick. Just measure the thickness of the adapter and make sure it's within ~0.05 mm all around, and you'll probably be fine regarding tilting of the lens.
p.3 #3 · Should M-to-E adapters be 9.80mm or 9.74mm?
Makten wrote:
Nor do I, but I think the lenses are nowhere close to those tolerances, and that's why ”more slack” is needed for generic adapters to work with every single lens.
Here's why I don't agree -- let's say we're just talking about M lenses and an M body -- it wouldn't "make sense" for the body to be intentionally short just to accommodate lenses that happen to be long.
The adapter should be built to whatever the spec is. If you want to build it just short of spec to accommodate your own +/- tolerance (to err on the side of being too short) -- that makes sense.
But it seems they're all being built considerably shorter than that.
p.3 #4 · Should M-to-E adapters be 9.80mm or 9.74mm?
turnstyle wrote:
Here's why I don't agree -- let's say we're just talking about M lenses and an M body -- it wouldn't "make sense" for the body to be intentionally short just to accommodate lenses that happen to be long.
The adapter should be built to whatever the spec is. If you want to build it just short of spec to accommodate your own +/- tolerance (to err on the side of being too short) -- that makes sense.
But it seems they're all being built considerably shorter than that.
I disagree, since I've owned a few M lenses that did not reach optically infinity focus on M cameras. The adapter manufacturers have to make the adapters too short because some lenses are off.
p.3 #5 · Should M-to-E adapters be 9.80mm or 9.74mm?
Let me begin by saying I don't have a simple answer. Hasselblad H takes the pain to match a particular lens to a particular camera body due to very minor manufacturing tolerance of the flange distance.
The body, the lens and the adapter can all be off by +/- 0.02mm or 0.03mm even when they are manufactured to the highest standards. So in the worse case scanario, everything is 0.03mm too thick and the total system is 0.09mm too thick. Given this, it can be argued the adapter should be made 0.1mm too thin to ensure infinity even in the worse case scenario. eBay Chinese adapters are typically 0.2mm too thin in order to "guarantee" infinity focus.
Not sure if you have read Roger Cicala's treatise that even expensive adapters are not good enough for critical optical bench testing. My personal experience has been that a difference of as little as 0.03mm has an impact on very fast or very wide angle lenses, even just pixel peeping but not lab testing.
So, we are in a limbo: make the adapter as exact as possible and it won't focus to infinity in the worse case scenario, or make it 0.1mm too thin and then it won't be fit for critical optical bench testing (and there would be a noticeable impact in the real world too for ultrawide and ultrafast). Metabones goes the as-exact-as-feasible route, which is why Metabones get the occasional infinity complaint, but big deal, they would always just offer a replacement to deal with it. To them, fouling up CRC for everyone is a greater evil than pumping up the return rate. I think the trick is, you can make it a hairline too thin and just take the beating of the occasional return, but resist the urge to overdo it.
The only way out of all this is to precisely match the thickness of an adapter to the exact lens and the exact camera body.
One more food of thought: every lens has field curvature and it may curve in or curve out. For those of you who expect the infinity stop of the focusing ring to be exact, is it exact at the center or in the corner?
p.3 #6 · Should M-to-E adapters be 9.80mm or 9.74mm?
cyberstudio wrote:
...or make it 0.1mm too thin and then it won't be fit for critical optical bench testing...
This only apply to lenses with floating elements. Any other lens will do fine if you use live view and not the scale on the lens. There is zero difference between changing the adapter thickness and moving the optical cell with the helicoid.
p.3 #7 · Should M-to-E adapters be 9.80mm or 9.74mm?
Makten wrote:
I disagree, since I've owned a few M lenses that did not reach optically infinity focus on M cameras. The adapter manufacturers have to make the adapters too short because some lenses are off.
If an M-mount lens cannot reach infinity when mounted on an M-mount camera, I would say there is a problem with the lens.
imho, it makes no more sense for all adapters to be intentionally short to compensate for such unintentionally long lenses, than it would make sense for the M-mount camera itself to be intentionally short to accommodate them.
However, if unintentionally long lenses are common, then I can see a case for two types of adapters: 1) technically correct adapters that are machined to the technically correct width, and 2) 'corrective' adapters that are machined short to help with such troublesome lenses.
But in general: just because some lenses are built wrong doesn't seem like a reason to build all adapters wrong.
p.3 #8 · Should M-to-E adapters be 9.80mm or 9.74mm?
fwiw, I just got a not-helpful reply from Metabones: "Sorry! I can’t give this information to you…..You can find on internet yourself." (It's certainly not posted on their own site.)
So far I have mostly bad info from the manufacturers:
Viogtlander: "The exact width is 9.5mm" -- but that shouldn't be right.
Novoflex: "9.74mm" -- at least a reasonable answer, even if it's short of the technically correct 9.80mm.
Metabones: "I can't give this information to you" -- why not?
Fotodiox: (replied with the adapter diameter, and I'm hoping to hear back from them)
I was trying to get the "manufacturer specs" so I could then measure the adapters and compare to their specs -- but it seems like I won't be getting good specs.
I'm still planning on getting a micrometer and a few adapters from B&H when they're back in stock.
I'm surprised that adapter width -- even from the premium manufacturers -- turns out to be so inconsistent.
