This discussion of adapter precision got me to thinking about my own two
adapters so I went off to measure them. I was surprised by your comment
about the Martin Cheung adapter since that's the first adapter I bought
some time ago for about $100 (when prices were much higher than today)
My lenses fit on the Martin Cheung adapter about the same as they would
fit on an OM body. Once the lens clicks into place there is a little
rotational play of about 1/4 to 1/2mm but no other motion. My other
adapter is too tight. Once the lens has locked into place there is zero
motion. While all of the Zuikos fit it's so tight fitting that 3 of my
non-Zuiko lenses won't fit. The ears on their bayonet mounts are just
the slightest amount too thick to even get started. I have to reserve
the Cheung adapter to be able to use those lenses. One of them is the
90/2.5 Vivitar macro.
I just measured the flange thickness of both adapters at 6 points around
each. Nominally, they should measure 2.00 mm since the EOS register
distance is 44.00 mm and the OM register distance is 46.00 mm.
The Cheung adapter measures 2.01 mm +/- 0.01. The other adapter
measures 1.95 mm +/- 0.01. Take my tolerances with a grain of salt
since I don't know that I can actually measure that precisely with my
digital calipers. I have to translate the difference in flange
thickness between these two to inches to mentally comprehend it. 0.05 mm
is approximately 0.002 inches. Two thousandths is a fairly large
discrepancy from a machining standpoint but I don't know whether it
translates into anything meaningful from a focusing standpoint;
especially considering that the error is on the shorter side making the
lens focus very slightly past infinity.
But these differences are all trivial in comparison with the adapters
pictured at the link below. The ones pictured below have such a
discrepancy that the difference between them is quite visible with the
eye. It looks like it might be as much as 1-2 mm. Yes, whole
millimeters, not hundredths of millimeters. That has to be quite
deliberate but I don't know why.
Chuck Norcutt
Jeff Keller wrote:
> I think it is more involved than just field curvature. I posted this
> link quite awhile ago
> http://www.16-9.net/lens_tests/15mm_2/sigma1224vnikon15f.html
>
> Initial testing made the N*kon 15/3.5 look pretty poor on the C*non.
> By changing to a thicker adapter, the N*kon was found to do very well.
> Probably any lens which has a close focusing group that moves
> separately form the other elements could be very sensitive to the
> register distance.
>
> The four OM adapters that I've tried on my 5D seem to be very close to
> the same thickness but I'm seeing some repeatability problems. I've
> mounted a couple lenses that seem to have had infinity focus way off
> but when I come back a couple days later to test/compare I'm not
> seeing the origninal problem.
>
> For what it's worth the cheap $20 chinese adapter (using an easily
> visible screw to stop down the lens) and the cameraquest adapter seem
> to have very similar thickness. The mid-priced adapter I got from
> Martin Cheung(?) is junk. The lens rocks back and forth on the mount.
> A recent adapter advertised as a high quality Japanese adapter ($86)
> appears to be assembled wrong preventing the lens from mounting
> correctly but seems to be very close to the thickness of the $20
> adapter & the cameraquest adapter. The cameraquest adapter was loaned
> to me recently by Mike. I've only had it available to me for the last
> 24 hours so I need to do some more testing.
>
> -jeff
>
> On 9/26/06, Moose <olymoose@xxxxxxxxx> wrote:
>
>>Doesn't seem so odd to me. Non-macro lenses from the MF era will
>>generally have noticeable field curvature. And it becomes more of an
>>issue at closer focal distances. There were some sample images from a
>>21/3.5 posted here a few years ago that showed a great deal of it in the
>>21/3.5 used close-up.
>>
>>This is just the thing addressed by the Minolta 24 mm VFC lens discussed
>>here in the last few days.
>>
>>A practical solution is to use the matte area to check center and edge
>>focus of parts of the image that should be in focus and pick something
>>in between. That way, your DOF is used to best advantage.
>>
>>Moose
>
>
>
>>Paul Martinez wrote:
>>
>>>I've found the Zuiko 28/2.8 to be the worst of the Zuiko 28's. I've tried
>>>them all on my FF Kodak SLR/c and found the /2 and /3.5 to be much better.
>>>Also, becuase many Zuiko's are old and out of adjustment, no solid
>>>conclusion should be made from a single sample. A good performing Zuiko 28/2
>>>will blow away any zoom lens on a FF DSLR in the corners, even at f/8.
>>>
>>>Other things also need to be considered, like the adapter you used and how
>>>you focused. The 28mm primes (including the Zeiss and Leica's) are very
>>>picky about what adapter you use on them. A poorly matched adapter will give
>>>poor results. It is often not as critical on a zoom lens at 28mm because
>>>they are much steeper retro-focus design at 28mm. For focus stopped down,
>>>leaning much more than you would think resonable to the infinity mark will
>>>give much better corner performance and DOF. Odd, but true.
>
>
> ==============================================
> List usage info: http://www.zuikoholic.com
> List nannies: olympusadmin@xxxxxxxxxx
> ==============================================
>
>
==============================================
List usage info: http://www.zuikoholic.com
List nannies: olympusadmin@xxxxxxxxxx
==============================================
|