Fuji X-Pro 1 grand test with Leica M Mount lenses

**Update** To rule out any mechanical faults with the Kipon adapter I plan on trying another adapter but may wait until the official Fuji one comes out. In the meantime, if you have tried an adapter with leica fit lenses wide open with any more success than I have had, please do drop me a line. Properly functioning Leica glass on this sensor would be fantastic.

 

**Update 2** For Geeks, the answer probably resides with the following:

1. Zeiss Lens document (Page 11 and 12)

2. Thickness of filter covering sensor:

  •      Leica M8 = 0.5mm
  •      Leica M9 = 0.8mm
  •      Fuji X-Pro 1 = 2.5mm

———————————————————————-

I really don’t like lens testing as it saps away what little creative energy I have and someone, somewhere is always unsatisfied with the testing methodology.
Also modern lenses are so good that any nitpicking under high resolution is, frankly, time you should be out shooting.
Nonetheless, I was so suprised at the results from my Leica fit lenses on my Fuji X-Pro (using the Kipon adapter) that I decided it was worth an hour of my life to delve further.
So here are all my Leica m-mount lenses (28 ASPH, 35 & 50 Summilux, Zeiss 21mm f2.8, Voigtlander 15mm and 35mm color-skopar) together with the Fuji 35mm 1.4 that came with the camera tested on a Fuji X-Pro 1, each at their maximum aperture and then f8. The Leica 28, 35 and 50 are the ASPH versions (the 35mm is not the most recent version)
Focus point for all was on the black pipe directly below the whiteish brick. Click through twice to get the full size picture.
The only aim of these tests was to examine edge resolution as this is often where these lenses fall down when used on digital non-Leica cameras. From experience, colour variation in the corners is not really an issue even with the 15mm wide lenses
All pictures were taken in a 20 minute timeframe, shot in RAW at c. 200 iso and converted in Silkypix with all defaults but with noise reduction turned off. I have no idea whether these results would be replicated with the official Fuji or other 3rd part adapters.
To rule out any difference between Raws and Jpegs I have included one Jpg from the Zeiss 21mm wide open – there appears to be no difference.
The M-mount adapter is a very tight fit with the lenses and if you’re regularly changing M Mount lenses I foresee you could damage the camera lens mount. Therefore, I’d suggest taking off the lens+adapter from the camera before separating the lens and adapter (which needs a very firm twist).
The Voigtlander 15mm is the first non-coupled rangefinder version and for some reason the Zeiss 21mm would not focus at infinity (irrelevant to these tests).
I’m finding that the faster lenses like the 50mm Summilux can be focused without zooming if used wide open as the shallow depth of field is replicated in the electronic viewfinder which allows the properly focused area to “pop”
There are a number of possible reasons for the performance of these high quality lenses as shown below:
1. The Fuji X-Pro 1 is just not a good optical match with M lenses.
2. My M lens adapter in some way is faulty.
3. Fuji know a trick or 2 for their own adapter which will increase the quality of results.
Unfortunately, at this time I have no real idea which of the above is the correct explanation for these results.
There is one simple way to rule out any errors in these tests – someone just needs to put an adapter on their X-Pro 1, shoot their Leica lens wide open and get sharp edges. I would be delighted !

Voigtlander 15mm wide open at f4.5

 

Voigtlander 15mm at f8.0

Carl Zeiss 21mm Biogon wide open at f2.8

Carl Zeiss 21mm Biogon at f8.0

Leica 28mm Summicron wide open at f2.0

Leica 28mm Summicron at f8

Leica 35mm Summilux wide open at f1.4

Leica 35mm Summilux at f8

Fuji 35mm wide open at f1.4

Fuji 35mm at f8.0

Voigtlander Color-Skoppar 35mm wide open at f2.5

Voigtlander Color-Skopar 35mm at f8

Leica 50mm Summilux wide open at f1.4

Leica 50mm Summilux at f8.0

Comments

  1. Hi,

    thank you for the work, very interesting! I will buy a m adaptor too, but at the moment i only find a Kipon and don’t know, if this is a good one.

  2. Tabrahm Shoonie says:

    Can you use a known good lens without soft edges such as a summilux or summicron? provided the adapter keeps the lens exactly parallel to the sensor, the adapter does not matter (you could hold it at the correct length

  3. Cyrus says:

    Hi Tabrahm
    Even the 28 Summicron, 35mm Summilux and 50 Summilux I tested were smeared to a greater or lesser extent when wide open.
    Having said that none are really bad wide open (apart from the Zeiss 21mm which is BAD) but when you’ve paid thousands of pounds for a lens you expect them to be better.
    All the lenses I tested were much much better at the edges on the Ricoh M adapter for the GXR and just about perfect on an M8 or M9.
    Having said that, shooting any of these lenses at f8 on the street will be fine.

  4. Cyrus says:

    The GXR ! It is a really excellent camera, and before the X-Pro 1 arrived it was by far the best small digital camera I had used.

  5. Gus says:

    http://www.flickr.com/photos/michaelhd/page2/

    could be helpful ???

    I still await my M adapter if I have any results that are different I will let you know,

    Gus

  6. J Spears says:

    Thanks for the work. I received the same adaptor and it feels close to the same fit as the two Fuji lenses I have. I have generally the same result with five Leica or Cosina lenses: Lower contrast in JPEG than the Fuji, and none of the Leica “pop”. I tried the 12mm thru a 135mm both wide open and stopped down, and my guess is that there is internal software correction for the Fuji lenses, although Fuji claims the 35mm is uncorrected and the wide f.2 lens is corrected.

  7. Timo Schwach says:

    Hi Cyrus

    I received my Fuji XPro1 a few days ago and made an identical test with the Voigtländer 15/4.5 and the Zeiss ZM 21/2.8 with the Kiopon Adapter and my results coincide with yours! Shot wide open these 2 lenses are unusable.

    I wrote a support inquiry to Fuji Switzerland and hope I will get an answer from them regarding the original Fuji adapter and if they think the performance will be fine with it and not show these problems.

  8. Peter says:

    Just ended my first trial with the X Pro 1 and the Kipon adapter with the same kind of disappointment. Lenses used were the Leica 16-18-21 Tri-Elmar and the Voigtlander 35mm f/1.2 Nokton Asph II. Very smeared corners unless stopped down to F8. In fact, while far from being perfect, the Tri-Elmar, which is not a retro-focus design, even performs slightly better on the NEX-7 (there is no lens cast problem as there is with the Zeiss ZM Biogon 21). The difference is even more pronounced for the Nokton, which is a great performer on the NEX-7. I had ordered both a Metabone and a Kipon adapter for the Fuji. Kipon arrived first. If there’s any improvement with Metabone, I will let you know. If the filter thickness is indeed the problem, than the adapter is of course irrelevant and Fuji’s proprietary one might cause the same issues.

  9. Peter says:

    Quick update: maybe part of the blame has to go to Kipon. Today I started experimenting with the four screws, varying the degree to which they were tightened. You will realize that there is a little play between the two rings of the adapter as you slightly loosen the screws, just enough to micro influence it’s behaviour. While it is hopeless to get it perfect in every aspect, I was able to significantly improve the results. The Tri-Elmar is now yielding very good results at f/5.6 while the Nokton has good edges now at f/2.8. Again, not perfect, but it leaves hope the a high precision adapter might achieve better results. Will keep you posted once Metabones’ arrives.

    • Cyrus says:

      Excellent detective work, Peter. I had a Minolta adapter for the Ricoh GXR M which was loose on arrival – tightening all the screws as far as I could put that adapter back in spec – is your Kipon adapter worse with all screws as tight as they will go ? (as I assume this is the default position)

      • Peter says:

        What got me started on this fine tuning quest was that in 2010 the very first adapter that was available for the NEX was a Kipon as well. Results were not okay, and – inline with your discovery – they improved once I had tightened all the screws. With the Fuji adapter, fully tightening all screws, though, negatively affects the c. 25% of the image to the right, actually across the frame, with the corners of course suffering the most. This is why in my case results improve if the screws are slightly loosened. Unfortunately my Metabone adapter did not make it for delivery this weekend and still is with DHL, so I have to wait.

  10. Peter says:

    Quick update since I just received the Metabones adapter. Not much of a progress, though. I tested it again with my Zeiss Biogon 21/4.5 ZM (which actually delivered slightly better results than the Tri-Elmar with the adjusted Kipon adapter). This lens is legendary since it is virtually free of any distortion. The results are largely consistent with the Kipon adapter. The thickness of the filter of the X Pro 1 might indeed be the cause for the problem. If so, I would be slightly surprised if Fuji could pull a trick or two when releasing their own adapter…

  11. H R says:

    I can assure you that the M adapter from Fuji will not work any better, or any other adapter for that matter.

    With the sensor design there is no way of avoiding smear with wide angle RF glass, unfortunately. If you check out Fuji’s own literature on the development of the X Pro1 you will see that they mention that they have designed and calculated the wide angle lenses (namely the 18/2) to cater for the sensor – and not the other way around. They haven’t even managed to get their own 18/2 to work perfectly with the sensor.

    I expect the X Pro 1 to work really well with normal and telephoto lenses as well as retrofocus SLR lenses, but not with wide angle RF/legacy glass. Sad, but true.

  12. H R says:

    I need to add that they probably released the M adapter to broaden the customer base, but that they did not devote that much of R&D into making it a suitable platform for all M lenses and given the sensor they have done the best they can. They can’t change the basic physics of their sensor layout to fit the wide M lenses
    and they cannot software compensate for it either.

    12>35 mm lenses: poor to fair. 35<200+ mm excellent.

    The current best APS-C platform for M lenses is still the GXR A12 M mount. Not the best sensor, not the best high ISO performance, but the best (as in the least problems) when shooting wider than 25mm RF glass.

  13. Calvin says:

    How does focusing a Leica M glass on the XP1 feels? Do you use the EVF or OVF function in the view finder?

    I was thinking of using the 35mm and xp1 combo for streets until the m10 comes out and the m9 price drops

    Your comments are greatly appreciated! Thank you!

  14. Steven says:

    I have an Olympus OMD.
    I am using the Voigtlander 15mm Heliar on it. A wonderful lens. But it is soft in the corners at all apertures.
    It gets better as you stop down but is still soft even at F8.
    Is this a product of using this full frame lens on a M4/3 sensor?
    (I would have expected it to be sharper in the corners because of the sensor crop factor only using the center of the lens.)
    But I have read some posts about small sensors causing corner blurring with adapted full frame wide angle legacy lenses.
    Why does this happen?

    Thank you,

    Steven

  15. Steven says:

    Here is some info I just came by about the lens blur issue around the periphery when using wide angle lenses on the OMD sensor and Nex Sensors.
    I talked to the guys at LensRentals.com.
    I explained my issue with the Voigtlander Super-Wide 15mm blurring around the periphery and they said they know all about it.
    They said this issue of edge blur is a problem that occurs when you try to use Wide angle “Leica or Leica mount” lenses on M4/3 sensors or the Sony Nex sensors.
    Here is the cause: These Sensors recess their photosites into little wells.
    In essence each photosite has a lens hood to block stray light from hitting it.
    This increases contrast and light concentration by only allowing direct light to hit the each photosite.
    The problem is the Lecia lenses have such a sharp angle of projection that the light is hitting the sensor at such a sharp angle on the sides of the sensor that the light is blocked by the photosite wells/hoods and cannot hit the photosite directly.
    This causes the blur effect seen in the OMD and the magenta color shift and heavy vignetting seen on the Sony Nex cameras.
    Stopping down cannot not correct this problem, it can only help a little.
    Wide Angle lenses designed for SLR should not show this problem as they project light at a more straight angle as they mount further away from the sensor.
    So sadly this is a limitation of the Sony Sensors that are used in the Nex and OMD. They are not compatible with “Wide Angle Leica mount” lenses.
    This Problem is gone with Leica lenses of 35mm and above on the OMD as the angle is not as sharp due to the longer focal length producing a less severe focus angle on the sensor.
    As an example my Zeiss 35mm Biogon does not show this problem.
    But a real bummer over all though for those wanting to use the great Leica Wide angle lenses on OMD and Nex.

  16. Ivan says:

    Test my Biogon 25 with Nex-5N(novoflex adaptor) and X-pro1 (official Fuji M-adaptor) .
    Biogon 25 is fine on Nex-5N (there are no colorshift and there are no corners smearing). Biogon 25 is bad on X-pro1 – corners smearing problem :(

  17. Clint Dunn says:

    I too tried the Zeiss 25mm Biogon with the Fuji M adapter and the outer third of the frame is outright soft…unusable. I also have a 50mm Summicron and it works pretty good with much better corners.

  18. Brian Rowland says:

    This has been a very informative thread as I own a few M lenses and seriously considering purchasing the XPro1.
    Frankly I’m a bit disappointed that a camera with so much potential has not been successful with wider M mount lenses. From the Fuj marketing literature they clearly recognise that many people own good M glass and actively promote using m mount lenses. But it seems there is a sensor compatibility issue that limits this application. Frustratingly other brands such as Sony and Olympus and even Ricoh seem more compatible. with M lenses. I hink Fuji needs to address this problem ……. Maybe an adaptor with a small amount of optical correction would help?
    It’s just as we’ll that the Fuj X lenses are excellent and in some cases as good as Zeiss and Leica

Speak Your Mind

*