Test results: which of 21 Leica M lenses will cover the Hasselblad X or Fuji GFX sensor?

[onasj]

Here are results of testing 21 Leica M lenses for their ability to cover the medium format 44x33 mm sensor of a Hasselblad X camera. I used a CFV-ii-50c back set to electronic shutter, a 907x, and a Novoflex LEM to HAX adapter, but I assume that these results will hold for other camera systems that also use the Sony IMX161 50 MP 44x33 mm sensor, including both Fuji GFX 50 cameras and both Hasselblad X1D cameras.

In general, don't expect that any Leica M lens will perfectly cover an entire medium format sensor with no vignetting—even the "cropped" 44x33 mm sensor (most of these lenses don't even cover a 36x24 mm sensor without at least some vignetting). Some lenses come very close to covering a 44x33 sensor though, and are more than usable to produce excellent images. As a general rule, you will see less vignetting at close focusing distances than at infinity, and you will see less vignetting at moderate apertures (around f/2.8, typically) than at very large apertures or very small apertures.

"NO" means dark corners are present at any focusing distance or any aperture—that is, the lens's image circle does not cover the entire sensor. You could still use these lenses, but you'll have to crop the black regions out during post.

"YES" means there might be some vignetting or, less common, some purple color shift in the corners, but you can get the subject image to cover the entire sensor. I provide notes to explain each YES when appropriate.

The version of each of the 21 lenses below is the *most recently produced version*, even though some (like the Stemar, the 50/2.8, and the 75/1.4) are no longer in production. The results may differ if you use an older version of the same lens since the optical formulae of many Leica M lenses changed as new versions were produced.



WATE at 16 or 18 mm: NO

*WATE at 21 mm: YES, some vignetting

21/3.4: NO

24/3.8: NO

28/1.4: NO

28/2: YES, but with strong vignetting especially at or near infinity, and at wide apertures

28/2.8: YES, but with strong vignetting and purple corners

Stemar: YES, but the divider produces some artifacts at top of the frame. Still, it's very cool to generate 3-D stereo image pairs using this old Leica lens and a current medium-format sensor!

MATE at 28: NO

MATE at 50: YES, but only at near distances

MATE at 35: YES, but with strong vignetting at the very corners

35/1.4: YES at closer distances or apertures around f/2.8, with some color shift

*35/2: YES, with some vignetting

50/0.95: NO

*50/1.4: YES, with only slight vignetting at infinity

50/2 APO: YES, but only between f/3.5-6.3

*50/2.8: YES, a little vignetting wide open at infinity

*75/1.25: YES, slight vignetting at infinity

75/1.4: YES, but only around f/2.8 and at close distances

*75/2 APO: YES, almost perfect until f/8, only slight vignetting

*90/2 APO: YES, with only slight vignetting

90/2.4: YES, but only at mid/near distances

90/4: NO

135/3.4 APO: NO

Overall, the lenses I marked with an asterisk are the most useable: the WATE at 21 mm, the 35/2, 50/1.4, 50/2.8, 75/1.25, 75 APO, and 90 APO. Which is really a not a bad outfit if your goal is to shoot your small Leica M lenses on a cropped medium format sensor at a range of focal lengths.

 

[FloatingLens]

Thanks a lot, onasj!

I am wondering how some of the other WA options esp. Zeiss Biogons perform when using 24x36 crop on the Hasselblad sensor. Will they yield visible smearing or color shifts even when they are used within their nominal image circle?

I am looking for a manual focus/aperture lens in the 21-28mm range to produce some crisp landscape details - cropping is OK. Any supplementary experience with C-V lenses in that range?

 

[Paratom]

Thank you, this is very interesting. I wonder if the disadvantage due to vignetting and color cast (and corner smearing?) would decrease IQ of the Hassy sensor in a way, that the same M lens with a M10 (or M10r) would overall lead to better IQ? What I am trying to ask: for anybody, with both a M10 and x1d or 907, does it make sense to use any M lens on the Hassy in stead of using it on the M10(r)?

 

[Paratom]

I own the Adapter for using X-PAN lenses on the x1d....but except a short test I dont use it so far.

 

[Satrycon]

ive been using my 50 summilux and 90 APO on my GFX50R for 2 years +
id like to add :
The adapter will make a difference to the kind of vignetting if any, i use novoflex for the M lenses
i recently tried a k&F adapter for leica R lenses, and my 180 Elmarit R works perfectly

"Soft Vignetting" can be killed in LR or C1 very easily,
"Hard Vignetting" = Crop

 

[onasj]

Thank you, this is very interesting. I wonder if the disadvantage due to vignetting and color cast (and corner smearing?) would decrease IQ of the Hassy sensor in a way, that the same M lens with a M10 (or M10r) would overall lead to better IQ? What I am trying to ask: for anybody, with both a M10 and x1d or 907, does it make sense to use any M lens on the Hassy in stead of using it on the M10(r)?

If you don’t mind the post-processing required to clean up the corners of the images, AND if you don’t have an M10-R, AND if live-view focusing with electronic shutter (with rolling shutter artifacts) is acceptable, then my sense from these tests is that you are better off using your CFV or X1D than your M10, M10-P, M10-D, or older M for COLOR photography IF image quality is your main consideration. That’s a lot of ifs. If you have an M10-R or M10-M, I’m not sure you will see any significant image quality benefits from the X sensor and in some ways image quality may actually suffer. Of course if you are planning to crop to a 33x33 sensor equivalent, you’ll be able to take advantage of a 26% larger sensor area, which can offer its own advantages.

 

[onasj]

Thanks a lot, onasj!

I am wondering how some of the other WA options esp. Zeiss Biogons perform when using 24x36 crop on the Hasselblad sensor. Will they yield visible smearing or color shifts even when they are used within their nominal image circle?

I am looking for a manual focus/aperture lens in the 21-28mm range to produce some crisp landscape details - cropping is OK. Any supplementary experience with C-V lenses in that range?

Unfortunately I only have one wide CV (a 15/4.5 for Sony E mount, so not adaptable to X or M cameras since the flange distance for the E mount is smaller than that of the X or M mount) and no Zeiss Biogon lenses.

 

[Godfrey]

A good list. I add:

If you restrict the CFVII 50c with the square crop (33x33 mm format, 39 Mpixel) instead of its full frame, MOST of the Leica M and R lenses I've tested become perfectly acceptable to use. Some have good, tight lens hoods for the original format which cause vignetting still, remove the lens hood and they work fine. Some have non-removable lens hoods (example: the Voigtländer HyperWide 10mm f/5.6 Aspheric) and have hard vignetting on all four edges as a result ... For this lens in particular (and others like it) you have to crop to within the hard vignetting and then judge whether doing that nets any particular advantage to your photography. (In the case of this specific lens, the resulting FoV is still perfectly usable (if not a lot more than it is on FF 24x36 format) and a good 20-30 degrees wider on the diagonal compared to the shortest focal length available in native mount (XCD 21mm f/4).)

I have a lot of fun testing and then using oddball lenses on the 907x and seeing what I can get out of them. I've made what I consider to be some very satisfying photographs with lenses that had all kinds of problems on the 'big' format. But if I'm looking for corner to corner, edge to edge coverage at a high quality level, I stick to the XCD and V system lenses for best results. It just makes sense to me.

G

 

[ggibson]

I'm curious which other M mount Voigtlander lenses work well. I've seen examples of the new Voigtlander Nokton 40/1.2 and 50/1.2 in used on the 44x33 sensor with pretty much full coverage. I love the CV40 on my Sony, so the idea of using the 50/1.2 with a similar FOV on a larger sensor is interesting. The new 75/1.5 might work well. The biggest limitation IMO is the restriction to e-shutter and the slow readout on the Hasselblads in this mode. Fuji GFX generally seems better for adapting M lenses for this reason.

 

[Godfrey]

I'm curious which other M mount Voigtlander lenses work well. I've seen examples of the new Voigtlander Nokton 40/1.2 and 50/1.2 in used on the 44x33 sensor with pretty much full coverage. I love the CV40 on my Sony, so the idea of using the 50/1.2 with a similar FOV on a larger sensor is interesting. The new 75/1.5 might work well. The biggest limitation IMO is the restriction to e-shutter and the slow readout on the Hasselblads in this mode. Fuji GFX generally seems better for adapting M lenses for this reason.

Having a focal plane shutter definitely helps for some situations. That said, I suspect it depends a lot upon your subject matter and the kind of lighting you work under as to how successful adapting shutterless lenses becomes, and also whether you tend to be a tripod shooter or a hand-holder. I've not had much difficulty with the eshutter or the lighting in my condominium, or outdoors, as long as I wasn't moving the camera and the subject was relatively static. Of course, you're also limited to ambient light with the eshutter, flash synchronization is disabled, so there's that as well ... but the Fuji's focal plane shutter presents only a fairly slow sync speed too. Plusses and minuses...

G

 

[fotofreak]

I own the Adapter for using X-PAN lenses on the x1d....but except a short test I dont use it so far.

Have you ever checked performance of the XPan lenses, according to the "false" electronic shutter ? I'm curious about that because a 907X with adapter and XP-30 or 45 mm lenses would be a very handy digital SWC.
I own a Fuji GFX50 that I use with these lenses and results are simply marvelous. But this camera has a REAL electronic shutter. A 907x set would be a more compact equipment.

 

[Godfrey]

Have you ever checked performance of the XPan lenses, according to the "false" electronic shutter ? I'm curious about that because a 907X with adapter and XP-30 or 45 mm lenses would be a very handy digital SWC.
I own a Fuji GFX50 that I use with these lenses and results are simply marvelous. But this camera has a REAL electronic shutter. A 907x set would be a more compact equipment.

Actually, to achieve the format and field of view of a Hasselblad SWC image, you need to crop the 33x44 format to square (33x33) and use a 21mm lens. The 45mm focal length on the 907x's full format is indeed a wonderful field of view, but it's more equivalent to a 35mm lens in 24x36mm FF terms, not a 38mm lens on 6x6 format.

(This notion of a close equivalent to the SWC is the reason why when I ordered my 907x Special Edition, I ordered the XCD 21mm lens along with it.)

G

 

[kinglang]

Can you test R 35-70/4 ?

 

[Godfrey]

I can't because I don't have one. Perhaps the thread originator or someone else has one of these lenses they can test.

G

 

[scho]

Can you test R 35-70/4 ?

I've used this lens with my GFX 50R on a Cambo Actus and it was fine for close/macro work, but does not focus to infinity. I'll try it again when I get the 907x/50c.