M lenses covering 44x33mm digital sensor

[Einst_Stein]

Advertisement (gone after registration)

Is there any list of M lenses that can cover a 44x33mm digital sensor?

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[Joakim]

No, you need to compiles one yourself from different sources and also people seems to have different opinions on what "cover" means in this context. Is it only for close focus or all the way to infinity, how much vignetting is ok etc.

Anyway some places I know are this thread at fredmiranda.com but it is for all kind of lenses and then there is this article by Ming Thein where he has tested some M lenses on the Hasselblad X1D. Maybe others can contribute with more sources.

Edited September 5, 2020 by Joakim

[UliWer]

For a rule of thumb you may assume long focal lengthes covering a larger format than short ones. 135mm was originally used for large formats, so they should cover 44*33mm, though show more vignetting than you are used to from the 24*36mm format.

[a.noctilux]

Documented in one of Marco Cavina article, definitely Tele-Elmar 4/135 can cover much more than 33x44.

this is link to his article in italian

[Einst_Stein]

I checked the MTF curve between 18mm~22mm. I assume if the curve shows significant degradation from 18mm to 22mm, it means the performance on 44x33mm would be unusable, if the curve degrades gradually or even turns up, the performance on 44x33 would be acceptable (w.r.t 36x24mm sensor).

Here performance means resolution and vignetting. If resolution chart is good bu vignetting no good, it could be fixed by post processing. 

Attached link shows the MTF of several Leica M lenses.   

https://www.overgaard.dk/pdf/Leica-M-Lenses-Their-Soul-and-Secrets_en.pdf

Using the above criteria, the following lenses are good for 44x33mm sensor.

1: 135mm f4

2: 90mm f2. APO;  90mm f2.8

3: 85mm f1.5

4: 75mm f1.4

5: 50mm f1.2; 50mm f1.4; 50mm f2(previous version); 50mm f2.8 Elmar (not Elmar-M)

6: 35mm f1.4 Summulux-M ASPH (the new version, not the previos)

7: 28mm f2 ASPH

8: 21mm f2.8 ASPH

I also checked R 35-70mm f4 and 80-200mm f4. 80-200mm f4 is really good, but 35-70mm f4 is only good for 50mm and longer.

[chasdfg]

7 hours ago, Einst_Stein said:

I checked the MTF curve between 18mm~22mm. I assume if the curve shows significant degradation from 18mm to 22mm, it means the performance on 44x33mm would be unusable, if the curve degrades gradually or even turns up, the performance on 44x33 would be acceptable (w.r.t 36x24mm sensor).

Here performance means resolution and vignetting. If resolution chart is good bu vignetting no good, it could be fixed by post processing. 

Attached link shows the MTF of several Leica M lenses.   

https://www.overgaard.dk/pdf/Leica-M-Lenses-Their-Soul-and-Secrets_en.pdf

Using the above criteria, the following lenses are good for 44x33mm sensor.

1: 135mm f4

2: 90mm f2. APO;  90mm f2.8

3: 85mm f1.5

4: 75mm f1.4

5: 50mm f1.2; 50mm f1.4; 50mm f2(previous version); 50mm f2.8 Elmar (not Elmar-M)

6: 35mm f1.4 Summulux-M ASPH (the new version, not the previos)

7: 28mm f2 ASPH

8: 21mm f2.8 ASPH

I also checked R 35-70mm f4 and 80-200mm f4. 80-200mm f4 is really good, but 35-70mm f4 is only good for 50mm and longer.

Speaking of 75lux, I think the 80lux covers the GFX/X1D.

I'm skeptical that the 75lux, 35lux FLE and 28cron cover the sensor adequately to avoid hard vignette. Never tried though but looked into it before.

[Einst_Stein]

Advertisement (gone after registration)

1 hour ago, chasdfg said:

Speaking of 75lux, I think the 80lux covers the GFX/X1D.

I'm skeptical that the 75lux, 35lux FLE and 28cron cover the sensor adequately to avoid hard vignette. Never tried though but looked into it before.

You can compare the MTF on page 63 and 64. 

[Joakim]

11 hours ago, Einst_Stein said:

I checked the MTF curve between 18mm~22mm. I assume if the curve shows significant degradation from 18mm to 22mm, it means the performance on 44x33mm would be unusable, if the curve degrades gradually or even turns up, the performance on 44x33 would be acceptable (w.r.t 36x24mm sensor).

Here performance means resolution and vignetting. If resolution chart is good bu vignetting no good, it could be fixed by post processing. 

Attached link shows the MTF of several Leica M lenses.   

https://www.overgaard.dk/pdf/Leica-M-Lenses-Their-Soul-and-Secrets_en.pdf

Using the above criteria, the following lenses are good for 44x33mm sensor.

1: 135mm f4

2: 90mm f2. APO;  90mm f2.8

3: 85mm f1.5

4: 75mm f1.4

5: 50mm f1.2; 50mm f1.4; 50mm f2(previous version); 50mm f2.8 Elmar (not Elmar-M)

6: 35mm f1.4 Summulux-M ASPH (the new version, not the previos)

7: 28mm f2 ASPH

8: 21mm f2.8 ASPH

I also checked R 35-70mm f4 and 80-200mm f4. 80-200mm f4 is really good, but 35-70mm f4 is only good for 50mm and longer.

But this doesn't tell us if the image circle of the lens covers the whole sensor or not and nothing about smearing in the corners or field curvature.

Edited September 10, 2020 by Joakim

[Einst_Stein]

9 hours ago, Joakim said:

But this doesn't tell us if the image circle of the lens covers the whole sensor or not and nothing about smearing in the corners or field curvature.

See the vignetting curve and the resolution curve. 

However,, this is only part of the reference information. Take it as you like. The final judge is still your eyes.

[Joakim]

True that the eyes are the final judge because I think in most cases that is the only way to determine if a lens can be used on a larger sensor and specially M lenses. I have actually never heard anyone use MTF curves for determining if a lens can be used on a larger sensor than intended before and I think that it says a lot of the usability of them for this.

[Einst_Stein]

50 minutes ago, Joakim said:

True that the eyes are the final judge because I think in most cases that is the only way to determine if a lens can be used on a larger sensor and specially M lenses. I have actually never heard anyone use MTF curves for determining if a lens can be used on a larger sensor than intended before and I think that it says a lot of the usability of them for this.

Now you've heard it.

Yes, a lot of people don't believe MTF at all, those are usually anti-scientifics, although there are some points in it. MTF is the only objective tool as far as I know, take it as what it is.

On the other hand, MTF does not tell you the camera/sensor's response to the lens. This is widely know since the digital camera. I had a Sony A7RII, which shows mediocre IQ from many excellent Leica lenses (mostly 35mm or wider), it's day and night compared to Leica M9 and M240. So, I see MTF is the necessary criteria for what I want, not a sufficient criteria. 

My take, whenever there is a doubt about a lens, the first thing is to check MTF, but don't stop there. The last thing I trust is the professional "lens reviewer". I assume you know why!   

[Joakim]

I have no issues with MTF curves and their usage, I just don't see how relevant they are in this context and why they in any way relate to your original question. 

[pgk]

15 hours ago, Einst_Stein said:

Yes, a lot of people don't believe MTF at all, those are usually anti-scientifics, although there are some points in it. MTF is the only objective tool as far as I know, take it as what it is.

But only when used appropriately. Image circles can be physically curtailed so regardless of how valid the data supplied is for 24x36mm it can't be extrapolated for larger coverage. You will have to try each lens to see if coverage is sufficient and viable I'm afraid.

[Einst_Stein]

4 hours ago, pgk said:

But only when used appropriately. Image circles can be physically curtailed so regardless of how valid the data supplied is for 24x36mm it can't be extrapolated for larger coverage. You will have to try each lens to see if coverage is sufficient and viable I'm afraid.

The Leica lens MTF covers 22mm. Take a look of the curves, use your common sense. 

[pgk]

Its a theoretical curve, not a measured one. The physical build of a lens can prevent the image circle being as great as it could be.

[CharlesL]

The Summicron 50 version v delivers a strong corner vignette. The Summilux 50 aspherical vignette is slightly smaller and much lighter.

If you want to use an M lens on a Fuji GFX or Hasselblad X1D camera - because you have the lens, because of the way it renders, because it is small - then work around the vignette. Expect to crop; the camera has a setting for the camera JPG and the viewfinder that helps you compose, while the raw file is the full sensor image for you to crop as you see fit for the particular shot. And when you crop, you get back closer to the field of view that the lens has on an M camera.

This was taken with a Summilux 35 pre-aspherical at f/11. As often happens, I left a little bit of the vignette.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[adan]

The things with MTF charts are:

1) they only measure out to 22mm because the 24x36 frame only extends to 22mm from the center (on the diagonal, to the corners). That doesn't necessarily mean the image circle is always limited to 22mm, as demonstrated in the nearby thread (or the post above) testing M lens coverage on a 33x44 sensor. But on an M film or digital body, anything beyond 22mm is irrelevant. Thus Leica's MTF charts don't bother looking further.

2) I take MTF charts with a grain of salt, although they are suggestive and indeed are at least a decent sketch of how a lens behaves. And one can correlate some things in the chart curves to actual lens behavior, and vice versa. However...

- MTF charts (at the moment, for M lenses) only measure down to 40 lpmm - I'm interested in performance at higher resolutions (up to at least 80 lpmm sometimes)

- MTF charts tend to overemphasize contrast, whereas, again, I look for raw resolution (the pickets in the fence, or whether 1cm letters on a sign are readable in a picture from 40m away). Some lenses have much better resolution than their MTF charts suggest (and some have less).

In other words, I'd prefer a lens that images as on the left in the sample below (low contrast, high resolution, low MTF) compared to the sample on the right (high contrast, lower resolution, higher MTF). Which is one reason I prefer, say, the Mandler 90mm Tele-Elmarit II to  - well, I won't name names, but some lenses with higher charted MTF curves.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

Because (below) I can fix the low contrast better than I can "sharpen" the lower resolution, especially with digital, and get a cleaner reproduction of the most delicate details. Within limits, of course.

MTF is more useful for film, where the clarity and contrast of the projected lens image has to "burn through" the blurring effects of gelatin and grain. As well as reducing lens performance to a single number, which is easier for figuring system resolutions (MTF lens x MTF film x MTF enlarger lens x MTF printing paper, for example).

And there are lenses that have both high resolution and high contrast.

And as Zeiss demonstrated way back when - most people perceive a high-contrast rendering (50%+) of a moderate resolution (40lpmm) at a moderate enlargement (A4/8x10, or on the Web) as "looking sharper" than a low-contrast rendering of a higher resolution.

But I ain't most people.

[Einst_Stein]

I don't buy the argument of the MTF curve being measured vs. calculated. They are closely correlated anyway. The M Pdf file mentioned how the MTF is measured, you can choose not to believe it.

I also do not argue with people of anti-scientifics. Sorry. You win.

Back to MTF, it's a graphic of engineering data. How to interpret it to match your human eye expectation has a lot of to do with your personal judge. Not to mention the  effects of the complete image forming paths, such as the sensor's performance and the post-processing. 

Enough theoretical talks, I think only the real image is the final judge. unfortunately I don't have a 44x33mm camera. I will see if I can borrow one.

Any real pictures are welcome! 

Edited September 12, 2020 by Einst_Stein

[Advertisement]

Hallo,

Look here: Leica M

[Einst_Stein]

26 minutes ago, CharlesL said:

The Summicron 50 version v delivers a strong corner vignette. The Summilux 50 aspherical vignette is slightly smaller and much lighter.

If you want to use an M lens on a Fuji GFX or Hasselblad X1D camera - because you have the lens, because of the way it renders, because it is small - then work around the vignette. Expect to crop; the camera has a setting for the camera JPG and the viewfinder that helps you compose, while the raw file is the full sensor image for you to crop as you see fit for the particular shot. And when you crop, you get back closer to the field of view that the lens has on an M camera.

This was taken with a Summilux 35 pre-aspherical at f/11. As often happens, I left a little bit of the vignette.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

The MTF do indicate the ASPH versions are generally better than the older version. Surprisingly, 35mm and 50mm seems to be worst than the 28mm and 90mm/135mm. (See the MTF curve in the Leica's document).

I think if it's just vignetting, it should be easier (relatively) to correct than the resolution degradation at the corner.  

The bottom line, a 35mm format lens would be hard to compete with the lens targeting the 44x33mm sensor. I am only interested if it is good enough or practically useful.

[pgk]

4 hours ago, Einst_Stein said:

I don't buy the argument of the MTF curve being measured vs. calculated. They are closely correlated anyway. The M Pdf file mentioned how the MTF is measured, you can choose not to believe it.

I also do not argue with people of anti-scientifics. Sorry. You win.

Back to MTF, it's a graphic of engineering data. How to interpret it to match your human eye expectation has a lot of to do with your personal judge. Not to mention the  effects of the complete image forming paths, such as the sensor's performance and the post-processing. 

Enough theoretical talks, I think only the real image is the final judge. unfortunately I don't have a 44x33mm camera. I will see if I can borrow one.

Any real pictures are welcome! 

I have a friend who has just calculated theoretical MTF data for an 1857 photographic lens for me using his design software - he is a lens designer by profession. He has only ever seen photos of the lens that this data is for and is working from the original (loose) patent information. MTF data is usually used in optimising designs for the intended use of the item being designed. It tells you a lot but only when used in conjunction with a myriad of other pieces of information which the software also generates (image curvature, distortion, etc.). MTF data IS largely about contrast but needs interpreting within the confines of the design requirements. The problem with general purpose photographic lenses (which is the type we are discussing here) is that they have subjectively analysed output. However as I have repeatedly stated, an M lens may not cover a large sensor simply because of the mechanical design of the housing into which the lens is placed because there is no need for it to do so and there may be the need for baffles and/or physical protection of the rear element. This is not rocket science its basic mechanics. Testing will quickly establish which lenses are usable acceptably and which aren't, however the M lenses (or at least modern ones) will be optimised to ensure good coverage of the 24X36mm format at infinity as there is simply no perceived need for good performance beyond this and so it won't be part of the design requirements.

FWIW I have a number of old large format lenses. Some state their coverage, others have to be looked up. Some appear to cover far more than the stated are but actually don't, in that the performance beyond it quickly falls off. Some have abrupt, physically curtailed image circles just outside their coverage. Obviously you need to choose a lens for its intended purpose so a larger, high performance image circle is needed if you are using movements. I have several old lenses with variable front (aperture) stop positioning. This enables the diameter of the image circle the lens produces to be varied, at the expense of performance, and produced a very versatile lens with true wide-angle capability, which was very valuable as it was unusual in its day (larger plates were needed though). The images produced were intended for contact printing only and the manufacturer was very conservative with suggested coverage, although photographers ignored this and worked with much bigger plates in practice.

Edited September 12, 2020 by pgk