40 & 28 combo, bit of a test

[Mike S]

Advertisement (gone after registration)

I'm planning a project which in my mind is going to need an M10, a 28 and a 40. I've recently picked up a nice 40mm Summciron-c, which i have been very impressed with. Traditionally I've been a 50mm type of chap but I do like a 40, I've had great times with the Voigtlander 40 1.2 on my Sonys and the little Ricoh GR3X has been a winner from day one. 

28 though, that's always been a struggle for me. but combined with a 40 it makes a lot of logical sense.

I've had the good fortune to borrow a 28 (Asph V1) from that deft exponent of the 28, my dear chum @pippy

I present my findings, with almost no scientific rigour.

Precis: Nice lens 

The detail: 28 has an edge at close and mid range, 40 might take it by a gnats at infinity.

Test conditions: Handheld M4-P for the wonky horizon (that's another conversation), UK version of sunny 16, 1/250 f/11, Fomapan400, EI250, developed in Adox XT-3 at 1+2 

And, I've got a theory... both lenses were running Leitz UV filters and hoods but 40mm is only single coated. In these conditions will that allow a little more clarity at infinity?

Conclusion: Happy with both lenses, now, who might lend me an M10?
 

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!

Edited August 21 by Mike S
Corrected @ tag

[Advertisement]

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

[pippy]

Silly question perhaps, Mike, but were both lenses set to infinity?

Philip.

[wda]

Hand-held or using a tripod ?

[Mike S]

38 minutes ago, pippy said:

Silly question perhaps, Mike, but were both lenses set to infinity?

Philip.

Good question. Both focussed on the girl’s bun. Seemed like a good hyperfocal point 

[Mike S]

10 minutes ago, wda said:

Hand-held or using a tripod ?

Handheld as noted, hence wonky horizon, but at least I’m consistent 😄

[Chris W]

As it is a test, to be harsh, the 28mm image has un-needed elements that the 40mm eliminates (successfully). More sky, edge of the breakwater, more bobbing heads in the water.

That's why a 28mm is harder to use IMO, if you want to make everything in the image count. 

[pippy]

Advertisement (gone after registration)

    

1 hour ago, Mike S said:

Good question. Both focussed on the girl’s bun. Seemed like a good hyperfocal point 

Initially I mis-read the 'n'.

OK.

I will have to have a look, myself, later as I've never noticed the lens to be lacking in sharpness at infinity. I wonder if there is some amount of diffraction going on here? The exit pupil being nearer the film plane and narrower elements in comparison to the 40 Summicron (etc...etc...) might be a factor.

I don't usually go down as far as f11.0 with a 28mm so perhaps the sharpness of the lens at the smaller apertures is something which will need to be examined further on down the line.

Over here I have the TTArtisan 28mm f5.6 with me so might do some small-aperture tests this afternoon just out of curiosity...

Philip.

[Mike S]

55 minutes ago, pippy said:

 

I will have to have a look, myself, later as I've never noticed the lens to be lacking in sharpness at infinity. I wonder if there is some amount of diffraction going on here? The exit pupil being nearer the film plane and narrower elements in comparison to the 40 Summicron (etc...etc...) might be a factor.

 

No lack of sharpness suggested, just that the 40 seemed a tiny bit “cleaner”, of course that could just be the 40 getting one closer to the action, as it were.

 

Diffraction is a thought well made, not that it isn’t in play here but generally less of a factor on film then on digital.

[Mike S]

2 hours ago, Chris W said:

As it is a test, to be harsh, the 28mm image has un-needed elements that the 40mm eliminates (successfully). More sky, edge of the breakwater, more bobbing heads in the water.

That's why a 28mm is harder to use IMO, if you want to make everything in the image count. 

Points well made perhaps, but not too harsh for Karsh as they say in some parts.

This wasn't really a test of compositional skills, but to your point I’m a whizz with a 21 and was known to whip out a Heliar 15 when I was in the mood. So composing with a wide has never been an issue other than with a 28, which is clearly a mental issue of my own that can join the queue for resolution. 

More work to be done at my end!

[pgk]

4 hours ago, Mike S said:

The detail: 28 has an edge at close and mid range, 40 might take it by a gnats at infinity.

Test conditions: Handheld M4-P for the wonky horizon (that's another conversation), UK version of sunny 16, 1/250 f/11, Fomapan400, EI250, developed in Adox XT-3 at 1+2 

And, I've got a theory... both lenses were running Leitz UV filters and hoods but 40mm is only single coated. In these conditions will that allow a little more clarity at infinity?

I would very much doubt that coatings would make any significant (visible) difference, especially using 400 film. I'd also doubt diffraction being a cause, again especially using 400 film. The 28mm Apsh v.1 seems to have had a good reputation, until the v.2 came out. I can't remember the difference, however at f/16 again I would be very surprised if you can tell them apart (I shoot a lot well stopped down and by f/16 most lenses work very well (even very old ones) and many are indistinguishable in practice). I'd suggest a few more 'test' shots to rule out straightforward variation due to something during shooting. I've always found the 40/2 to be a great little lens although I no longer have one.

Edited August 21 by pgk

[pippy]

3 hours ago, Mike S said:

...not too harsh for Karsh as they say in some parts...

...Diffraction is a thought well made, not that it isn’t in play here but generally less of a factor on film then on digital...

I do wonder if 'Big Malky' (Sp?) ever went into the business? Never warmed to the fellow. Having said that I do hope he's not here......😸......

3 hours ago, pgk said:

...I'd also doubt diffraction being a cause, again especially using 400 film...

Having just carried out a non-scientific test I also doubt (understatement) that diffraction is at work.

Took three 'contre-jour' frames on the M Monochrom with the TTArtisan 28mm f5.6 at f11.0, f16.0 and f22.0. Subject matter was a 'beach-scene' with my focus-group (Ho!Ho!) perhaps 15m away and with masts in a harbour and a flotilla of yachts out on the horizon all of which (checking using google maps FWIW) were roughly 800m to 1Km away.

No real point in posting all three as there was b'gger-all difference between them and, in any case, forum restrictions require a drastic down-grading of IQ but, FWIW (Pt. 2), here's the f16.0 pic of the general scene with (despite what it says in the dialogue box!) a 200% crop of central area. On-screen in 'real-life' the definition of the masts and sails is quite remarkable;

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!

Main lessons learned from today's mucking-about? Firstly; don't worry about possible IQ drop-off at tiny apertures and, secondly, clean that bloody sensor when I get home...

I would also like to try out the 40mm Summicron properly at some point. Not that I need a third 40mm lens......

Philip.

Edited August 21 by pippy

[Vanillasludge]

If the 40 behaves similarly to the somewhat contemporary v4 Summicron 35, it will have a fair amount of focus shift as it’s stopped down.  This is natural for non-asph high speed lenses and has the effect of stretching depth of focus.  
 

 

[Mike S]

On 8/22/2025 at 12:45 AM, Vanillasludge said:

If the 40 behaves similarly to the somewhat contemporary v4 Summicron 35, it will have a fair amount of focus shift as it’s stopped down.  This is natural for non-asph high speed lenses and has the effect of stretching depth of focus.  

Definitely renders similarly to the v4 35 Summicron I used to have. Aiui as DoF increases as one stops down any focus shift is mitigated, and certainly by f/11 there will be no issues. Also see @pippy above re: diffraction being a non event. Bottom line - they’re all good lenses 😉

I am interested in the “non-asph high speed lenses comment” are you suggesting with an aspheric lens DoF does not increase as you stop down or that they don’t suffer focus shift ? Assume the latter but I’m no authority 😄 though I have plenty of spherical lenses with no discernible focus shift on digital or film. 

[pippy]

On 8/22/2025 at 12:45 AM, Vanillasludge said:

...If the 40 behaves similarly to the somewhat contemporary v4 Summicron 35, it will have a fair amount of focus shift as it’s stopped down.  This is natural for non-asph high speed lenses and has the effect of stretching depth of focus...

 

1 hour ago, Mike S said:

...I am interested in the “non-asph high speed lenses comment” are you suggesting with an aspheric lens DoF does not increase as you stop down or that they don’t suffer focus shift ? Assume the latter but I’m no authority 😄 though I have plenty of spherical lenses with no discernible focus shift on digital or film...

I'll try to keep this brief...

A fair number of years ago, having read several comments concerning 'horrendous' focus-shift evident with the Voigtlander 40mm f1.4 Nokton, I undertook some carefully controlled studio tests with around 20 M-mount lenses. My interest was due, primarily, to my never once having noticed such a phenomenon with my own example and this had been my #1 lens ever since I bought my first digi-M (an M8.2) over a decade ago.

The bottom line? All the spherical lenses - ranging from 21mm to 135mm - exhibited some degree of focus-shift to a slightly greater or slightly lesser extent. Equally, in all cases, the focus-shift was only noticeable within a certain camera-to-subject-matter distance range and at certain apertures. Typically the distance would be approx MFD x 1.5 to 2 or 1.5 to 2.5 and the aperture range would be from Max.Ap to 1.5 to 2 stops below Max. Ap. depending on the lens.

In the case of the Nokton (IIRC) the lens started to back focus from c. 1.2m to 1.8m and the shift would go from c.1.5cm to 2.5cm when the lens was used at f1.4 and f2.0. By the time the camera-to-subject distance had gone beyond 1.8m and the aperture was stopped-down to f2.4 / f2.8 the phenomenon was no longer visible.

It is worth bearing in mind that this test was Pixel-Peeping with knobs on. In 'The Real World' absolutely none of this matters whatsoever for 99.999% of the stuff which I shoot so I don't give the matter a second's thought; happily I just get on with my snapping.

Things might well be different, of course, for other folks with different shooting preferences to myself.

Unsurprisingly the 35mm Summilux v2 shared almost exactly the same behaviour yet no-one seems fit to comment on that lens' 'Horrendous' focus-shift. I do wonder why this should be the case...

Philip.

[Vanillasludge]

2 hours ago, Mike S said:

Definitely renders similarly to the v4 35 Summicron I used to have. Aiui as DoF increases as one stops down any focus shift is mitigated, and certainly by f/11 there will be no issues. Also see @pippy above re: diffraction being a non event. Bottom line - they’re all good lenses 😉

I am interested in the “non-asph high speed lenses comment” are you suggesting with an aspheric lens DoF does not increase as you stop down or that they don’t suffer focus shift ? Assume the latter but I’m no authority 😄 though I have plenty of spherical lenses with no discernible focus shift on digital or film. 

Focus shift is typically reduced with asph lenses and especially in less aggressive, lower speed designs. 

[Vanillasludge]

31 minutes ago, pippy said:

 

I'll try to keep this brief...

A fair number of years ago, having read several comments concerning 'horrendous' focus-shift evident with the Voigtlander 40mm f1.4 Nokton, I undertook some carefully controlled studio tests with around 20 M-mount lenses. My interest was due, primarily, to my never once having noticed such a phenomenon with my own example and this had been my #1 lens ever since I bought my first digi-M (an M8.2) over a decade ago.

The bottom line? All the spherical lenses - ranging from 21mm to 135mm - exhibited some degree of focus-shift to a slightly greater or slightly lesser extent. Equally, in all cases, the focus-shift was only noticeable within a certain camera-to-subject-matter distance range and at certain apertures. Typically the distance would be approx MFD x 1.5 to 2 or 1.5 to 2.5 and the aperture range would be from Max.Ap to 1.5 to 2 stops below Max. Ap. depending on the lens.

In the case of the Nokton (IIRC) the lens started to back focus from c. 1.2m to 1.8m and the shift would go from c.1.5cm to 2.5cm when the lens was used at f1.4 and f2.0. By the time the camera-to-subject distance had gone beyond 1.8m and the aperture was stopped-down to f2.4 / f2.8 the phenomenon was no longer visible.

It is worth bearing in mind that this test was Pixel-Peeping with knobs on. In 'The Real World' absolutely none of this matters whatsoever for 99.999% of the stuff which I shoot so I don't give the matter a second's thought; happily I just get on with my snapping.

Things might well be different, of course, for other folks with different shooting preferences to myself.

Unsurprisingly the 35mm Summilux v2 shared almost exactly the same behaviour yet no-one seems fit to comment on that lens' 'Horrendous' focus-shift. I do wonder why this should be the case...

Philip.

In the film era we hardly noticed these issues because the majority of lenses still stayed within the accepted circle of confusion to “look sharp” on 35mm film.  My purchase of a grain focusing magnifier for my darkroom was what keyed me onto the Summicron shift.  I thought my rangefinder was off!😂

As the owner of the king of focus shift, the C-Sonnar, I can tell you a few interesting points about the phenomenon.  

First:  The EFFECT of shift is not a “set amount”.  If the focal plane shifts say, 05mm, the plane of sharpest focus is moved farther when focusing to greater distances.  The visible shift on my Sonnar may be 4 cm between f1.5 and 4 at minimum focusing distance, but it is substantially more the farther away the focus point.  

Second: The issue of infinity sharpness in some lenses is due to their focal plane shifting past infinity as they are stopped down.  The sonnar is the absolute worst at this and if set to infinity needs f8 to be “sharp” on a 24mp sensor at 100% enlargement.  (Use live view and you can get great detail at all apertures, you just won’t ever use your infinity bump stop)   

To add to the variables, flatness of field comes into play here.  Focus shift on the center of the frame may cause a visible shift, but that very change can actually bring outer zones INTO focus if the plane of focus on the lens curves back. (Or in and out and in, like on the c-sonnar)

As you said, stopping down the lens past to the point where the phenomenon is no longer meaningful negates all of this.  (You just have to promise yourself never to go to 100% enlargement on your M11 files😂)

In the end, these factors are mostly pixel sniffing. If the image looks good at normal viewing distances, stop thinking about it. 

PS:  The Summilux v2 is probably so terrible at 1.4-2 that nobody noticed that it was also all over the place.   

 

[Mike S]

@Vanillasludge @pippy Thank you for the updates. Not entirely news to me but helpful detail. Always fun to find I’m not alone in loving this stuff. 

[pippy]

2 hours ago, Vanillasludge said:

In the film era we hardly noticed these issues because the majority of lenses still stayed within the accepted circle of confusion to “look sharp” on 35mm film.  My purchase of a grain focusing magnifier for my darkroom was what keyed me onto the Summicron shift.  I thought my rangefinder was off!😂

As the owner of the king of focus shift, the C-Sonnar, I can tell you a few interesting points about the phenomenon.  

First:  The EFFECT of shift is not a “set amount”.  If the focal plane shifts say, 05mm, the plane of sharpest focus is moved farther when focusing to greater distances.  The visible shift on my Sonnar may be 4 cm between f1.5 and 4 at minimum focusing distance, but it is substantially more the farther away the focus point.  

Second: The issue of infinity sharpness in some lenses is due to their focal plane shifting past infinity as they are stopped down.  The sonnar is the absolute worst at this and if set to infinity needs f8 to be “sharp” on a 24mp sensor at 100% enlargement.  (Use live view and you can get great detail at all apertures, you just won’t ever use your infinity bump stop)   

To add to the variables, flatness of field comes into play here.  Focus shift on the center of the frame may cause a visible shift, but that very change can actually bring outer zones INTO focus if the plane of focus on the lens curves back. (Or in and out and in, like on the c-sonnar)

As you said, stopping down the lens past to the point where the phenomenon is no longer meaningful negates all of this.  (You just have to promise yourself never to go to 100% enlargement on your M11 files😂)

In the end, these factors are mostly pixel sniffing. If the image looks good at normal viewing distances, stop thinking about it. 

PS:  The Summilux v2 is probably so terrible at 1.4-2 that nobody noticed that it was also all over the place.

I did preface my earlier post (#15) with "I'll try to keep this brief..."......😸......

I won't go into details (you will thank me for this) but, start-to-finish, the aforementioned test took the best part of a day and the subsequent plotting of the results on graph paper (yes; I really did) drove me half-nuts and yes; field-curvature can complicate matters considerably!

I don't have an example of the 50mm f1.5 C-Sonnar (if this is the lens to which you refer) but the 'worst' performing lenses I tested were (unsurprisingly again) a pair of 50mm f1.1 using the classic Sonnar optical design. It really is in 'The Nature of the Beast' but, equally, there are photographers who will lean towards Bertele's design specifically because of its general characteristics in terms of rendering whereas others (myself included) might favour results obtained when using a Double-Gauss (or a variety thereof) design. And, of course, Apples and Oranges are not the only fruit......😺......

On the other hand one lens which has become very much a favourite hereabouts is the Rollei 40mm f2.8 HFT Sonnar. Absolutely superb in every way.

The v2 35 Summilux? What can we add to that which has already been said over the years? At f1.4 it really is, in many ways, quite 'terrible' yet in day-to-day use it possesses this 'Jekyll-and-Hyde' duality of rendering which is so unique that, for all its faults, it is somehow rather magnificent. Having given the matter some considerable thought over the last few years if I could keep only one lens to use for the rest of time it would be this. No question about it.

Philip.

Edited August 27 by pippy

[Advertisement]

Hallo,

Look here: Leica M

[jaapv]

 

On 8/21/2025 at 6:01 PM, pippy said:

forum restrictions require a drastic down-grading of IQ

Well, you dumbed them down way beyond forum restrictions. 

[pippy]

46 minutes ago, jaapv said:

Well, you dumbed them down way beyond forum restrictions. 

"dumbed them down" ?...

......Edited......but Noted.

P.

Edited August 27 by pippy