Home

Please support mflenses.com if you need any graphic related work order it from us, click on above banner to order!
SearchSearch MemberlistMemberlist RegisterRegister ProfileProfile Log in to check your private messagesLog in to check your private messages Log inLog in

Possibly stupid question: flange distance and image quality?
Post new topic   Reply to topic
Home   Manual Focus Lenses
View previous topic :: View next topic  


PostPosted: Tue Jul 16, 2024 4:40 pm    Post subject: Possibly stupid question: flange distance and image quality? Reply with quote
This might be a dumb question, but I am trying to educate myself.

Is there any relationship at all between flange distance and image quality? I am not thinking about 1 millimeter one way or the other between various SLR mounts, but more along the lines of between SLR (generally around 40 - 45mm or so) and rangefinder (about 30mm) or mirrorless (±20mm).

Put differently, If you have company X which makes the same basic lens for a rangefinder and an SLR, what if any difference might we expect to see in image quality?

Regards, Christine


PostPosted: Tue Jul 16, 2024 5:11 pm    Post subject: Reply with quote
No such thing as a stupid question
https://en.wikipedia.org/wiki/No_such_thing_as_a_stupid_question

"Is there any relationship at all between flange distance and image quality?"

I think so. For example, rangefinder lenses on dslr tend to be better than dslr lens on dslr. Why? Perhaps because design tolerances are tighter?

"If you have company X which makes the same basic lens for a rangefinder and an SLR, what if any difference might we expect to see in image quality?"

The distance from the sensor to the rear element would be the same for mirrorless, rangefinder, and SLR, yes?

Sorry, more questions than answers from me. Smile


PostPosted: Tue Jul 16, 2024 5:28 pm    Post subject: Reply with quote
visualopsins wrote:
"Is there any relationship at all between flange distance and image quality?"

I think so. For example, rangefinder lenses on dslr tend to be better than dslr lens on dslr. Why? Perhaps because design tolerances are tighter?


Sounds plausible. Although not strictly speaking a rangefinder but half frame, I find my Olympus PEN-F lenses to be very good. Approximately the same flange distance though.

visualopsins wrote:
"If you have company X which makes the same basic lens for a rangefinder and an SLR, what if any difference might we expect to see in image quality?"

The distance from the sensor to the rear element would be the same for mirrorless, rangefinder, and SLR, yes?

Sorry, more questions than answers from me. Smile


I am thinking more along the lines of: Company X makes/made both an SLR and a rangefinder camera each with their own flange distances. The 50mm lens on each type of camera is the same basic lens design, except scaled for a 30mm or a 40mm flange distance. All else being equal, would you see be able to tell which camera a picture was taken on? How about if you take both those lenses and put them on a mirrorless camera with appropriate adapters? The rangefinder lens would be 10mm closer to the sensor than the same lens designed for the SLR.

Regards, Christine


PostPosted: Tue Jul 16, 2024 6:23 pm    Post subject: Re: Possibly stupid question: flange distance and image qual Reply with quote
connloyalist wrote:
This might be a dumb question, but I am trying to educate myself.

Is there any relationship at all between flange distance and image quality? I am not thinking about 1 millimeter one way or the other between various SLR mounts, but more along the lines of between SLR (generally around 40 - 45mm or so) and rangefinder (about 30mm) or mirrorless (±20mm).

Put differently, If you have company X which makes the same basic lens for a rangefinder and an SLR, what if any difference might we expect to see in image quality?

Regards, Christine


IMO, there would be no difference. If the optical design of both lenses is the same, then the optical components of the lens (i.e. the bits that matters for imaging) would be positioned at the same distance to the film/sensor, irrespective of the mechanical flange distance.


connloyalist wrote:

I am thinking more along the lines of: Company X makes/made both an SLR and a rangefinder camera each with their own flange distances. The 50mm lens on each type of camera is the same basic lens design, except scaled for a 30mm or a 40mm flange distance. All else being equal, would you see be able to tell which camera a picture was taken on? How about if you take both those lenses and put them on a mirrorless camera with appropriate adapters? The rangefinder lens would be 10mm closer to the sensor than the same lens designed for the SLR.

Regards, Christine


The wording chosen here is a little confusing. There is no such thing as "scaling" a 50mm lens for a 30mm or 40mm flange distance. A 50mm lens is a 50mm lens and the optics will need to be mounted such that the rear nodal plane of the optics is located 50mm from the film/sensor surface (for infinity focus). I know that sounds technical but essentially it means that the only difference between using the same 50mm (optical) lens design on an SLR or on a range finder is that for the range finder the optics would be mounted much further ahead of the mount flange position, such that the actual lens elements would be the same distance from the film/sensor as they would be for the SLR.

I.e. on that same optical design 50mm lens, for the SLR the rear element's lens surface might be almost flush with the lens' mount flange, whereas on the lens for the range finder that same rear element's surface might be quire deeply recessed relative to the lens' mount flange. It is no different from needing an adapter (i.e. empty spacer) to mount a FF SLR lens onto a FF mirrorless camera; there is no impact on image quality of the optics themselves, but there may be an impact from different internal reflections resulting from the different mounts.

In practice many (esp. wide-angle) lenses for rangefinders have an optical calculation/construction that differs from their SLR counterparts of the same focal length. The absence of the mirror box allows for a more symmetrical design of the optics which makes them easier to correct for distortion amongst other things, but then you are talking about different optical designs of the same focal length, which obviously will have a different image quality.


PostPosted: Tue Jul 16, 2024 6:43 pm    Post subject: Re: Possibly stupid question: flange distance and image qual Reply with quote
RokkorDoctor wrote:

There is no such thing as "scaling" a 50mm lens for a 30mm or 40mm flange distance. A 50mm lens is a 50mm lens and the optics will need to be mounted such that the rear nodal plane of the optics is located 50mm from the film/sensor surface (for infinity focus). I know that sounds technical but essentially it means that the only difference between using the same 50mm (optical) lens design on an SLR or on a range finder is that for the range finder the optics would be mounted much further ahead of the mount flange position, such that the actual lens elements would be the same distance from the film/sensor as they would be for the SLR.


Of course, that makes sense. I hadn't considered that Smile

Regards, C.


PostPosted: Tue Jul 16, 2024 7:36 pm    Post subject: Reply with quote
connloyalist wrote:

Is there any relationship at all between flange distance and image quality?


connloyalist wrote:
Put differently, If you have company X which makes the same basic lens for a rangefinder and an SLR, what if any difference might we expect to see in image quality?


visualopsins wrote:
The distance from the sensor to the rear element would be the same for mirrorless, rangefinder, and SLR, yes?

Sorry, more questions than answers from me. Smile


If we go from a 35mm (D)SLR (about 40-50mm flange distance) to a rangefinder (around 30mm) and to a mirrorless (about 20mm), the optical designer of a lens can move the last element of the lens much closer to the film/sensor. This gives him/her a lot of additional freedom to design the lens, especially when it comes to wideangle lenses.

Designing wideangles for SLRs first was impossible (an 4.5/40mm being the limit). Then, around 1950, the "inversed telephoto" lens (later known as "retrofocus") was invented by Angénieux and Zeiss Jena, and 35mm lenses (about 65° image angle) became possible. N At the same time Bertele was designing his 90° 4.5/21mm Biogon for the rangefinder Contax. Only much later - around 1965 - such lenses were introduced for SLRs as well. This gives you an idea about the difficulties involved when going from 30mm flange distance to 45mm flange distance!

Similar - albeit not as dramatic - difficulties were encountered when designing fast (f1.4 and f1.2) 50mm normal lenses for SLRs. f1.2 was achieved only around 1978! Again, fast 1.2/50mm rangefinder lenses had been developped at least ten years earlier ...

Tele lenses of the same generation (!) however usually do not differ much. The Leitz 2/90, 2.8/90 and 2.8/135 were the same optical constructions for both the SLR and the rangefinder Leicas, as far as I know.

There's another twist to the story when it comes to digital sensors. At least the early sensors were pretty insensitive to oblique light rays. This leads to stronger-than-usual vignetting and/or color cast in the corners if traditional symmetric (rangefinder) wideangles are used. AND there's the thick filter stack in front of the sensor which may cause massive performance degeneration if traditional symmetrical wideangles are used.

So it's a somewhat complicated issue, but easy to understand if you are aware of the three main points mentioned above.


visualopsins wrote:
No such thing as a stupid question
I think so. For example, rangefinder lenses on dslr tend to be better than dslr lens on dslr. Why? Perhaps because design tolerances are tighter?


I don't really understand what you're meaning here. Apart from were few exceptions you cannot mount rangefinder lenses on DSLRs unless you restrict yourself to macro work. How can you can you judge their preformance on DSLRs then?

S


PostPosted: Tue Jul 16, 2024 9:10 pm    Post subject: Reply with quote
There are less design restrictions on wide angle lenses when the mount has a shorter flange distance, so there can be IQ improvements.
For lenses where the focal length is moderately above the flange distance such constraints are gone and there should be nothing related to the mount. There may be issues with deciding to make the lens cheaper, smaller or lighter.

In each case there will be a load of other factors involved.


PostPosted: Tue Jul 16, 2024 9:36 pm    Post subject: Reply with quote
stevemark wrote:
...
visualopsins wrote:
I think so. For example, rangefinder lenses on dslr tend to be better than dslr lens on dslr. Why? Perhaps because design tolerances are tighter?


I don't really understand what you're meaning here. Apart from were few exceptions you cannot mount rangefinder lenses on DSLRs unless you restrict yourself to macro work. How can you can you judge their preformance on DSLRs then?

S


Good catch! Misspoke there, meaning to substitute mirrorless for dslr.


PostPosted: Wed Jul 17, 2024 4:15 am    Post subject: Reply with quote
I have the Zeiss Distagon 21mm 2.8 in e-mount. I see it also comes in Canon or Nikon DSLR. But those are obviously a different calculation and much larger. Shorter flange allows closer optics allow more flexibility for designers to solve for optimal designs.


PostPosted: Wed Jul 17, 2024 11:32 am    Post subject: Reply with quote
stevemark wrote:

Designing wideangles for SLRs first was impossible (an 4.5/40mm being the limit). Then, around 1950, the "inversed telephoto" lens (later known as "retrofocus") was invented by Angénieux and Zeiss Jena, and 35mm lenses (about 65° image angle) became possible. N At the same time Bertele was designing his 90° 4.5/21mm Biogon for the rangefinder Contax. Only much later - around 1965 - such lenses were introduced for SLRs as well. This gives you an idea about the difficulties involved when going from 30mm flange distance to 45mm flange distance!

Similar - albeit not as dramatic - difficulties were encountered when designing fast (f1.4 and f1.2) 50mm normal lenses for SLRs. f1.2 was achieved only around 1978! Again, fast 1.2/50mm rangefinder lenses had been developped at least ten years earlier ...

Tele lenses of the same generation (!) however usually do not differ much. The Leitz 2/90, 2.8/90 and 2.8/135 were the same optical constructions for both the SLR and the rangefinder Leicas, as far as I know.

There's another twist to the story when it comes to digital sensors. At least the early sensors were pretty insensitive to oblique light rays. This leads to stronger-than-usual vignetting and/or color cast in the corners if traditional symmetric (rangefinder) wideangles are used. AND there's the thick filter stack in front of the sensor which may cause massive performance degeneration if traditional symmetrical wideangles are used.

So it's a somewhat complicated issue, but easy to understand if you are aware of the three main points mentioned above.


Thank you, this is what I was looking for.

Regards, C.


PostPosted: Thu Jul 18, 2024 1:13 pm    Post subject: Reply with quote
stevemark wrote:

There's another twist to the story when it comes to digital sensors. At least the early sensors were pretty insensitive to oblique light rays. This leads to stronger-than-usual vignetting and/or color cast in the corners if traditional symmetric (rangefinder) wideangles are used. AND there's the thick filter stack in front of the sensor which may cause massive performance degeneration if traditional symmetrical wideangles are used.
S


Steve, could you please expand on the aspect of symmetric lens formula's (I understand this as usually being 6/4 double Gauss?) influencing results on sensors?

As I understand the issue of light hitting the sensor at an angle, the actual light sensitive part of a sensor is (or, used to be; relation with BSI sensor?) at the bottom of a "tube". Light coming at this tube at an angle would decrease the efficiency resulting in the color casts and vignetting you mention.

What I don't understand is how symmetrical lens design affects this? I do notice that for example the Canon LTM 50mm 1.4 is a 6/4 while the Canon FD 50mm 1.4 is a 7/6.

Also, does this suggest that lenses designed for digital sensors tend to have the light hitting the sensor straight, not at an angle?

Regards, Christine


PostPosted: Thu Jul 18, 2024 1:40 pm    Post subject: Reply with quote
connloyalist wrote:
stevemark wrote:

Designing wideangles for SLRs first was impossible (an 4.5/40mm being the limit). Then, around 1950, the "inversed telephoto" lens (later known as "retrofocus") was invented by Angénieux and Zeiss Jena, and 35mm lenses (about 65° image angle) became possible. N At the same time Bertele was designing his 90° 4.5/21mm Biogon for the rangefinder Contax. Only much later - around 1965 - such lenses were introduced for SLRs as well. This gives you an idea about the difficulties involved when going from 30mm flange distance to 45mm flange distance!

Similar - albeit not as dramatic - difficulties were encountered when designing fast (f1.4 and f1.2) 50mm normal lenses for SLRs. f1.2 was achieved only around 1978! Again, fast 1.2/50mm rangefinder lenses had been developped at least ten years earlier ...

Tele lenses of the same generation (!) however usually do not differ much. The Leitz 2/90, 2.8/90 and 2.8/135 were the same optical constructions for both the SLR and the rangefinder Leicas, as far as I know.

There's another twist to the story when it comes to digital sensors. At least the early sensors were pretty insensitive to oblique light rays. This leads to stronger-than-usual vignetting and/or color cast in the corners if traditional symmetric (rangefinder) wideangles are used. AND there's the thick filter stack in front of the sensor which may cause massive performance degeneration if traditional symmetrical wideangles are used.

So it's a somewhat complicated issue, but easy to understand if you are aware of the three main points mentioned above.


Thank you, this is what I was looking for.

Regards, C.


Even within the standard lens focal length the rangefinder lenses tend to be symmetric or even slightly telephoto where the SLR lenses are somewhat retrofocus. The 45mm focal length fixed rangefinder lenses I have salvaged are near symmetrical, my 50-55-58 SLR lenses are all retrofocus. A rough method: Set the aperture at say 5.6 and keep the lens at arm length, first look at the front side then back side, aperture opening size differs with asymmetric designs. Retrofocus lenses show a larger aperture hole at the back side, telephoto lenses a smaller aperture hole at the backside, if symmetrical there is no difference.



10 fixed rangefinder lenses all within 45-50mm 1.7-2.8: 6 slightly telephoto at the left side, 3 symmetric in the center, 1 slightly retrofocus in front of the right side (Petri 45mm 1.Cool.
The SLR lenses 50-58mm 1.2-2.0: they are all at the right side, all retrofocus, the Canon SSC 55mm 1.2 at the rear left the most retrofocus.
The goal to create a more compact camera is probably the reason that there are standard focal length rangefinder lenses with a slight telephoto effect. The slightly retrofocus Petri 45mm 1.8 has a measured focal length of 46,9mm, the longest measured focal length of my 45mm rangefinder lenses.

Yes, the A7RII FF sensor shows more edge issues with the rangefinder lenses. True they are more wide angle than the SLR standard lenses that I have, on the other hand the standard SLR lenses have that focal length mainly because the register distance needed for a flipping mirror didn't allow a focal length equal to the diagonal of the film format, either go for a longer focal length or/and make them more retrofocus. By going even more retrofocus some 45mm standard SLR lenses became possible later on.


PostPosted: Thu Jul 18, 2024 1:59 pm    Post subject: Reply with quote
@ Ernst:

Be careful not to mix up the two pairs

1) "retrofocus <=> tele lens"
2) "symmetrical <=> "asymmetrical"

None ofyour normal lenses shown above are retrofocus constructioins (the Zeiss APO Distagon 1.4/55mm however is!).

"Retrofocus" lenses have a negative front group followed by a positive rear group (which is the master lens). "Tele" lenses have a positive front group (=master lens) followed by a negative rear group.

Typical fast vintage 1.4/50mm SLR lenses of course are not symmetric, but you'll never find a negatve front group - they are NOT retrofocus constructions!

S


PostPosted: Thu Jul 18, 2024 2:43 pm    Post subject: Reply with quote
stevemark wrote:
@ Ernst:

Be careful not to mix up the two pairs

1) "retrofocus <=> tele lens"
2) "symmetrical <=> "asymmetrical"

None ofyour normal lenses shown above are retrofocus constructioins (the Zeiss APO Distagon 1.4/55mm however is!).

"Retrofocus" lenses have a negative front group followed by a positive rear group (which is the master lens). "Tele" lenses have a positive front group (=master lens) followed by a negative rear group.

Typical fast vintage 1.4/50mm SLR lenses of course are not symmetric, but you'll never find a negatve front group - they are NOT retrofocus constructions!

S


Alright, that is the strict definition of retrofocus and telephoto. However it is quite obvious that the asymmetry in the SLR standard lens samples I have is there to overcome the longer register distance of an SLR camera. It is not due to the magnification ratio of the 35mm frame format in normal use as both rangefinder and slr standard lenses have to cope with that. I should have written less positive for the front cell and more positive for the rear cell then. Which in the end will deliver less oblique rays on the sensor.


PostPosted: Mon Jul 22, 2024 3:06 pm    Post subject: Reply with quote
stevemark wrote:
Designing wideangles for SLRs first was impossible (an 4.5/40mm being the limit). Then, around 1950, the "inversed telephoto" lens (later known as "retrofocus") was invented by Angénieux and Zeiss Jena, and 35mm lenses (about 65° image angle) became possible.


Not to be overly pedantic -- slightly pedantic may be OK -- but TTH was there first, in 1932, with 25 and 35 mm lenses for 3 color Technicolor cameras.
Post new topic   Reply to topic
Home   Manual Focus Lenses
Page 1 of 1