Shorter lens register = "crisper" images?

[DeeJayK]

A sort of "lifestyle" article on Leica in today's NYTimes caught my eye with one particular paragraph:

Unlike DSLR (digital single-lens reflex) cameras, in which a photo subject is viewed through a mirror, range finders offer crisper images because the lens is closer to the sensor of the camera. The shutter on a range finder is also incredibly still and quiet — unlike the DSLR, in which the mirror flips up and down at the moment a photo is snapped — so little vibration is added to a photo when the shutter is pressed.

I'm no expert, but I've never heard or read anything else that espouses a relationship between flange distance and resolution, but there it is in black and white (or perhaps "monochrom" considering the subject of the article) in the New York Times.

It seems that :43: has both "advantages" of the Leica design only more so. Given that the lens register of :43: is over 31% shorter than that of Leica M, I guess that we should expect a 31% gain in image crispness.

All joking aside, is there any reason to expect there to be any advantage to having the lens mounted closer to the imaging plane or is this another case of a "tech" writer just making things up? I looked into this writer's bio and found some interesting qualifications including that he "helped design the first Britney Spears doll" and that he is "currently writing" a book which was published in 2010 -- pretty impressive résumé, I must admit.

 

[AceAceBaby]

I would think there must be a limit. To take it to extremes, if the distance were 1mm, so much of the light from the lens would be hitting the sensor at hugely oblique angles that must degrade the image quality?

I was thinking about this the other day in relation to, eg, Sigma lenses refitted for different mounts, and thinking that longer might be better with these relatively short distances, in order to try and keep the light hitting the sensor as directly as possible. My hunch about this also came from the fact that the Nikon mount (one of the longest register distances in SLRs) was chosen for the first Metabones adapter. I'd guessed that the longer RD would reduce image resolution headaches from more acute angles, meaning their first product would be the easiest.

I am looking forward to all the maths and diagrams explaining it all.

 

[CiaranCReilly]

I think we should take the article as it is, an enjoyable lifestyle article, and not get too caught up! If people who know no better buy Leicas as a result, the market may be flooded with cheap used Ms in six months

The "crisper" claim is right beside a phrase describing a rangefinder shutter as "still". I can't see how register distance would affect image quality though, I would have thought the opposite if anything (due to rays hitting the sensor at smaller angles).

Maybe he means the view through the viewfinder is crisper?

 

[Ulfric M Douglas]

I like the word "codswallop" myself.

 

[Jonathan F/2]

I think having a longer registration distance also has it's benefits. Nikon for example has the awesome 14-24mm 2.8 that even Canon hasn't been able to replicate. Saying that, using a Nikon f mount adapter on M43, even with the large gap, sharp glass is still the most important component in getting crisp images.

 

[slothead]

I like the word "codswallop" myself.

Look it up. There is at least one "interesting" definition.

... edit ...

Having said that, I can't see where you saw it used.

 

[wildwildwes]

Yes, the article writer's assertion is absolutely correct. The fact that there is NO mirror & it's associated mechanism between the lens optic and the sensor (or in the age of film cameras -- the film plane) does in fact give the upper hand to rangefinder cameras. This advantages are at least three-fold (again, as noted in the article) in that there is 1.) no chance of mirror vibration issues in a rf style camera AND 2.) having the lens optic nearer the sensor allows for sharper more contrasty images due to better corrected lens designs & 3.) mirror-less designs are able to be built much smaller and lighter.

All of that said, these same attributes hold true for mirror-less camera designs! It is my opinion that within a few years ALL major manufacturers of cameras will start offering mirror-less cameras (whether for general consumer use or professional). Technology has negated the need for the mirror-reflex design!

Cheers-

 

[RobWatson]

So c mount lenses are "crispyest" on account they have even less distance between rear optic and sensor? Codswallop is too dignified a name for balderdash.

Better and more corrected designs compared to the sloppy craptacular deigns to date from Leica, Nikon Canon, Zeiss, etc? Anyone that even attempted to understand optical design theory would appreciate that the closer the optic gets to the sensor the more difficult the design becomes. I'm not one to play the logical extremist card but imagine the fun with a wide field optic in contact with the sensor - just exactly how is that even going to work at all?


About the best one can postulate is when looking through the view finder the mirrored systems add an optical component so image quality may be degraded (crappy/dirty mirror, bad alignment, etc). Given the the mirror is actually moved out of the way during capture ...

 

[DeeJayK]

So c mount lenses are "crispyest" on account they have even less distance between rear optic and sensor?

How about smart phone cameras...the distance between the lens and sensor of something like the iPhone can't possibly be more than a millimeter or two. With that much "crispness" to offer, there's no wonder it's such a popular camera.

 

[Sammyboy]

Ideal register = diagonal of the format.

 

[RobWatson]

At some point the angles are severe enough that real problems are encountered.

Total internal reflection - Wikipedia, the free encyclopedia

 

[Ned]

Those of us who were four-thirds users before micro four-thirds existed, are familiar with the term "Telecentric Design". Micro Four-Thirds does not have this telecentric design because of the need for such a short register, and therefore exhibits the disadvantages of high light angles which are also present in full frame DSLRs. This is why Micro Four-Thirds cameras require automatic distortion correction and in some cases automatic Chromatic Aberration correction. Both of these problems are increased by the short register (though if you're adapting SLR glass then you return the telecentricity to the system through the dead space of the adapter).

The advantage of a short register is the ability to make wider angle lenses a lot easier without the need for retrofocus design. However, I cannot see "crisper images" as being an advantage. If anything, I would expect the opposite (for reasons already mentioned). I call BS.

PS, another advantage of a longer register is easier macro capabilities. The long register is like adding an extension tube to every lens. The close focus distance of many rangefinder lenses is horrendously long, making close-up photos difficult to take. Kudos to Olympus for being able to engineer such good close-focus into their m4/3 lenses, as well as combating Chromatic Aberration through optical measures and not post-production. Most people don't appreciate the feats they had to perform in order to overcome all the optical hurdles of a system as small as m4/3.

 

[goldenlight]

I suspect this is one of those esoteric theories that has some basis in truth but is irrelevant in practice.

 

[meyerweb]

Those of us who were four-thirds users before micro four-thirds existed, are familiar with the term "Telecentric Design". Micro Four-Thirds does not have this telecentric design because of the need for such a short register, and therefore exhibits the disadvantages of high light angles which are also present in full frame DSLRs.

Actually, I believe that most m43 wide to normal lenses ARE telecentric, to compensate for the problems of a short register.

 

[dwig]

... I can't see how register distance would affect image quality though, ...

The register distance has absolutely nothing to do with potential image quality or the angles at which the light strikes the film/sensor, at least directly. It is merely a measure of where the physical mounting flange is relative to the film/sensor.

That said, a small register distance, when accompanied by a lack of any other obstruction (e.g. reflex mirror, ...), does give the optical designers some additional freedom and can sometimes aid in producing better optical designs.

 

[drd1135]

I like the word "codswallop" myself.

It sounds so much more genteel than "crap".

 

[conurus]

In my humble opinion it might be educational to look at the 21/2.8 Biogon for Contax G and also the Distagon 21/2.8 for Contax/Yashica. Same manufacturer, same FL, same max aperture. Both are reckoned as some of the best wide angles ever made.

Biogon: simpler 9 element design, excellent flare resistance, no distortion. However, more vignetting.

Distagon: a lot more complicated design, ghosting problem rare but does occur, mustache distortion.

I think a retrofocus wide with a lot of room behind the lens can be made very high quality if you put enough effort into it. Even recent Leica wides employ retrofocus designs. The primary issue with a very short backfocus (distance from rear element to image plane) is vignetting. There are some design trade-offs involved here and it all depends on the design objective of the lens.

 

[Djarum]

The first thing I thought about after reading just the quoted line was that, gee, the extra few millimeters of air the light doesn't have to go through is soooo much better...

 

[AceAceBaby]

The register distance has absolutely nothing to do with potential image quality or the angles at which the light strikes the film/sensor, at least directly. It is merely a measure of where the physical mounting flange is relative to the film/sensor.

Rear elements can't get (much) closer to the sensor than the mount, though*, so it directly influences that in being a minimum limit.

*Adapted C-mounts can get closer, but they are pushing it.

 

[flash]

Once upon a time this was true. When the SLR was first introduced the manufacturers had aan awful time with wide lenses because the rear element was forced away from the film plane. Designs of lenses became very complex and there wasn't the technology to play with exotic glass and design precision required. After about 50mm it didn't matter anymore because the rear element was naturally placed further away than the flange distance anyway.

Today, with CAD and improvements in manufacturing as well as the development of ASPH, APO and special glass (UD, flourites etc) many modern SLR lenses are superb, in spite of the design differences. So I don't think the statement made holds true any more. The Nikon 12-24 is an excellent example of this. This lens would have been nearly impossible to make, even 15 years ago.

Gordon

p.s. poor corners in lenses are often as a result of the flange width, rather than depth. That's why Canon used an all new EOS mount when it went to AF. It took Nikon some years to catch up to Canon because they kept their legacy mount and they are still not making some lenses quite as fast as the Canons. But again technology has helped Nikon get around some of the limitations of the smaller mount and the playing field is now level.