resolution of LF lens

greenspun.com : LUSENET : Large format photography : One Thread

I'm going to buy a LM system. Sb. told me the following today, The resolution of large format lense is lower than 35mm lenses. The resolution of 35mm lenses is about 120 line pair per mm The resolution of MF lenses is about 40 line pairs per mm The resolution of LF lenses is about 20 line pairs per mm I can not beleive on it due to the LF lense is more accuratly produced and the structure looks more simple no more moving elements. I have no clare idea to support meself. Does sb. have expeirences or any data which could answer it with yes or no. Thanks Simon

-- Simon ZENG (szeng@cn.schneider-electric.com), October 05, 2000

Answers

I'd like to see that 35mm lens that could deliver 120 lppm to the film!
The root of this misconception is based in diffraction. This sets a fundamental limit to the resolution of any lens. As numerical aperture increases (the hole in the lens gets smaller), then resolution falls in proportion.
Because lenses for larger formats tend to be used at higher f numbers, then this diffraction limit shows itself more readily, and yes, theoretically LF is worse off than 35mm. But if you take a picture at f/22 with a 35mm camera, and with a medium format and a 5x4 camera, and then enlarge them to the same print size, then the 35mm will have the poorest image quality, and the LF the best. In fact, if depth of field wasn't an issue, then the same would be true at any given aperture

There is absolutely no point in having a lens that can resolve 120 lppm (??!!) if you can't get anywhere near that resolution on normal film.
In addition, resolution is not quality. Quality has to do with smoothness of tone, contrast, accuracy of drawing, and a whole range of other factors that aren't easy to put a number to.
If all you want to do is photograph black and white lines very close together, then stick to 35mm and Technical Pan film. But if you want to take pictures, then move up to a larger format.

-- Pete Andrews (p.l.andrews@bham.ac.uk), October 05, 2000.


Even the resolution of an ideal lens is diffraction limited according to the laws of physics. The limitation is imposed by the opening of the diaphragm. For obvious reasons, the resolution behind the lens cannot exceed the resolution of this "point". Since this opening is usually smaller than the film format size, the image must be expanded. Expanding the image will lower the overall resolution, since there is no way to "add information" behind the diaphragm.

If you compare two *ideal* lenses of the same speed for different film format sizes, they will yield exactly the same amount of "information". The only difference is their "density".

In an ideal system, the question is not film size but film format. The square format will capture most of the resolution provided in the image *circle*

So LF has actually less usable resolution, because the image circle is and needs to be larger than film format. On the other hand, film and lenses are still far from beeing ideal. Since film does have a limited resolution on its own, the combination of lens and film is an optical system, too. The combined resolution of this optical system is always less than the resolution of the weakest element. In LF, film is not the limiting factor, because film resolution is always higher than lens resolution. As long as film resolution is not improved far beyond todays limits, LF will always capture more information than MF or even smaller formats.

-- Thilo Schmid (tschmid@2pix.de), October 05, 2000.


Read the "35 mm, medium format, or large format?" article at http://www.photodo.com/nav/artindex.html and compare pic from different negs: #4(35mm),#5(6x6) and #6(4x5).

-- Andre (5745@my-dejanews.com), October 05, 2000.

Many LF test at 60-80 LPM and some MF lenses beat that. That is test ing at the middle of the aperature. Getting hung up on resolution figures is a 35mm amateur type of thing and doesn't have much place in LF unless you are making wall murals. Much more important in LF work is coverage and contrast. Incidentally, the highest test results I ever read about was for a Leitz 50mm Summicron, which resolved 150 LPM. Many of the lenses in all formats will resolve more LPM than the film is capable of recording unless you use something like Tech Pan. I you are planning to shoot LF, forget about LPM and shoot pictures for content. A sharp lens is less important than a sharp eye and mind.

Regards, Doug.

-- Doug Paramore (dougmary@alanet.com), October 05, 2000.


Last night I was looking with an 8x loupe over a contact print from an 8x10" TMX negative of a market square taken with a pre-WWII ultra-wide lens--no match for any modern lens certainly. I saw things I didn't see when I fired the shutter--a girl with some friends looking back at the camera from over 30 feet away, detailed images of people drinking at a cafe about 45 feet away, a headless rollerblader maybe 60 feet away who must have bent into a tuck during the 1/5-second exposure, print on signs all the way on the opposite end of the square at least 100 feet away reminding me of another cafe where I had a cappucino and a camera shop next door where I recalled glancing at some used lenses. If I had used the newest 35mm camera with the best lenses, I might have seen the girl looking back, but the rest would all be blobs.

-- David Goldfarb (dgoldfarb@barnard.edu), October 05, 2000.


Sorry Thilo, but image circle has absolutely nothing to do with resolution. You can have a triplet design with a narrow coverage angle and poor resolution, or a well corrected symmetrical lens with wide coverage and excellent resolution.
Also, since the aspect ratio of LF is closer to square than 35mm, then LF makes more efficient use of a given image circle.


-- Pete Andrews (p.l.andrews@bham.ac.uk), October 05, 2000.

Check this site for some good real world testing of lens resolution : http://www.hevanet.com/cperez/index.html

-- Ron Shaw (shaw9@llnl.gov), October 05, 2000.

Pete, in the first part of my explanation, I was talking about an ideal lens. The amount of information availiable to the film is independent of the image circle. But the resolution, measured in dots per in inch, is not. Otherwise you would be able to increase the amount of information availiable to the film by just increasing the image circle, regardless of the diameter of the diaphragm. This ist physically not possible. Diffraction Limits mean "Information Limits".

The more suqare format sized LF will not save you more information als long as a larger image circle is desireable. The amount of information captured by film depends on the percetage of the image circle actually used, not on its aspect ratio. In LF, a much larger image circle is usually desired. This is why Coverage is more interesting than Resolution. The cost for that is loss of information. But LF will compensate for this. And it will do so forever, because even if manufacturers will push film resolution far beyond todays limits, the sharpness and color saturation of their enlargements will be diffraction limited by physics, too. Diffraction happens on every pixel.

BTW: the most potent 35mm photographic lenses are those built for microdocumentation systems with 35mm non-perforated film, like the Zeiss S-Planar 5,6/60 and Zeiss S-Biogon 5,6/40. There resolution is at least 250lp/mm in the outer image circle and near 350lp/mm in the center. They are "the perfect lens" for 35mm-enlargers, if you can get one...

-- Thilo Schmid (tschmid@2pix.de), October 05, 2000.


well, i read the article and examined the pictures referenced above in andre's post. frankly, i dont know what they think they are saying. the 4x5 images are uniformly superior in every way compared to the smaller formats. additionally, i would guess that most of you LF users are like me - we all started out using small fomat cameras, and hesitated at every step before moving to larger formats becasue of cost, size, weight, flexibility - all the drawbacks you hit as you move upward in format size. there was a time when i knew i would never want to deal with all the rigamarole that goes along with film holders, no zoom lenses, having to use a tripod on every shot, etc. i loved my nikon F2, but after my first experiences with medium format (a beautiful old rollie 3.5F), i could not even look back - the first prints i made from 6cm negs were far sharper and better than anything i had previously done in 35mm. a few years later, i held my breath and tried LF, and again, there is simply no way i can go backwards - the quality improvement moving from MF to LF was striking, even at 8x10 print size. not to say that a MF print isnt sharp - they are, and can be superb - but it was the additional richness of tonality that did it for me. the extended gray scale reproduction (range of middle tones) is critical in my architectural work, and careful use of LF is the only way i know to acheive it. what i am saying is the proof is in the pudding - i doubt any of of would bother using all this bulky equipment if our experiences had taught us that smaller formats could do the same job just as well.

-- jnorman (jnorman@teleport.com), October 05, 2000.

I don't know who "Sb" is, but they are only qualitatively correct. The numeric values are wrong. LF lenses have modestly lower resolution in line pairs per mm, but the greatly decreased enlargement required for a given size print more than compensates for the modestly decreased resolution on a negative. As an example, suppose that a 35 mm lens can do 80 lppm and a 4x5 lens 50 lppm. On an 11x14 print these become about 7 and 18 lppm, predicting the LF resolution advantage that we know is true from experience.

In practice, few 35 mm photos are taken with a tripod, so the theoretical resolution of 35 mm is not achieved and the resolution advantage is even more on the side of LF. Of course, there are some photos that can't be made with LF or with a tripod.

-- Michael Briggs (MichaelBriggs@earthlink.net), October 05, 2000.



That's not the way diffraction works Thilo. The smallest point of light that can be theoretically focussed by a lens is called the Airy disc. The radius of the Airy disc, to its first bright fringe, is given by the formula: r = 1.22lambda*Nf. Where lambda is the wavelength of light, and Nf is the numerical aperture of the lens. So for yellow light this is approximately 0.72Nf microns in diameter, or in terms of resolution 1392/Nf lppm.
This is true for any diffraction limited lens, regardless of focal length, and regardless of image circle.
Any theoretically perfect lens with an aperture of f/5.6 will therefore have a resolution limit of 248 lppm, whether it covers 35mm or 20"x16".
If you care to think of diffraction as an 'information density' limit, then it works against smaller formats.

What limits resolution in practise is mainly chromatic abberation, which is inextricably linked to focal length for refractive lenses. Even so, at apertures of f/16 and above (numerically), large format lenses can achieve resolutions close to the theoretical maximum.
Anyway, this obsession with pure resolution is still a complete red herring when it comes to assessing the quality of an image, and the ease with which a given quality can be acheived.

-- Pete Andrews (p.l.andrews@bham.ac.uk), October 06, 2000.


Mr.Jnorman, 1)i just want tell Mr.Simon that 'line pairs per mm' is _not_ so important in final print, compare #4 and #6 photos, may be #6 is little bit less sharp than #4 and so what? The _total_ quality of photo from 4x5 negs is much better. 2) i switch myself from 35mm to 6x6 and 4x5 due to quality, but I _understand_ that each piece of equipment has it's own place. It's like a screwdriver sets: if you good mechanics, you have in your vest Philips #1,#2,#3 screwdriver. It's simply impossible to made all you work with one size screwdriver because of different screws sizes.

-- Andre (5745@my-deja.com), October 06, 2000.

The photodo article is talking about sharpness, not tonality or grain. A less contentious example than the format comparison is the way that film with hard-edged grain can be very sharp but have dreadful tonality: lith film is a limiting case. Similarly, in the photodo examples the edges of the test patterns are well defined in the 35 mm shots but the areas of solid colour have bad grain. It's sharp, but ugly.

Roughly, the tonality of film is given by the grain density, so it scales with the area of the format. Sharpness is related to the differential of the grain density - how fast it varies from place to place - so it only scales linearly with the format size. For the same angle of view, the required focal length grows linearly with the film size, but the difficulty of making that lens grows at least quadratically so there is the possibility that the overall system sharpness will actually get worse. In real life there is so much variation between lenses, focussing systems and the way the film is held that all bets are off.

The sharpest civilian lens Kodak ever made was a plastic moulded singlet used in their disc cameras. That should give some clue as to how well sharpness alone relates to image quality :-)

-- Struan Gray (struan.gray@sljus.lu.se), October 10, 2000.


Moderation questions? read the FAQ