Image quality when focussing closer...

Hi there,
Not sure how to phrase this, but I have a thought/question about lenses and image circle when close focussed:
As a lens is focussed on increasingly smaller distances from the film plane or camera, the distance of (parts of) the lens is moved away from the camera, either by the focus ring, extension tubes or bellows unit.
Q. As the lens moves further away, does the image circle (at the film plane) get bigger?
If so, and one is using less of the complete image produced by the lens, is this reducing the resolution of the lens? Well, I guess that the resolution of the lens is still the same but could one be using 'less' of the resolution as a lot of the image from the lens is outside the area of the film plane?
Not very well put I know but perhaps a summary thought could be; would the captured image get worse as the lens is focussed closer esp with tubes and bellows?

Thanks for staying with the rambley question - any thoughts/answers welcomed!
Sim2.

Yes, the image circle grows as the lens is moved away from the film plane.

Whether increasing magnification improves or reduces image quality depends on the lens. A lens can be optimized for only one film plane-to-subject distance. If it is optimized for a distant subject, it will give worse image quality on a close one. And vice versa.

Most lenses for general use are optimized for big subject in front of the lens, small negative behind it. This is why reversing the lens when shooting at magnifications above 1:1 is recommended. Reversing the lens for big front, small behind makes best use of the lens' optimizations.

<sigh> the way urban legends grow!
Yes, lenses are optimized to perform well within a certain RANGE of distances...for example, typical non-macro lens is optimized for good performance in the range of about 9x FL to Infinity.
Lenses which focus closer than that need to have ADDITIONAL optimization so that it better focuses the different wavelengths of light that comprise white light. Apochromatic (APO) lenses are designed to bring different wavelengths (red, green, and blue) into focus in the same plane. The residual color error (secondary spectrum) can be up to an order of magnitude less than for an achromatic lens of equivalent aperture and focal length. Apochromats are also corrected for spherical aberration at two wavelengths, rather than one as in an achromat. That is why, in larger format lenses, the APO (apochromatic) designated lenses are used as 'process lenses' primarily designed for the graphics industry, they are flat field lenses which are optimized for 1:1 reproduction.
That does NOT mean that they perform poorly at Infinity...it means that their design takes in special consideration of close field work and focusing different wavelengths of light extra well in that environment.

So lenses for enlargers have a design target where they perform especially well (NOT that outside that range they are mediocre!) For example, Rodenstock 105mm and 150mm APO enlarging lenses are among the finest they make (better than a 105mm or 150mm non-APO lens from them), and these lenses are designed with a stated optimum magnification of 6X...up to a 24x30" print from 4x5" sheetfilm, the lens introduces near-zero distortions. But as enlargement magnification gets to 10x, the lens can introduce -0.1% distortion (barrel/pincushion) out near the edges of the image area. That is an example of optimization for a point in the range...outside that range of 6X (at 10X with 0.1%) is absolutely not 'bad', it is still quite spectacular!

Thanks for the replies.
I knew that reversing a lens can be better than using it the right way around when getting around 1:1, but never really thought about why... ah, the joys of having a little techie learning!
Perhaps where my thought might have been leading could be for, say a medium format system/lens, when the medium format lens is far enough away from the image plane that the image circle is rather large - I don't know any figures for this, could a 'better' image be projected from a lens for a smaller format eg 35mm, as the image circle from the 35mm lens is now large enough to cover the medium format film? Thought being that a complete image circle from a smaller format lens might be better than part of an image circle from a correct lens for the system???

Apologies for asking these odd queries but occaissionaly the brain goes off on a random train of thought and today, you guys are the *lucky* receipients of this ramble! Your considerations appreciated.

Entertaining. The answers are mostly correct so I don't have much to add except to say that the formal definition of apochromatic has everything to do with the number of wavelengths brought to focus and nothibg to do with object / image distances.

Object distance information is loosely tied to the design terms "infinite conjugate" and "finite conjugate" or macro.

I'll also verify that designs can be corrected over a range of distances with no detectable falloff in optical MTF performance. That is separate from the idea of best focus based on depth of field. As the range of corrected object distances increase, the design becomes more complex...especially at short distances where the magnification is changing so much.

Here is what Rodenstock says about one of their APO lenses designed to operate well around 1:1 macro range.


Certainly Rodenstock also makes non-macro versions of APO lenses, too.

''Thought being that a complete image circle from a smaller format lens might be better than part of an image circle from a correct lens for the system???''

I suppose it depends on what you mean by ''better''. I would have thought that as the lens moves further away from the film plane; the image on the ground glass will broaden somewhat if not much, but it will also darken. Although in shooting the picture; if the time is extended sufficiently then there might not be any difference in the exposure of the negative. Film speed, development procedures, and enlargement intentions would then be relevant as far as ''better'' is concerned.
But I'm no techie so take that with a pinch of salt. Just my immediate thoughts.

I have been scratching my head with something similar; as I wanted to know the pros and cons of using a very long (and relatively inexpensive) 135 format lens on a home-made camera with a 4x5 back. i.e. how much coverage could I expect to get on the ground glass with that combination.
So frustrating, I wish there was more info online for Andreas Feininger's 1200mm camera construction. If there was, I'd definitely give it a go.

JP

As far as 'better' is concerned - I have made an assumption that whatever I throw in front of the film plane, I can achieve a correctly exposed and correctly focussed image.
The 'better' would be, if all other things were the same, one print would/might look better than the other to the viewer, whatever the tech reason. i.e. the viewer need not know or understand if the lens was apo corrected, a macro lens, a reversed enlarging lens or a lens made for a different format; when comparing A to B, one could be selected as the 'better' of the two!

Its a question of whether more or less magnification makes a better picture. Yes. No.

Example. I used to shoot flowers and such with KM, a MicroNikkor, usually 105/2.8 AIS, and electronic flash. If I wanted the slide to contain good detail in the flower, I shot at higher magnification and lost the setting. If I wanted to show the flower in its setting I shot at lower magnification and lost good detail in the bloom. Which was better? With 24x36 I couldn't have everything, had to choose.

In other words, now that you've explained what you meant, your question is unanswerable. And it has nothing to do with optics, everything to do with photographic purpose.

I solved my problem to my satisfaction by moving up to 2x3, which let me get reasonable detail in the main subject and enough of its setting in the frame. Usually, not always. 2x3 brought other problems with it, not germane here.

The question of using

Sim2 said:

Hi there,
Not sure how to phrase this, but I have a thought/question about lenses and image circle when close focussed:
As a lens is focussed on increasingly smaller distances from the film plane or camera, the distance of (parts of) the lens is moved away from the camera, either by the focus ring, extension tubes or bellows unit.
Q. As the lens moves further away, does the image circle (at the film plane) get bigger?
If so, and one is using less of the complete image produced by the lens, is this reducing the resolution of the lens? Well, I guess that the resolution of the lens is still the same but could one be using 'less' of the resolution as a lot of the image from the lens is outside the area of the film plane?
Not very well put I know but perhaps a summary thought could be; would the captured image get worse as the lens is focussed closer esp with tubes and bellows?

Thanks for staying with the rambley question - any thoughts/answers welcomed!
Sim2.

Click to expand...

Yes, the image circle gets larger as the lens is moved away from the plane of projection, which is precisely how the 'enlargement' of macro occurs.
That is why the 'size' of the object grows larger and larger as framed at the focal plane of the camera, and also why the exposure increases...the light is spread across a larger and larger area, diminishing the amplitude of light seen at a given point on the focal plane.

Think of the total resolution in line-pairs it can resolve across its image circle, so that total resolution is spread as its image circle is spread across a larger and larger area.

As for using 135 format lens vs. 4x5 format lens, if the 135 format lens is designed to give 120 line-pairs per millimeter resolution across an image circle of 50mm, and a 4x5 format lens is designed to give 50 line-pairs per millimeter resolution across an image circle of 150mm (at Infinity...the image circle is larger at the closer distances), the 4x5 lens gives 50 line-pairs/millimeter to cover 4" while the 135 format lens has only 30 line-pairs/millimeter also covering 4"!

wiltw said:

the 4x5 lens gives 50 line-pairs/millimeter to cover 4" while the 135 format lens has only 30 line-pairs/millimeter also covering 4"!

Click to expand...

Hmmmm, how do the lenses know which format is on the back? If you place a 35mm piece of film instead of the 4x5 sheet, does the lens revert back to a 120 line pairs per mm?

Kodak's 203mm f7.7 Ektar (and the earlier uncoated anastigmat version) is an excellent performer from Infinity to 1:1, it's a feature of the dialyte design.

Ian

cowanw said:

Hmmmm, how do the lenses know which format is on the back? If you place a 35mm piece of film instead of the 4x5 sheet, does the lens revert back to a 120 line pairs per mm?

Click to expand...

Lenses DO NOT CARE what is at the focal plane...they deliver the SAME RESOLUTION as originally designed to give.

But a given lens is designed to have either:
A) 80mm image circle with 120 line-pair/mm of resolution -- in the case of a 150mm FL lens for 135 format, or
B) 200mm image circle with 45 line-pairs/mm -- in the case of a 150mm FL lens for 4x5 sheetfilm.

That is why using an adapter to mount B on a 135 format film camera (with an adapter) will result in optically inferior results to mounting A on the same camera! B is designed to cover 6" diagonal or more, while A is not designed to cover much more than 2" diagonal

Apologies for the tardy return to my thread. I appreciate all the answers and points of views.

It is good to get some confirmation from the wisdom of crowds to my 'gut feeling' about the image circle getting larger as the focus point gets closer.
I reckon that my other thought of using the smaller format lens at a distance that enables the image circle to cover the larger format is rather like the 'how many angels on a pinhead' discussions in theological circles.
Without being able to calculate how many lines per millimetre for both lenses are actually on the film plane at the same focussed distance, it is all rather nebulous! One for a rainy afternoon when I've run out of chems.
I think that actually using a macro lens designed for the format is the easier route to getting better quality when focussed close - or it will, at least, give better quality than a 'normal' lens at the same focus point.
Thanks to all – I'm off to restart the counting as I lost track at 385,321 angels. I wish they would stand still.

Skiver101 said:

I have been scratching my head with something similar; as I wanted to know the pros and cons of using a very long (and relatively inexpensive) 135 format lens on a home-made camera with a 4x5 back. i.e. how much coverage could I expect to get on the ground glass with that combination.
So frustrating, I wish there was more info online for Andreas Feininger's 1200mm camera construction. If there was, I'd definitely give it a go.

JP

Click to expand...

Easy to see for yourself...hold the lens up to a wall so it focuses the image on it and measure. I just did a very quick test with 135mm Topcor (made for Topcon RE SLR), and it projects an image circle of only about 70mm across (diagonal)

I have been scratching my head with something similar; as I wanted to know the pros and cons of using a very long (and relatively inexpensive) 135 format lens on a home-made camera with a 4x5 back. i.e. how much coverage could I expect to get on the ground glass with that combination.
So frustrating, I wish there was more info online for Andreas Feininger's 1200mm camera construction. If there was, I'd definitely give it a go.

JP

Click to expand...

Do you mean something like this: Dead Link Removed

To add to what wiltw wrote, using a lens for a 35 mm camera isn't a good solution.

But assembling a long lens 4x5 is no big deal these days. Just get a process lens of the focal length you want and then cobble together a camera from used Cambo or Sinar standards, rails and bellows. I built one using Cambo bits and, instead of a real Cambo rail, a long piece of 80/20 1" x 1" t-slotted extrusion.

The longer the lens, the more expensive. Up to around 600 mm is fairly affordable, 1200 mm lenses are scarce and much more expensive.

If you want relatively easy-to-find and length, the best bets are Apo-Artars (coated only), Apo-Nikkors and Apo Ronars. My longest is a 900 Apo Saphir. Apo Saphirs longer than 360 mm are uncommon, my 900 was a very lucky find.

The biggest problem is a shutter. I have an industrial Compound #5 (no diaphragm) with an adapter that holds it in front of my 900.

^ Those damned Canadians Peter

wiltw + Dan Fromm...
Thanks for the response guys.
I think it's as I feared, there would be massive wastage of film area using a 135 format lens on a 4x5 back.

The process-lens route is what I had in mind; although I would have to break my meagre budget for a worthy focal length.
But cobbling together the ''camera attachment'' smile to the process-lens would be relatively straightforward.
I had planned on using the ''hat method'' anyway, so a shutter is the least of my worries.

...The dream endures.

Thanks for that !

JP

How meager is your meager budget? What is a worthy focal length? Are you aware of the practical problems of using long lenses?

Re camera attachment. Process lenses were delivered with mounting flanges. Used ones that have been well cared for are sold with their flanges. Given a lens, a flange to fit. and a board, bore a hold in the board large enough to accept the rear of the lens, drill holes in the board to accept bolts that pass through mating holes in the flange, and there you are. No special attachment needed as long is the lens is a bit smaller than the board. If the lens is larger than the board, then heroic measures are called for.

Re worthy focal length. As I said, process lenses up to around 600 mm are relatively inexpensive. Even a 600 mm lens requires a very stable tripod etc. One of my friends who's an exponent of long (up to at least 1200 mm) process lenses uses two tripods, one at each end of the monorail. Mr. Polidori's approach, visible in the link in post #16 above, is to use a stout tripod and a couple of Manfrotto Magic Arms. Magic Arm equivalents can be cobbled up.

FWIW, my failed Baby Bertha what had a 900 mm lens (and a 610 and a 480 and ...) sat on a tripod that was almost stout enough and was stabilized by a single Magic Arm. That was a good enough solution to the stability problem for me to move on to the next step in bringing up baby, test shots. Baby was unusable with lenses longer than around 250 mm and the back in portrait orientation because of the shape of my 2x3 Graflex RB's mirror box. Oh, well.

Before you spend a new p, think hard about the photographic purpose for which you want a long lens. I have a heap of longish lenses, have rarely hit a situation in which a lens longer than 360 mm or so makes sense on 6x9 or 6x12. I've shot birds and such with a 700 on a Nikon. All too often the subjects were still too far away. Getting closer is often better than getting a longer lens.

Sim2 said:

I knew that reversing a lens can be better than using it the right way around when getting around 1:1, but never really thought about why... ah, the joys of having a little techie learning!

Click to expand...

For a camera lens there are two planes where the image is in focus; the focal plane and the object plane. Camera lenses are designed so that focal plane is flat. However, except for some specialized lenses the object plane is not flat but rather curved. When dealing with flat objects like a postage stamp some people recommend reversing the lens so that the object plane is now flat. However this means that focal plane is now curved. When the lens is correctly focused this reversal has no effect on the resolution of the image at the center of the focal plane (one of the two the cardinal points). You are in effect just shifting the problem. It is extremely difficult to design a lens where both planes a truly flat and such lenses command a high price. Such lenses contain more elements with different kinds of optical glass. Some may use fluorite elements or even plastic ones.

''How meager is your meager budget? - What is a worthy focal length? - Are you aware of the practical problems of using long lenses?''

Think of zero budget, then half it. - I'm inspired by Feininger's long shots, so maybe 600-1200mm. - Yes, the practicalities are foremost in my thinking; that's why I ask so many seemingly dumb questions. But thanks for all your input, it IS appreciated.

Little by little, I approach my goal.
...Softly softly, catchee monkey.

JP

Gerald C Koch said:

For a camera lens there are two planes where the image is in focus; the focal plane and the object plane. Camera lenses are designed so that focal plane is flat. However, except for some specialized lenses the object plane is not flat but rather curved. When dealing with flat objects like a postage stamp some people recommend reversing the lens so that the object plane is now flat. However this means that focal plane is now curved. When the lens is correctly focused this reversal has no effect on the resolution of the image at the center of the focal plane (one of the two the cardinal points). You are in effect just shifting the problem. It is extremely difficult to design a lens where both planes a truly flat and such lenses command a high price. Such lenses contain more elements with different kinds of optical glass. Some may use fluorite elements or even plastic ones.

Click to expand...

Are you sure? I ask for two reasons.

First of all, I'd always understood that photographic objectives are designed to image a flat surface on a flat surface. We say plane of best focus because that's what it is for a perfect lenses. Lenses as made may suffer from field curvature (image a flat surface in front of them on a slightly curved surface behind them) but that's a fault, not a design objective.

Secondly, lenses can be optimized for only one film plane-to-subject distance. Lenses for general use are optimized for large subject in front of the lens, small negative behind it. When a lens is used at magnifications > 1:1, it has a small subject in front of it and a large negative behind. Reversing a lens made for general use when using it at magnifications above 1:1 makes best use of its optimizations.