Highest Sharpness & Resolution

[sanking]

it's not that it doesn't matter, but that it's a small part of the puzzle of what makes an image look sharp and detailed.

I agree that detail is a part of the puzzle, but IMO it is not always a small part. In contact prints and in moderate magnifications from LF film resolving power is probably not as as important as edge contrast. But as the amount of magnification increases the amount of detail in the negative becomes a very important issue if critical sharpness at a standard viewing distance is to be retained.

Sandy King

[Gordon Moat]

Hello Sandy,

A question for you.

. . . . . .

Regarding previous comments about Mamiya 7 versus 4X5 it should be noted that the magnification factor is more important in determining the capacity of these two systems to record detail than the total area of the negative. The long dimension of a 6X7 negatives is 55% of the long dimension of 4X5 and to capture the same amount of detail a 6X7 system would only need to resolve about 1.8 times as much as a 4X5 system, not more than twice as was earlier stated. Given the fact that it is not at all uncommon for lenses of the Mamiya 7 system to resolve over 100 lppm I think it not unreasonable to suggest that in many circumstances Mamiya 7 is capable of producing as much detail as 4X5 when using lenses with similar angles of coverage on the two formats. And especially so given the fact that for equal DOF the Mamiya 7 lenses can be used at apertures where diffraction is much less. There are of course other factors that determine final image quality, but if the issue is detail and nothing more my own experience is that Mamiya 7 is very close to 4X5, and may actually beat it in some circumstances.

Sandy King

I am using Linhof 56x72mm for my 6x7 comparison. When I place two frames from that on top of a single 4x5 frame, there is still room left over on the 4x5 frame, hence where I get the more than 2x difference. So is the Mamiya 7 frame size larger than the Linhof 6x7 frame? Or is there some other math I am missing here? Thanks in advance.

Ciao!

Gordon Moat
A G Studio

[sanking]

Hello Sandy,

A question for you.





I am using Linhof 56x72mm for my 6x7 comparison. When I place two frames from that on top of a single 4x5 frame, there is still room left over on the 4x5 frame, hence where I get the more than 2x difference. So is the Mamiya 7 frame size larger than the Linhof 6x7 frame? Or is there some other math I am missing here? Thanks in advance.

Ciao!

Gordon Moat
A G Studio

Gordon,

For the earlier comment I just based my figure on the difference between nominal values of 6X7cm and 5".

However, I just measured the actual long dimensions of some of my Mamiya 7 negatives and 4X5 negatives, and actual is even more favorable to the 6X7 format than nominal. My Mamiya 7 negatives are a full 7cm on the long dimension, whereas the 4X5 negatives are only 12cm on the long dimension. In other words, the 6X7 negative is 58% as long as the 4X5 negative, so resolution with the Mamiya 7 would only need to be 1.7 times 4X5 to give the same detail on film.

If you are actually getting 72mm with the Linhoff your comparison would be even more favorable, assuming equal lens quality. However, I really doubt that anything else out there in medium or LF will equal Mamiya 7 optics.

OK, that is how I did the calculations. Am I making a conceptual mistake of some sort?

Sandy King

[sanking]

resolution numbers also tempt people to think of the image chain in terms of weakest links ... like if the film resolves x lp/mm, then that's the limit of the whole system, and you won't see the benefit of a lens that resolves more than x. which of course isn't how it works at all.

Paul,

Williams makes exactly that point, i.e. that the weakest link in the image chain is the one that ultimately determines the amount of information that can be conveyed.

In the chapter on the Theory of Image Degradation he write: "It can be seen from the degradation equation that the minimum size of the photographic-spread function and the greatest image quality a system can produce will always be limited by the largest component-spread function in the system. When one spread component system is conspicuously larger than others, it imposes a resolution limit on the system."

Sandy King

[Gordon Moat]

Okay, using the long dimension:

3.175x for 11.9" by 15" print from 4x5
5.3x for 11.7" by 15" print from 4x5

So 1.67 difference. Assuming 60 lp/mm within reach from a 4x5, then a Mamiya 7 lens would roughly match at 100 lp/mm. I think that is within reach with most B/W films, based on C.Perez Tests. I have not seen Fuji Astia 100F or Kodak E100G (supposedly the two best) resolve much beyond 90 lp/mm (Erwin Puts published in LFI magazine a few years ago), so the comparison on colour film might be tougher.

Anyway, the comparison is very close, with 4x5 staying within the capabilities of the film though lens limited, and the Mamiya 7 lens being limited by the film. Then it would be a comparison on other aspects, such as movements capability on a 4x5, compared to arguably better edge resolution on Mamiya 7 lenses.

If the OP is still reading, this might give him another idea. If he does not need movements in the camera, perhaps a Mamiya 7 system might work better for him. With B/W film, he could potentially do better than most 4x5 systems and lenses.

Ciao!

Gordon Moat
A G Studio

[sanking]

Anyway, the comparison is very close, with 4x5 staying within the capabilities of the film though lens limited, and the Mamiya 7 lens being limited by the film. Then it would be a comparison on other aspects, such as movements capability on a 4x5, compared to arguably better edge resolution on Mamiya 7 lenses.


Ciao!

Gordon Moat
A G Studio

That is pretty much the conclusion I reached. Course, you must also factor grain in there since that is going to be 1.7 times as large with 6X7 as with 4X5, assuming the same film.

You could actually get the same DOF at equal aperture by using a much finer grain film with 6X7, and at this point I think you are about par with the two formats.

In order to get the equivalent of 100 lppm of detail out of a 6X7 negative you will need a scanner capable of "effective" resolution of 5000 ppi. You will need an Imacon or drum scanner for that. This assumes of course that you want to print at the highest possible size and still retain about 5 lppm of detail in the print (which would be about 36"X44"). Up to 18X22" effective scanning resolution of about 2540 ppi would be enough.

Sandy King

[dslater]

well, he was also comparing enlargements to contact prints, which is probably a bigger variable than the lenses, and almost as big a variable as the format difference.

I don't think you can conclude much at all about lenses from a comparison like that, except that contact prints from old lenses can look great (not really a news flash!)

He is not trying to make a comparison about lenses - the point he's making is that if you want sharper prints, then move up to a larger format so you enlarge your negatives less or make contact prints.

[Sheldon N]

Of course, the discussion of MF lenses having higher resolution goes out the window when simple movements like tilt come into play.

If you are shooting with a mild wide angle lens and taking a picture with a simple flat-foreground-to-infinity composition, having front tilt makes all the difference in the world.

To get an acceptably sharp (sufficient DOF) 16x20 print with everthing from 5 ft to infinity sharp, you need to stop down to roughly f/22 (assuming a 50mm lens on 6x7) to get everthing within DOF. At f/22, you are diffraction limited to roughly 60 lp/mm on film resolution so the "better" medium format lens has no advantage over 4x5, which can shoot at f/16 with front tilt and also put 60 lp/mm on a larger piece of film.

If you make the lenses longer, then the advantage to having tilt on 4x5 is even greater, since the MF lenses just have to stop down farther to get sufficient DOF.

This advantage doesn't hold true for all compositions and all scenes, but I would venture to say it is the primary reason why most of us choose large format instead of MF for our primary landscape camera.

[sanking]

This advantage doesn't hold true for all compositions and all scenes, but I would venture to say it is the primary reason why most of us choose large format instead of MF for our primary landscape camera.

I use LF primarily for perspective control, not control of sharpness.

With regard to the specific scenario you mention, let me throw in this. Some years ago I heard the photographer David Muench discuss his working habits on 4X5 with this type of near to far scene. My memory is that he stated that with near distance to far distance scenes he almost always relied on stopping the lens down to get adequate DOF rather than using tilts. Reasoning was that the use of tilts, which does indeed give sharpness from near to far on one plane, may grossly distort images on the other plane.

Sandy King

[paulr]

Paul,

Williams makes exactly that point, i.e. that the weakest link in the image chain is the one that ultimately determines the amount of information that can be conveyed.

In the chapter on the Theory of Image Degradation he write: "It can be seen from the degradation equation that the minimum size of the photographic-spread function and the greatest image quality a system can produce will always be limited by the largest component-spread function in the system. When one spread component system is conspicuously larger than others, it imposes a resolution limit on the system."

Sandy King

The flawed thinking I'm talking about is this: people assume that if you've identified the weakest link in the image chain, then the other (stronger) links cannot degrade the image farther. They can. All parts of the optical system degrade the image. Higher quality parts of the system degrade it less. Final image quality is not determined by the weak link, but by the sum total degradations of all the links.

There is NO WAY to predict the optical quality of an optical system by looking at resolution numbers of its individual parts (however you choose to measure them). Unless one of the parts is radically weaker than all the others put together. But MTF, since it shows percentage of contrast lost, makes this easy. you just multiply.

Williams goes into some detail on this, but all it takes is general understanding of MTF theory see how it all works.