Does LF handle as much enlargement factor as we all think?

[Bill Glickman]

A while back, Danny B started a thread on difference in resolution between LF an d MF. Pat Chase stated that good human vision can only detect about 7 lpmm of r esolution, anything less would be viewed as blurred, or poorly focused. Anythin g greater would be perceived as sharp. (He has a great deal of experience in t his area) Assuming great LF lenses can only deliver on 56 lpmm to film, when en larged, assuming no losses in the process, a 3x enlargement would yield on paper  7 lpmm. Therefore a 4x5 chrome enlarged only to 12x15 viewed up close, sa y 1 foot, would be the max. enlargement this print can withstand before it appea rs to be blurred or poorly foccussed. I realize a prints viewing distance shou ld be equal to the diagonal of the print. But thats not usually the case, spec ially with LF shots, everyone wants to inspect the detail up close! So for this example, I am considering how much scrutiny a print can handle up close.

After viewing Chris Perezs LF lens tests (thank you Chris for sharing this ver y useful information with us) linked to this home page, it is clear that many le nses deliver poorly on the edges or sometimes even in the center at certain f st ops. If the lens delivers only 30 lpmm to the edges, hence the weak link in the chrome, which now allows for only 2x enlargement max. before it becomes perceiv ed as not sharp. Is this math correct, or is there some other missing factor? Could this explain the erratic behavior LF sometimes delivers on 4x5? We could very well be experiencing very low LPMM due to shooting at a poor resolving f s top on the lens, using poor optics, poor focus, etc. Any one of these variables can produce much poorer results then we would expect in LF, specially 4x5, 8x10 is much less prone due to the same resolution at 2x the film size. Any input w ould be helpful..

[Wayne_6692]

Can't understand your math. Onfilm resolution of 56lpmm will allow for an 8x enlargement (56/7= 8).

[Bill Glickman]

Sorry Wayne, maybe this is where I went astray in my never ending quest for sharpness.... but if you start with 50 lpmm on film and enlarge 2x, it was my understanding the resolution on the enlargement will be 25 lpmm, or cut in half. If you enlarge again, for a total of 3x, this will cut the 25 lpmm to 12.5 lpmm. In other words, each time you double the enlargement you halve the resolution. Do you disagree with this?

[Patrick Chase]

Hi Bill;

Are you dividing by the enlargement factor squared?

Line pairs per millimeter is a linear measure of resolution (i.e. it measures resolution along a single axis rather than the amount of information contained per unit area), so an LF original with 56 lp/mm of "perfect" resolution (i.e. MTF is close to 100% all the way out to the 56 lp/mm limit) would produce exhibition-quality results when "perfectly" enlarged up to 8X (=56/7), or ~32" x 40".

There are two catches though:

1. Stated lens resolutions are usually the perceptible resolution limits, which typically exhibit MTF in the 10-20% range, if that. The oft-applied Rayleigh criterion for diffraction-limited resolution yields about 18% MTF at the calculated limit (of course, the Rayleigh criterion also assumes point sources, so it's open to question whether it's really applicable to this discussion at all).

2. Enlargement is _never_ perfect (though with modern techniques it's getting pretty close!).

Regards,

Patrick

[Wayne_6692]

Bill; Going from 2x to 3x doubles the area of the image, but it is only a 50% increase in linear dimensions. The lpmm is a linear function.

Wayne

[Bill Glickman]

Wayne, Pat, totally understood. This was a classic case of confusing myself - after all these years of doing the math correctily, somehow I just started doubted myself....Not sure how I tricked myself into this one... Who says we can not show signs of senitlity at 40? I'm living proof... Thanks guys... Bill

[Glenn Kroeger]

Two other factors to consider. The 7 lp/mm is an extreme number. Consider that most people find LightJet images exceptioally sharp looking, and they are only imaged at 12 pixels/mm which means nothing beyond 6 lp/mm gets on paper. At my age (and yours) I am probably scraping along at 3 lp/mm!

The other factor is film. While I agree that some medium format lenses are very sharp, color films are pretty much shot by 50 lp/mm. Not that you can't squeeze out slightly more resolution, but by 50 lp/mm the MTF contrast is dropping precipitously, and so edge sharpness is gone.

I am finding with digital work that edge contrast is much more important than resolution in producing perceived sharpness. I am astounded at how sharp some low resolution images can appear if careful unsharp masking is used to bump up edge contrast.

So you can make good 30x40's from 4x5... but you can't look at them with a jewelers loupe!

Glenn

[Ellis Vener]

There is another set of non theorectical factors to consider as well: the enlarg ing lens and the enlarger. Also considerthat lens tests like those that ChrisPer ez performed or only really valid for thoise specif lenses. The next one in the pro duction cycle will undoubtedly perform differently. the difference may be slight or it might be 2x slight (assuming that the manufacturer has a +/- 10% range of acceptable performance, if one lens is near one extreme and another lens is at the other this could nearly be a 20% difference between two "identical" lenses.) The moral of my story is that theorectical math will only take you so far, to get the rest of the story you have to test your own equipment. Ideally in both a laboratory and real world context.

If your "real world" photographs aren't as sharp as you think they should be you need to examine all parts of your system, starting from the ground upwhat are the qualities of of the gross parts of the system: tripod, head, camera bodies, lenses, film holders. Then see where these systems couple: tripod to ground, tripod to head, head to camera, tripod to head to camera, camera to lens, ground glass alignment to camera body, groundglass to film plane, etc. perhaps there is a weakness in your system? Something that induces a slight bit of vibration?

[Ellis Vener]

Some corrections of some pretty gross errors: There is another set of non- mathmatic factors to consider as well: the enlargin g lens and the enlarger.

Also consider that lens tests like those that Chris Perez performed are only really valid for thoise specific le nses. The next one in the production cycle will undoubtedly perform differently. the difference may be slight or it might be 2x slight (assuming th at the manufacturer has a +/- 10% range of acceptable performance, if one lens i s near one extreme and another lens is at the other this could nearl y be a 20% difference between two "identical" lenses.) The moral of my story is that theory will only take you so far and to get the rest of the s tory you have to test your own equipment. Ideally in both a laboratory and real world context.

[Pete Andrews]

Leaving aside the errors in Mathematics, I think the original question was a very valid one. I was glad to hear that I'm not alone, in that I have never had a very high opinion of the optical qualities of LF lenses, and I have used some of the best (Symmars, Nikkor-Ws, Grandagons). Critical examination of LF negatives through a high-power magnifier has always fallen short of what is obtainable on MF, or even 35mm.

The issue, IMHO has always been one of tonality, rather than absolute sharpness. Until the introduction of T-grain type films, 35mm just couldn't compete in terms of smoothness of tone with MF, and MF was always a rung down from LF. Now T-grain has changed all that.

I can now get the sort of quality from 35mm that I used to get from MF, and from MF what I used to get from 5x4, but it stops there. Using T-max or Delta in my 5x4 doesn't give me 10x8 quality, it simply shows the shortcomings of the lenses more clearly. In fact, if lens movements aren't an issue, MF can give better results. I do still use LF, but less and less these days, and more for the leisurely way of working, than any real difference in quality.