Evaluating lens performance <> a practical approach?

[l2oBiN]

What is the best method to quickly get an idea of what the lens coverage is and how good is the performance of that coverage. I have a 8x10 camera at my disposal and would like to evaluate the coverage and performance/quality of the projected image circle from the centre to the edge.

This came about through an observation that a pair of lenses which according to specified image circles should not have much movement on 8x10 seem to have considerable shift/rise capabilities. I realise that some of that could be attributed to the circle of illumination rather than the image forming circle and thus would lik to quickly evaluate whether this apparent bigger projection posses good image quality.

[E. von Hoegh]

Tape a piece or pieces of newsprint to a wall at a distance of least 10x the focal length of the lens, make a negative and then inspect the negative under high magnification. Be very careful to use filholders which are consitent and accurate, focus with a loupe at the working aperture, etc.
You'll find that sone lenses will illuminate considerably more than they cover sharply, while others cut off just beyond the sharp zone.

[l2oBiN]

Could this be simply done with a loupe on the gg without using up negatives?

[Dan Fromm]

No. The GG is too coarse.

[E. von Hoegh]

Could this be simply done with a loupe on the gg without using up negatives?

What Dan said. But as long as we're on loupes, make certain yours is focussed sharply on the GG; check this without a lens on the camera by illuminating the GG with a lamp or pointing the lenseless camera at a widow.

[ridax]

Glass matted mechanically with a too hard an abrasive (diamond, SiO₂) comes out too coarse indeed. Al₂O₃-matted ones are reported to be far better (though I've limited experience with those). But a proper chemically etched ground glass has more then enough quality for testing lens sharpness.

Smaller format (including 4x5") camera backs and SLR viewfinders are indeed not usable for tests like that, but for a different reason. The negative is just going to be enlarged quite a bit more then a loupe can magnify still not showing too much of the loupe's own aberrations. But with 8x10", a good loupe is just OK. (I use a number of short focal length cine lenses for my loupes as those used to be a dime a dozen not long ago - just until the digital guys and gals began to put them on their NEXes.) How much would you enlarge you 8x10"? Some 4x already yields a mural that would probably be looked at from a couple of yards. In that mural, perhaps nobody would ever notice any sharpness defects that were not visible on a quality ground glass through a quality loupe.

A more accurate but still pretty fast method of testing is watching an enlarged image projected by the lens. I use an enlarger for shorter focal length lenses and horizontal projection for the longer ones, putting a cold lamp inside my camera. Just holding a small (medium format) fluoride-etched ground glass in one hand and a loupe in another, move into the sharp-focused zone of the projected image - and voila, you see all the aberrations and sharpness and contrast and all the rest that one may ever want to see. No film processing involved.

I've just noticed a picture of a factory QC operator testing lenses via a projector system in a Leica brochure one day. I thought if the method was OK for Leica (and Leica apparently thought the method was more then just OK; they looked being proud of using it) then it was good enough for me, too. And in the decade that passed since that day, I was never disappointed.

And speaking of the ground glass - it should be a real GROUND GLASS, not any sort of bright screens. Bright screens are too transparent and make a lot of the so called aerial image visible though them, and the aerial image is not a thing you want to see to get the idea of the resulting photographic sharpness. A ground glass usable for sharpness tests is very fine in grain and also very dense, probably denser then the one you'd find the most comfortable to use in the field.

And yes the grain of the glass is very visible through a loupe. But no it does not interfere with observing the lens sharpness when the image you watch is magnified quite a bit in projection.

[Bob Salomon]

Do you want to do copy work of flat objects or use it for scenics? If flat field copy work do the news paper test. Or just shoot a brick wall. If for scenics go shoot a scene with fine detail across the field.

If you want to know what the lens covers:

Point the lens out a window. Darken the room. Hold a large, white piece of cardboard behind the lens (open the lens and set it at the largest F stop, move the board towards and away from the back of the lens till the image on the board is sharp. measure the diameter of the circle on the board. That is the coverage at infinity.

Note, the circle of illumination will be larger then the acceptable coverage so look carefully at the circle that you measure. As you stop down the circle will become larger.

[Jim Jones]

Decades ago I improvised a projection system for testing 35mm camera lenses with a projector as a light source, a micro copy of the entire Bible in a 2" square chip for a target, and a variety of lens mount adaptors. The Bible was then available from Edmund Scientific, and I found some recently online from another source. While the detail in the text of that Bible isn't perfectly filmed, it is fine enough to show lens performance shortcomings that probably won't be noticed in practical 35mm photography. The results of such a test are more subjective than objective, but so is most photography. The same method of testing should be applicable to any format, once one buys or makes a suitable target. Sanding a rigid sheet of Plexiglass in a random pattern with fine sandpaper might make a useable target for subjective testing in any format.

[ridax]

Jim, excellent point on this type tests being subjective. For visual tests, a well-known lens is nearly a must for reference - even after years of practical experience.

As for the test target, I prefer a real life object to any image on film. Whatever the quality, an image is never perfectly sharp itself; while usable, it's certainly not the best of test targets. I actually use a mesh-like cloth ribbon with wide gaps between the threads - the type little girls use in their hair braids - but a black one. The crossing threads are the bigger 'lines', and individual fibers are the finest ones... very handy.

I also prefer my test target not to be flat. I put it at an angle to the lens so I can easily find a piece of sharp image without trying hard to get into focus. Besides, this way I can see both the foreground and the background defocused zones to the left and to the right of the sharp zone - all at the same time.

I include a transparent ruler with my black ribbon to be able to see how far off-center is the part of the image I am looking at.