Real image circle for 80mm XL

The 80mm XL lists an image circle of 212 mm at f22, and a max rise of 34mm portrait. I am trying to test this on my Ebony 45SU, which has a max rise of 70mm. I set the camera at max rise and film in portrait orientation to map out the vignetting and to see how much difference various size filters make. Much to my surprise, at max rise I do not see a well defined vignetting line, just a gradual fade over the top couple of centimeters. With step up rings and filters (72, 77, 82) I see a sharp line from about 3 to 2 centimeters from the top edge of the film. This is still better than the stated image circle. What is going on with this lens? Is the stated image circle really conservative, or is there image degradation toward the edges I am not seeing since I am shooting a blue sky? Someone should stick one of these on an 8x10 and see what the full image circle really is.:-)

This particular lens does have some drop off towards the edges (I have one). It's supposed to cover 8x10, but I haven't tried it myself. I normally use a center filter with that lens.

It won't cover 8x10 at normal distances.

It does not drop off like a light switch so you can shoot sky with lots of rise and never see the image degrade.

You can infer what is going on looking at the modulation transfer function of the lens. Below is the MTF for infinity focus at f22 for the 80mm SSXL. On the X-axis 100 represents 106mm which is the specified image circle (212mm diameter) in radial distance from the center. I added a red line to indicate the edge of this IC. The blue line represents the radial distance to the corner of the 4x5 format with zero movement.

The curves on the graph are modulation transfer in the radial (solid line) and tangential (dashed line) at 5, 10 and 20 cycles per mm. This basically represents the percentage of contrast transferred by the lens to the image (y-axis) as a function of distance from the center of the image (x-axis) and frequency (curves).
General lens characteristics inferred from the graph are that 1) MTF decreases with distance from the optical center and as frequency increases and 2) as you move away from the optical center of image you lose resolution in the tangential direction much faster than the radial direction (that's just the way it is).

As to the original question, it is clear from the graph that this lens performs extremely well at f22 with no movements. It would be difficult to see the loss of resolution at the very corners even with at a very big enlargement. So no compromises for the 4x5 format with no movements. At the edge of the image circle there is a substantial loss of MTF at high frequency. 5/mm, 10/mm and 20/mm MTFs have fallen to approximately 75%, 50% and 20% respectively. So at the specified IC you've already lost a lot of your high end and everything is heading south fast if you keep going farther out.

People will argue about what the actual IC of a lens is relative to the specified IC, about whether the spec is "conservative" but that becomes very subjective. If you look at the MTF charts (assuming they are available) the intent of the manufacturer is clear and you can determine what that means for your own application. For highly technical applications where detail to the very corners is critical I would call the specified IC overly liberal and try to shift the film as little as possible. For contact printing or very modest enlargement I would say the specified IC might be overly conservative and go right up to it or farther. Certainly if you are just rising up into the sky you can get away a lot. I think I've done about 4 cm rise in portrait orientation without blacked out corners. I would not use 4 cm of fall without expecting the bottom foreground to go noticeably soft.

Just remember that even if you can illuminate a corner it doesn't mean the lens performs acceptably there.

Thanks! I was surprised because some of my lenses mechanically vignette near their IC and you get a clear end of the image circle. This one does not, but as you point out, unless there is sky on the edge, you lose a lot of detail, esp. when you get beyond the 212. I estimate that the light fall off started to be visible at 5.5 cm, but that is 2 cm beyond the image circle. So the answer with this lens is to be careful to stay with moderate movements - 3 cm or less rise in portrait, and use my 72mm XL for the extremes. My goal with this lens is to have a light lens set for carrying around, and keep the 72mm XL for working close to the car. A main saving is that with the 80, 110, 150, and maybe 300 f9 I can use 72mm filters on adapters, and not have to carry the whole Lee kit I need for the 72mm.

Ed,

Do you see more IC from the 110XL? I know the 90XL covers 8X10 with very little vignetting and makes excellent photographs on the 8X10 sheet of film. Interesting how well your 80XL does.

The 110 is like my 120SW - it allows greater movements than my 4x5 camera can do. I never run out of coverage. What amazes me is that it is a stop faster and 1/2 the size of the 120. Plus, according to the manufacturer, it has very good close up performance. I have not had a chance to test that.