Teach me to see soft corners...

[Tim Meisburger]

Why am I constantly surprised by soft--or non-existent--corners in my negatives? I don't use extreme movements much, but when I travel I often shoot what's available, and that is often architecture. I just came back from a business trip to the US an have several negative that, but for soft corners, would be nice. In one, I had an old mission church with a fence and gate. The gate and front door were not aligned in reality, but sneakily I used shift (I never use shift) to line them up and make a pleasing (or boring) symmetricallity (is that even a word?). Bam - I develop the negative and the whole upper left corner is gone. Why don't I see it? And how can I improve my technique so I do? (I'm already plotting to replace my lenses with some with better coverage).

Thanks for suggestions

[djdister]

How confident/cognizant are you of the parallelism of your front and rear standards? Unintentional tilt or swing (or both) can lead to soft corners. If two adjacent corners are soft, it could just be off from one camera movement (i.e. tilt), but if only one corner is soft, or opposite corners are soft, then you have affected two planes (tilt and swing) at the same time.

[Tim Meisburger]

Unless I have forgotten to zero after a tilt (I did once), I don't have problems with parallelism. My problem seems to be seeing what is plainly on the ground glass in front of me. Maybe I need one of those square loupes to look in corners, but I suspect I just need to focus (my mind) less on the subject and more on the corners.

[Mark Sawyer]

Tim, I'm not sure if your camera's ground glass has clipped corners or not, but that's what they're there for! Peek through each corner and see if you can see the whole aperture opening at the shooting f/stop. If you see the whole aperture, you're good; partial aperture circle and you're into partial illumination; no aperture visible, no light hitting the corners.

[Old-N-Feeble]

It's possible the front or rear standard is moving between the time you compose and make the exposure... perhaps the rear standard when inserting the DDS? How sure are you the movements all lock solidly? Is it possible the GG isn't seated properly?

[Doremus Scudder]

Tim,

First, and something I don't see mentioned very much, is that when you use movements, stopping down will often get you less coverage than when viewing wide open. This occurs because, when wide open and using, say, lots of shift or rise, the farthest corner gets fairly well illuminated by the part of the lens aperture opposite it and farthest away (I hope this is making sense...). When you stop down to taking aperture, this side of the aperture basically goes away, leaving the corner in the dark. This what I suspect to be your problem.

A similar thing happens when using filters in front of the lens; stopping down can often lead to vignetting from the filter ring.

So, when setting up a shot like this, you need to remember to check with all filters in place and with the lens stopped down to taking aperture. (Reread the previous sentence for emphasis).

Check by either looking through the clipped corners of your ground glass and making sure you can see the entire lens aperture, or, by looking back through the lens aperture and making sure you can see all corners of the ground glass. If you can't, you'll vignette.

As far as "soft corners" go. Most modern lenses are mechanically vignetted so that if you have coverage, you'll get a relatively sharp image. Older lenses, however, often have a larger circle of illumination than sharp image circle of coverage. If you're using older lenses (even some of my Ektars are really soft at the extremes), you'll just have to shoot with that awareness. A soft corner in a dark blue sky makes little difference, but if there is important detail there, you may want to reconsider.

At any rate, if you think you have lenses that are soft at the extremes of coverage, you'll have to check the sharpness on the ground glass at the taking aperture, and then later on the neg itself till you get a feel for how much you can stretch coverage for a particular lens.

Hope this helps,

Doremus

[Tim Meisburger]

Thanks Mark. That makes sense, and explains why you get greater coverage as you stop down (as the aperture moves towards the centre there is less mechanical vignetting from the lens barrel). Slowly I figure this out.

I'll have to practice checking the corners and including that in my normal routine. (plus buy lenses with more coverage).

[Tim Meisburger]

Thanks Doremus, that explains even more. So the whole glass can look illuminated at 6.8, then I close the lens, insert the holder attach a yellow filter, then stop down to 45 or 64, and at this point I am introducing mechanical vignetting from the front of the lens. Interesting, and illuminating:).

[Doremus Scudder]

Thanks Doremus, that explains even more. So the whole glass can look illuminated at 6.8, then I close the lens, insert the holder attach a yellow filter, then stop down to 45 or 64, and at this point I am introducing mechanical vignetting from the front of the lens. Interesting, and illuminating:).

What can happen is that the corners are mostly/partially illuminated when the aperture is wide, but a smaller aperture just doesn't "see" the corner of the film. Try shifting to an extreme and then look through the front of the lens at a corner as you stop down and you'll see what I mean. Filters can make things worse, but even without one, you can clip a corner pretty easily if you forget to check when using extreme movements. I make the mistake every now and then myself when I'm not paying enough attention, so don't feel bad.

Best,

Doremus

[Dan Fromm]

Doremus, I'm sorry, but I don't agree with you about the effect of stopping down on mechanical vignetting.

I hang some of my lenses in front of a leaf shutter, so am always at risk of mechanical vignetting by the rear of the shutter's tube. Stopping down makes the lens' exit pupil smaller, increases the angle from the edge of the exit pupil to the rear of the tube. This is why I don't agree with you.

I also have a tiny 45/9 CZJ Goerz Dagor in barrel. This little gem has (choose one) too small outer elements or too narrow a tube. Either way, stopping it down makes the exit pupil smaller and increases the angle from the edge of the exit pupil to the rear of the barrel. Another reason not to agree with you.

[Jac@stafford.net]

A minor question: are some (especially modern) lenses built with intentional mechanical vignetting, possibly to enforce reasonable coverage claims?

[Doremus Scudder]

Dan,

Think of it this way; let's say you've shifted enough so that the center of the lens can't "see" a corner of the film. However, you've got the lens wide open, and the aperture is wide enough so that the outer parts of the iris do "see" the corner. Then you stop down... oops, suddenly the part of the aperture that was illuminating the corner is gone. Result, dark corner. I've observed this often in the field; (partially) illuminated corners when wide open, dark corners when stopped down.

Maybe it's lens-specific, but, think about angles for a minute. The angle from the far side of a wide-open aperture to a corner on the opposite side is surely more acute than one from the center of the lens. If there is anything in the way that is mechanically vignetting the center part of the aperture already, stopping down can only make things worse.

That said, I'm going to go double-check to make sure, since I respect your knowledge and expertise, especially when it comes to lenses. However, I've drawn the scenario out on paper and it seems to confirm my observations.

I'll report back.

Doremus

Update: Just checked with my 90mm SA f/8: At extreme rise, and when looking back through the lens aperture for the corners of the ground glass, I can only see the far corners through the edge of the iris wide open. When stopped down to f/32, the corners are not visible at all: vignetted. If I were to introduce a filter into the equation, the effect would be much more pronounced.

Doremus

[Vaughn]

Seems I have observed the opposite, Doremus. Looking thru the cut corners, with the lens wide open, one can see vignetting (the aperture is partly blocked), but as one reduces the aperture, one can see the entire aperture -- no vignetting.

The corner does not get darker as one closes down relative to the rest of the GG. In fact, the corner gets brighter relative to the rest of the GG since only the corners were getting vignetted.

[Doremus Scudder]

Vaughn,

What you observe is true as long as light from the lens' center is still reaching the corner in question. In that case, the near edge (i.e., on the same side as the corner) is vignetted wide open, but not when the lens is stopped down. And, as you say, the corners in question end up being illuminated more evenly.

Carry the thought experiment a bit further though, to where light from the center of the lens aperture and then some is being blocked and not reaching a corner or corners (due to rise/shift or a combination), but a crescent of the wide-open aperture is still able to illuminate the corner(s) in question. So, when observing the ground glass, the corners are illuminated, albeit only partially. The darker corner, however, is not always so readily apparent, especially when dealing with a hot spot, wide angle lens, etc. Now, when you stop down, you'll effectively eliminate the illumination from this crescent, leaving only light coming the center of the lens, which never reached the corner in the first place, thereby darkening the corner entirely.

Plus, it's really common to be on the edge of coverage like this and then add a screw-on filter to the front of the lens without checking. The effect I'm speaking about is only exacerbated when a filter comes into the equation.

I hope I'm making sense here. And, I think this could be Tim's problem.

Best,

Doremus

[mdarnton]

What I want to know is what exactly are the lenses that are causing this? It doesn't sound like movements are extreme, so maybe you just need to buy lenses suited to the job. I know I have a couple of lenses where my bellows would tie in a knot before I'd run out of coverage.

[Vaughn]

Now, when you stop down, you'll effectively eliminate the illumination from this crescent, leaving only light coming the center of the lens, which never reached the corner in the first place, thereby darkening the corner entirely.

No, if you can see the entire opening of the lens thru the cut corner, that corner is receiving all the light possible thru that aperture. If you see only the cresent, then you are not seeing all the light that is passing thru the lens (while all the light is still hitting the center of the GG). The light from the center of the lens does hits the corners.

If you add a filter that vignettes, you will see the filter thru the lens via the cut-corners.

"Hot spots" (brighter areas on the GG than the surrounding areas of the GG) are not due to the amount of light hitting the GG -- but instead it is the light hitting the GG and not all going towards our eyes. The film will not 'see' that hot spot.

[Doremus Scudder]

No, if you can see the entire opening of the lens thru the cut corner, that corner is receiving all the light possible thru that aperture. If you see only the cresent, then you are not seeing all the light that is passing thru the lens (while all the light is still hitting the center of the GG).

If you add a filter that vignettes, you will see the filter thru the lens via the cut-corners.

"Hot spots" (brighter areas on the GG than the surrounding areas of the GG) are not due to the amount of light hitting the GG -- but instead it is the light hitting the GG and not all going towards our eyes. The film will not 'see' that hot spot.

Vaughn,

We're saying the same thing. I, myself, have made the mistake of thinking my lens was covering when it was not and the corners were only illuminated by a crescent of the aperture when wide open. This then disappeared entirely when I stopped down, resulting in real vignetting, not just light fall-off in the corner. I'm talking about a situation here where the photographer is not checking through clipped corners or by looking back through the taking aperture at the corners. Certainly, if the OP had been checking, he would have seen the vignetting. It's when you're not checking and not really aware that that detail in the corner is only coming from a sliver of the aperture that this mistake is easy to make.

Best,

Doremus

[Tim Meisburger]

I think what Doremus says makes sense geometrically, as changing the location of the far edge of the aperture will change the angle that light passes through the barrel. Of course, not checking through the clipped corners is my error, but I can certainly imagine seeing part of the aperture when wide open and none of it when stopped down. This seems to me the most likely scenario, as I was applying extreme movements for the lens in question (yes, I need to buy lenses with more coverage), and evaluating the whole image on the ground glass wearing readers, then buttoning up and stopping down to 45 or 64, then adding a yellow filter.

This discussion has been very illuminating for me, and my takeaways are:

- when using movements, check the clipped corners with the lens stopped down to taking aperture and with any filters attached
- buy lenses with more coverage

Thank you gentlemen. Its 2:45 am in Bangkok (I'm still jet-lagged from my US trip), but at least I am making productive use of my time!

[Robert Opheim]

There is mechanical vignetting from lens edges, filter edges, and lens hoods edges - these can be checked for (as said previously) by checking through the cut corners of the ground glass - after you stop down - you should see a full circle of the aperture, and not one that is cut off. Also, there are lenses that get "soft" toward the edge - or said differently the focus goes down. My lenses are from the 1920's through the 1980's or so. Some of my older lenses do get soft toward the edge of coverage. I believe the later lens designers cut back the area of the "circle of confusion" from the edge of lens design through mechanical means. This is different than light fall-off toward the edge - that is compensated for in wide angle lenses by center filters. I have heard as one lens with focus fall off at the edge - that some photographers have mentioned angulon lenses having more focus fall off at the edge of coverage - otherwise these can be very sharp lenses - as documented by tests. It seems that out of focus edge areas in prints are dependent on how large a print is that you are making. 4x5 to 16x20 prints show more issues than a smaller print. A larger print will magnify visual perception of out of focus areas more.

[Robert Opheim]

Tim, I re-read your original question where you used shift to correct the composition. If your front standard and rear standard were parallel than only the lens edge focus quality could be involved. If you used swing (which I have done) and used too much swing, than it can really change the plane of focus - and throw another area out of focus area.

[Tim Meisburger]

Thanks Robert. On the particular image the problem was rise and shift combine, leaving one corner blank. I've been confused because I wondered how I could not notice on the ground glass something so apparent in the negative, but I think Doremus explained that; decrease in image circle due to mechanical vignetting from filters or stopping down when the camera is adjusted to the limits of lens coverage (where these factors would not decrease coverage if the image was entered).

[Vaughn]

Vaughn,

We're saying the same thing. I, myself, have made the mistake of thinking my lens was covering when it was not and the corners were only illuminated by a crescent of the aperture when wide open. This then disappeared entirely when I stopped down, resulting in real vignetting, not just light fall-off in the corner.

Best, Doremus

Yes, perhaps we are just thinking differently on the same subject. Next time I get the camera out, I'll see if I can duplicate your description of what is happening as I have never observed it in actual use (going from a crescent to no light at all when stopping down). I do not think closing down the lens could cause physical (real?) vignetting -- but certainly enough to get a corner not exposed enough to register a tone. It will be interesting to check it out.

But I do not use short lenses, so that may be the reason. If all one is seeing a thin crescent when wide open, that is some serious vignetting going on -- and that closing down probably will not cure!

[Doremus Scudder]

Yes, perhaps we are just thinking differently on the same subject. Next time I get the camera out, I'll see if I can duplicate your description of what is happening as I have never observed it in actual use (going from a crescent to no light at all when stopping down). I do not think closing down the lens could cause physical (real?) vignetting -- but certainly enough to get a corner not exposed enough to register a tone. It will be interesting to check it out.

But I do not use short lenses, so that may be the reason. If all one is seeing a thin crescent when wide open, that is some serious vignetting going on -- and that closing down probably will not cure!

Vaughn and Dan,

Here's the phenomenon I'm speaking of. I mounted a yellow filter on my 90mm Super Angulon f/8 (a lens that I run out of movements with more often than I'd like...). I then mounted it on a camera, removed the back and racked the bellows all the way back to reveal the rear element and exit pupil easily. I then took photos from an extreme angle, but not more extreme than could be achieved with movements (I have vignetted with this lens without a filter before).

The first photo is shot with the lens wide open, as it would be for viewing/focusing, and shows the filter ring impingeing on the exit pupil, but not completely covering it. There is still a sliver of light that would be illuminating the corner (it's not really a crescent, but whatever that shape is called created by two circles intersecting...).

Photo two is simply the same shot from the same location, but with the lens stopped down to f/16. The sliver of light that was illuminating the corner is all but covered by the iris now. Stopping down to f/22 or further would eliminate it entirely.

And yes, the first photo represents "some serious vignetting," however, when working under the dark cloth, it is amazing how much detail one can still see in the corner, even with 50% or more of the light from the aperture being blocked (that's just a stop or two...). It often appears that the corner is fine, just suffering from light fall-off. However, when stopped down to taking aperture, the corner goes dark (remember, we didn't check through the corners or back through the lens for vignetting because we thought the corner was okay...).

This happened to me too many times in the field until I figured it out. Now I check :)

Best,

Doremus

128814128815


P.S: For the curious, the shape created by the intersection of two circles is called a vesica piscis or a mandorla http://en.wikipedia.org/wiki/Vesica_piscis