Focusing Consumer vs Pro SLR's

[Timmyjoe]

Aizan, Don't know that collimator. I use an old Richter Collimator.

Correct, the focus screen has to be exactly the same distance optically from the lens flange mount as the film plane is mechanically from the lens flange mount. It has nothing to do with contrast. It has to do with the beams of light coming through the lens and focusing at a certain spot (or a certain range known as the depth of focus). On a wide angle lens, or a normal lens with a really big aperture like f1.2 or f1.0, the depth of focus is really small, so the focus screen and film plane have to be precisely at the same distance. With a lens stopped down to f5.6 or more, or with a long telephoto lens (both having a depth of focus that are very large), the focus screen and film plane don't have to be as precisely set.

Best,
-Tim

 

[Steve M.]

Not to doubt you, but no camera I have ever seen, still or motion picture, needs that sort of tolerance. Not unless you're shooting w/ a .95 lens or something. Motion picture cameras focus totally differently than a still camera, which is why they require a second person for follow focus. They are focused by distance, not through the lens, unless you're talking about something small like a Bolex. Besides, there is no difference between a "pro" film camera and a consumer camera in this regard. I have a Leica R lens that came to me w/ a shim (and a thick one at that) missing from the rear element. I was able to find one close enough at a hardware store. Before I installed it, I shot some test photos w/o any shim and they were out of focus wide open, but by f8 everything was tack sharp, even though the lens absolutely would not focus to infinity w/o that shim. f8 will cure a multitude of sins!

 

[Timmyjoe]

Steve, I'll point out the one thing you got right. By f8 everything was tack sharp. Yes, by f8 everything is sharp.

Best,
-Tim

 

[BillBingham2]

Not to doubt you, but no camera I have ever seen, still or motion picture, needs that sort of tolerance. Not unless you're shooting w/ a .95 lens or something. ........

Steve,

There are many more cameras that have been around over the years than I think you, me, and everyone else registered at RFF have seen. Lots of special government and scientific work has been done with cameras that would need that level of tolerance.

Tim,

Excellent site. Good to see you are a surviver. Stay healthy.

B2 (;->

 

[Timmyjoe]

Thanks Bill.

 

[Timmyjoe]

Now I just need to find someone who can do this on my FM and FM2n. I've always had a little suspicion that my pics from those cameras weren't as sharp as the ones from my '59 F. I always chalked it up to pilot error.

Vic,

One way you can test your two cameras for yourself is print up some focus test targets, like the one here:

Focus Test Pattern

Mount them on foam core or something similar. Then place them about fifteen or twenty feet from your camera. Stagger them about six inches apart, so the middle "Focus" target is at a set distance, let's say twenty feet, and from the camera's POV, place a target to the left of the "Focus" target at nineteen feet, six inches, and to the left of that, one at nineteen feet, and to the left of that, one at eighteen feet, six inches. Then to the right of the "Focus" target, place a target at twenty feet, six inches, and to the right of that one at twenty one feet, and to the right of that one at twenty one feet, six inches. Use the largest aperture 50mm lens in your kit, and open the aperture wide, focus on the "Focus" target, adjust your exposure, and press the shutter. Do this three or four times to take into account "pilot error". Then stop the lens down one stop, and repeat. Stop the lens down another stop, and repeat. Maybe one more stop, and repeat. Process the film and scan the negs at the highest resolution your scanner permits (or print the largest size you can in your wet darkroom).

By examining your images, you will be able to tell if your focus screen is off or if it is spot on.

Doing something similar to this is what first tipped me off that there was something wrong with my freshly serviced Canon A-1. Same with my FM-2n.

Best,
-Tim

 

[sevo]

On a wide angle lens, or a normal lens with a really big aperture like f1.2 or f1.0, the depth of focus is really small, so the focus screen and film plane have to be precisely at the same distance. With a lens stopped down to f5.6 or more, or with a long telephoto lens (both having a depth of focus that are very large), the focus screen and film plane don't have to be as precisely set.

You only look at one half of the system there. With a wide angle, the depth of field is inversely large, and with a long lens, proportionally small. I don't know off-hand whether depth of field and depth of focus exactly cancel out, but the issue seems to be much less relevant on SLRs than you make it be. Born out by the fact that I have never heard of any test that found a particular SLR not fit for wide-angle use due to poor mount or ground glass adjustment...

 

[Timmyjoe]

Depth of field and depth of focus do not cancel each other out.

And hey, if you find that any SLR camera you slap a wide angle lens on makes acceptable images for you, more power to you. That's great.

If you look at my original post, I called it an observation. I've observed that the focus screens on the consumer cameras I tested were not to spec. And I can observe how that effects image sharpness. I'm not telling anyone they need to have their camera serviced. I am trying to help those who have noticed the same thing I have and would like their images to be sharper when focusing with a consumer SLR. Nothing more.

Best,
-Tim

 

[kxl]

This is probably the most informative post that I have ever read on RFF. Thanks for your valuable insight.

 

[Timmyjoe]

Thanks Keith. Was out shooting a few rolls yesterday, comparing a Canon F-1 and Nikon F3HP with the shimmed Canon A-1 and shimmed Nikon FM2n, and the two consumer cameras held their own against their more "professional" brothers when it came to focusing. The two Canons were each in turn mated with a Canon 50mm f1.2L @ f1.4 and a Canon 85mm f1.2L @ f1.4. The two Nikons were each in turn mated with a Nikon 50mm f1.2 @ f1.4 and a Nikon 85mm 1.4 AF-D @ f1.4.

Not saying it's for everyone, just glad I shimmed these two cameras, as they are mine.

Best,
-Tim

 

[Spider67]

Is there a place where on can have his camera shimmed that way? (Bevor we all start to pester Tim 😉 ) . It's interesting that the FM2👎 and the FE 2 were considered consumer models. They were unbelievably expensive in their days. Even when sold used.

 

[hamradio]

Very, very interesting. I now wonder if my dslr is off; it has some type of aftermarket focusing screen in it, and I've noticed that achieving sharp focus with faster manual focus glass (1.4) is sometimes a challenge; what appears sharp in the VF and makes the focus confirm blink usually is slightly off, but I just chalked that up as my own mistakes, and need to click a few shots before finding perfect focus.

 

[Godfrey]

FUD

The tolerances of the film*position in the film gate of a 35mm pro SLR camera vary by as much as .005" (0.127mm). Trying to collimate focusing screen to image plane position more accurately than that is a waste of time.

The film plane is not flat anyway, it's slightly convex, because there has to be room for the film to move in the format gate; the required free space means that the film bends a little. (Yes, I know about vacuum pressure plates and clamping pressure plates ... I only know of one pro-grade 35mm DSLR that was produced with that feature, the rest are technical, specialist, and industrial cameras.)

While there's good reason to check focusing screen/image plane collimation if you have consistently unfocused images after testing focus with a good test setup, remember also that focus shift with auto-aperture SLR lens diaphragms is another source of improper focus at the taking aperture. So you have to check with several lenses to be sure that there is actually a mechanical issue with the camera body.

The long and the short of it ... Yes, if the positions of focusing screen and film plane are out of calibration, the quality of your focusing will suffer. But the far far more likely issue is that the camera or subject moved between focusing and exposure, or that the lens has focus shift issues, unless the camera is seriously beat up or very poorly made.

G

 

[tunalegs]

I'd like to see a visual demonstration of the havoc wreaked by a focus screen which is 100 thousandths of a millimeter out of whack.

I'd imagine, relative to things like camera shake, film flatness, or lens related issues, it'd be pretty hard to pick out and identify when looking at prints.

If you normally use a heavy tripod and magnifying viewfinder, it may be something really worth worrying about. For anybody else, I doubt one usually focuses the lens to within 100 thousandths of a millimeter in most practical situations anyway.

 

[David Hughes]

Hmmm, I reckon most of us have had a camera that was in focus properly at the time of shooting and then was out of focus when the prints were delivered. And some of us have used the same lens in the back up body and found everything was sharp. QED

Regards, David

 

[sevo]

Hmmm, I reckon most of us have had a camera that was in focus properly at the time of shooting and then was out of focus when the prints were delivered. And some of us have used the same lens in the back up body and found everything was sharp. QED

No, I haven't. Well, except for rangefinders (a technically different matter). The original poster was talking about misaligned ground glass.

written on the road

 

[ferider]

FUD

The tolerances of the film*position in the film gate of a 35mm pro SLR camera vary by as much as .005" (0.127mm). Trying to collimate focusing screen to image plane position more accurately than that is a waste of time.

: :

The long and the short of it ... Yes, if the positions of focusing screen and film plane are out of calibration, the quality of your focusing will suffer. But the far far more likely issue is that the camera or subject moved between focusing and exposure, or that the lens has focus shift issues, unless the camera is seriously beat up or very poorly made.

G

Not fair, Godfrey. The OP talked about adjustments of 0.101mm and 0.190mm, which is in your tolerance ballpark. And calibration of register distance for shallow Depth of Focus lenses (wide angles) is orthogonal to focus shift that is mostly an issue 50mm and longer.

Tim, this has been an excellent post, thank you. Register distance is most important for wides, focus screen position for teles, and for the fast moderate wides that I like to use (40/35 or 28), both have to play nicely together. The one consumer grade camera that has performed very accurately for me is the Nikon FM. Maybe because the focusing screen is not interchangable, or was I just lucky on ebay ? Come to think of it, the cheap fixed screen OM2000 that I once had was very good as well, out of the box.

Thanks,

Roland.

 

[Steve Bellayr]

Going back to the original post: This is why pro level cameras, even used, are better than amateur cameras new. Professionals can't afford to use cameras that go out of focus and make money. Thank you for that test information.

 

[Timmyjoe]

I'd like to see a visual demonstration of the havoc wreaked by a focus screen which is 100 thousandths of a millimeter out of whack.

I'd imagine, relative to things like camera shake, film flatness, or lens related issues, it'd be pretty hard to pick out and identify when looking at prints.

I must say I'm a bit baffled by this attitude. Having your camera set up properly, so that what is in focus on the ground glass/focus screen, is also what is in focus on the film in the camera, is really basic. From testing the few consumer cameras I have at my disposal, I found the accuracy of the above statement to be off, whereas the pro level cameras I have at my disposal, were spot on.

You need to start with your piece of mechanical equipment, have that spot on, then make sure to limit camera shake and other variables. But if your mechanical equipment is not spot on, then even if you control all other variables, you will still not get properly in focus pictures.

My test method is to shoot focus charts on Tri-X film with a 50mm f1.2 lens (set at f1.2), process, scan with a Nikon 9000, and look at the scan at 100%, and I have no problem seeing the difference.

But as I said before, if you're happy with the focusing of your consumer SLR, great, more power to you.

Best,
-Tim

 

[Timmyjoe]

Okay, here is an 100% scan/crop (no sharpening done on any of these) of the test chart shot with the Canon NEW F-1 and the 50mm f1.2L lens at f1.2:

Here is the same chart, shot with the same set up, at an 100% scan/crop shot with the Canon A-1 and the 50mm f1.2L lens at f1.2:

I repeated the shots with each camera to eliminate user error, and found the A-1 shots all to be softer than the F-1 shots.

I then proceeded to shim the A-1 focus screen, after checking it with my collimator.

Here again is an 100% scan/crop (no sharpening done on any of these) of the test chart shot with the Canon NEW F-1 and the 50mm f1.2L lens at f1.2 for the second comparison test:

And here is an 100% scan/crop of the test chart shot with the Canon A-1 and the 50mm f1.2L lens at f1.2, after I shimmed the focus screen 101 thousandths of a millimeter:

If you can see no differences in these images, as far as sharpness goes, then I would just ignore this thread, you will be happy with the images you get from your camera. I, on the other hand, noticed a difference, and I want my cameras set up so I can get the best images possible from them. I still need to deal with camera shake, subject movement, etc. But at least I know my camera is spot on, even though I may not be.

Best,
-Tim