Critical focus : ground glass vs infinity

Paul Ozzello · Aug 6, 2022

Dan Daniel said:
Second, it could be that the film is not sitting at the same plane as the ground glass.

Interesting question. I just loaded a tank with two test rolls. I'll have a better idea later tonight.

brbo said:
Crazy thought, is it possible that the ground glass was placed with the wrong orientation into the ground glass holder?

Nope - the frosted side is facing the lens. The ground glass adapter hooks into 4 notches in the back of the camera. Everything is machined aluminum - Impossible to reverse it.

grat said:
Now, the OP is using a very high power loupe to focus-- and I'm personally wondering if he's focusing on the smooth side of the ground glass, instead of the rough side (where the image is actually being projected). Could explain a persistent slight difference between the lens and the ground glass.

No. I'm using a Peak 20x loupe with adjustable focus, the grain on the glass is perfectly focused. But you got me thinking - maybe the grain of the glass is thicker/deeper than the film emulsion - and I'm actually focusing a few microns into the glass? But that would probably have the opposite effect and I would have to focus past infinity.

maltfalc said:
try again, but focus on the moon

Great idea . Maybe my concept of infinity is incorrect.

I wrote in a previous post that two other cameras I have exhibit the same "phenomenon". Another horseman SW612 does the same thing, and also my Hasselblad SWC/M (it has always done this since new - even after service from Hasselblad NJ). The head tech was never able to explain to me why.

To be clear, none of the cameras are way off, usually 1/4" (at most) away for the infinity mark. I mostly should wide lenses (72mm/90mm schneider, 45/55/65 Rodenstock) and I almost always shoot 2 stops from fully stopped down. I have never really noticed a sharpness problem even with my drum scans. But it's something that baffles me - and I always wonder if I'm focusing optimally.

The Horseman SW617 is similar to the Linhof 617 (same schneider lenses) but uses detachable film backs. The backs were made by Mamiya and keep the film very tight and flat.

I thought maybe the ground glass adapter was off but I have 2 of them for the SW617 (both glass without fresnel) screwed tightly inside an aluminum frame. When attached there is absolutely no play either with the ground glass or film back. My SW612P does the same - and I have two different ground glasses - one factory glass - and another custom one with a fresnel made by SKGrimes; both focus identically and both are equally off ~1/4" from infinity.

I'm severely myopic (optically ;-) and wear thick glasses - could that possibly throw off my perception of focus??

The reason I bring this up is that I want to shoot more often with the external viewfinders and not always deal with the hassle of using the ground glass. I mostly shoot distant mountain ranges which should be at infinity. And I'm stumped at where I should set the focus

reddesert · Aug 6, 2022

A few thoughts:

- obviously what matters is the on film test. Often it is easier to measure the point of sharpest focus if it's closer than infinity, for example shoot a fence running obliquely away from you and focus at ~20-30 feet away. That way you can (ideally) see degraded focus on either side of the focus point, which helps to locate the best focus. Testing at infinity one only sees one side so it's hard to locate the "best."

- The other thing that matters is how far out the nominal error is. You mention about 1/4" which I assume means 1/4" of rotation of the focus ring. But this could be different for different lens helicals, etc. What matters is how much extension (lens-film distance) past nominal infinity. If you say for example "when the GG is focused at infinity the lens focus scale reads 20 meters for a 75mm lens" that would allow us to compute the extra extension from the lens equation. (For those example numbers, it would be 0.3 mm of extension. Ideally the GG and lens position would be better calibrated than 0.3mm, but it's within the depth of focus at f/5.6, and you would likely not see it in practice if you're regularly stopping down to f/16.)

- Related to that, you can also check where on the focus ring your best position is, relative to the DOF scale, if there is a DOF scale. If the offset is much smaller than the DOF at the taking apertures you usually use, then it may not be worth worrying about. If it is larger than the DOF, clearly it needs to be checked.

- If you are worried about focusing your eye on an incorrect surface relative to the ground glass, you can use a fiducial index mark on the ground glass. Focus your eye on the mark and the image at the same time and you'll be guaranteed to be focusing on the GG surface. If the GG doesn't have crosshairs on it already, you can draw a light mark on the GG with pencil (For a true glass groundglass. Pencil will erase easily off a glass screen; not sure it will be as easy if the screen is molded plastic.)

John Wiegerink · Aug 7, 2022

Paul Ozzello said:
Interesting question. I just loaded a tank with two test rolls. I'll have a better idea later tonight.

Nope - the frosted side is facing the lens. The ground glass adapter hooks into 4 notches in the back of the camera. Everything is machined aluminum - Impossible to reverse it.

No. I'm using a Peak 20x loupe with adjustable focus, the grain on the glass is perfectly focused. But you got me thinking - maybe the grain of the glass is thicker/deeper than the film emulsion - and I'm actually focusing a few microns into the glass? But that would probably have the opposite effect and I would have to focus past infinity.

Great idea . Maybe my concept of infinity is incorrect.

I wrote in a previous post that two other cameras I have exhibit the same "phenomenon". Another horseman SW612 does the same thing, and also my Hasselblad SWC/M (it has always done this since new - even after service from Hasselblad NJ). The head tech was never able to explain to me why.

To be clear, none of the cameras are way off, usually 1/4" (at most) away for the infinity mark. I mostly should wide lenses (72mm/90mm schneider, 45/55/65 Rodenstock) and I almost always shoot 2 stops from fully stopped down. I have never really noticed a sharpness problem even with my drum scans. But it's something that baffles me - and I always wonder if I'm focusing optimally.

The Horseman SW617 is similar to the Linhof 617 (same schneider lenses) but uses detachable film backs. The backs were made by Mamiya and keep the film very tight and flat.

I thought maybe the ground glass adapter was off but I have 2 of them for the SW617 (both glass without fresnel) screwed tightly inside an aluminum frame. When attached there is absolutely no play either with the ground glass or film back. My SW612P does the same - and I have two different ground glasses - one factory glass - and another custom one with a fresnel made by SKGrimes; both focus identically and both are equally off ~1/4" from infinity.

I'm severely myopic (optically ;-) and wear thick glasses - could that possibly throw off my perception of focus??

The reason I bring this up is that I want to shoot more often with the external viewfinders and not always deal with the hassle of using the ground glass. I mostly shoot distant mountain ranges which should be at infinity. And I'm stumped at where I should set the focus

Try focusing without your glasses, you might just be surprised.

Paul Ozzello · Aug 7, 2022

reddesert said:
A few thoughts:

- obviously what matters is the on film test. Often it is easier to measure the point of sharpest focus if it's closer than infinity, for example shoot a fence running obliquely away from you and focus at ~20-30 feet away. That way you can (ideally) see degraded focus on either side of the focus point, which helps to locate the best focus. Testing at infinity one only sees one side so it's hard to locate the "best."

- The other thing that matters is how far out the nominal error is. You mention about 1/4" which I assume means 1/4" of rotation of the focus ring. But this could be different for different lens helicals, etc. What matters is how much extension (lens-film distance) past nominal infinity. If you say for example "when the GG is focused at infinity the lens focus scale reads 20 meters for a 75mm lens" that would allow us to compute the extra extension from the lens equation. (For those example numbers, it would be 0.3 mm of extension. Ideally the GG and lens position would be better calibrated than 0.3mm, but it's within the depth of focus at f/5.6, and you would likely not see it in practice if you're regularly stopping down to f/16.)

- Related to that, you can also check where on the focus ring your best position is, relative to the DOF scale, if there is a DOF scale. If the offset is much smaller than the DOF at the taking apertures you usually use, then it may not be worth worrying about. If it is larger than the DOF, clearly it needs to be checked.

- If you are worried about focusing your eye on an incorrect surface relative to the ground glass, you can use a fiducial index mark on the ground glass. Focus your eye on the mark and the image at the same time and you'll be guaranteed to be focusing on the GG surface. If the GG doesn't have crosshairs on it already, you can draw a light mark on the GG with pencil (For a true glass groundglass. Pencil will erase easily off a glass screen; not sure it will be as easy if the screen is molded plastic.)

Lots of interesting points.

The helicals are made by Horseman and the lenses are mounted to fixed lens cones which screw into the camera body. The lens I tested is the Schneider 90mm SA XL and the focus falls between 10m and infinity (there are not intermediate values). There are exactly 64mm of rotational distance between infinity and 3m, and my infinity focus is off by roughly 6mm (but within the f5.6 depth of field scale).

I could try a closer focus test next week depending on what my test roll shows (it's hanging up to dry). And I have a measuring wheel I could use to test a series of closer distances - but how accurate is the distance scale on a lens?

Based on the ground glass - it seems the lens is focusing past infinity, so if the film plane is identical to the gg the test should show a similar result to what you described showing both sides. I'll follow up in the morning after I look at the film.

You're right about it probably being negligeable especially stopped down. I routinely make 72" wide prints and I'm a little OCD about my technique .

Here are a few images of the camera lens and ground glass :

Dan Daniel · Aug 7, 2022

Paul Ozzello said:
I'm severely myopic (optically ;-) and wear thick glasses - could that possibly throw off my perception of focus??

Not sure if this will resolve your issue, but yes, I have found that the glasses I wear when checking focus using a loupe make a large difference. I have two different strength of reading glass prescription, one for reading and one for working on cameras and such. If I have the wrong pair on, I never get a clear image. I also have found that one loupe that I use with the reading glasses has needed to be shimmed up to work.

So this avenue is worth exploring. A a general troubleshooting principle, if you keep getting the same problem, look for the one common element from example to example, and your eyes may be it.

Paul Ozzello · Aug 7, 2022

So I took a look at my test roll this morning and have come to the conclusion that my lenses are OK and the ground glass must not be properly calibrated.

To recap I shot 4 test images of a distant bridge to try to get the best focus at infinity.

1) f5.6 focusing on the ground glass
2) f5.6 setting the lens on the infinity stop (and slightly unsharp on the ground glass)
3) f5.6 focusing at 10 meters
4) f32 focused on ground glass.

Using a 10x loupe it's really hard to notice any differences at all. But at 20x it becomes clearly evident.

Frame 2 set at infinity was by far the sharpest, and the bridge in the distance was extremely detailed. At f5.6 Depth of field was obviously very poor.

Frame 4 focused on the ground glass and shot at f32 was also very sharp and had great depth of field. But the bridge was still slightly less sharp than in frame 2 at infinity (so even stopped down - proper focus is still critical). Depth of field really helped keep the bridge sharp and it was significantly sharper than frame 1 despite both being improperly focused on the ground glass.

I don't know if diffraction came into play - I wish I had taken a test shot at infinity at f32 to compare. I chose f32 since it's two stops short of the maximum f64.

In Frame 1 you can clearly see the focus plane is much closer than what was sharp on the ground glass - the focus plane was roughly 50 meters away - not the bridge at infinity I had focused on using the ground glass.

Luckily my ground glass is removable from the frame so with some trial and error I should be able to shim it - it should only require 1-2 thicknesses of film. But at least for now I think I can safely assume that the lenses are perfectly calibrated for infinity and use my external viewfinders.

I was also really surprised by how much 1/4 inch of helical movement can effect the focus plane so dramatically.

Alex Varas · Aug 7, 2022

Congrats, Paul!! You know the problem so you are half way to the solution already!

For shims I use 0.2mm thick envelop paper (light brown used in post) it has some wax (I think) and it will last.
I would start putting just one on each side of the ground glass.
Before anything I would measure the distance from the border of the ground glass frame till the ground glass itself, using the lower part of the caliper, I have been measuring things this way and it might sound not accurate but without any other tools you will do fine (maybe you need to measure 5 times each spot for having a real number). After having numbers already I would put shims and I would check the real difference by doing same operation as above.

After new tests shots you will see if you need to add another 0.2mm shims or remove or whatever until you get same photo with infinity stop and ground glass focusing image.

You are gonna need a caliper to measure correctly, I use a digital one and it serves well.

Paul Ozzello · Aug 7, 2022

Alex Varas said:
Congrats, Paul!! You know the problem so you are half way to the solution already!

For shims I use 0.2mm thick envelop paper (light brown used in post) it has some wax (I think) and it will last.
I would start putting just one on each side of the ground glass.
Before anything I would measure the distance from the border of the ground glass frame till the ground glass itself, using the lower part of the caliper, I have been measuring things this way and it might sound not accurate but without any other tools you will do fine (maybe you need to measure 5 times each spot for having a real number). After having numbers already I would put shims and I would check the real difference by doing same operation as above.

After new tests shots you will see if you need to add another 0.2mm shims or remove or whatever until you get same photo with infinity stop and ground glass focusing image.

You are gonna need a caliper to measure correctly, I use a digital one and it serves well.

Good idea - I have some of those calipers and a micrometer, but I don't have any of that paper - do you think a piece of film would be too thick?

I don't think I need to shoot anymore test rolls - just need to get the ground glass properly shimmed with the lens on infinity.

@reddesert had an interesting comment about calculating the focus plane movement based on the rotational distance of the helical . It would be interesting to compare the theoretical value with what I measured. What is the equation for figuring this out?

grat · Aug 7, 2022

Does your camera have a fresnel lens? If so, would be worth verifying it's in the right location.

Paul Ozzello · Aug 7, 2022

grat said:
Does your camera have a fresnel lens? If so, would be worth verifying it's in the right location.

No fresnel. Just ground glass.

@Alex Varas I measured both the gg and film back with digital vernier calipers. Average distance of 5.09 mm for the film back vs 4.82mm for the ground glass. So the ground glass needs to be moved back about 0.3 mm. This also corresponds to about the distance the lens moves when turning the helical 1/4". Film thickness is 0.12mm so I'll start with two pieces on each side

reddesert · Aug 7, 2022

Paul Ozzello said:
@reddesert had an interesting comment about calculating the focus plane movement based on the rotational distance of the helical . It would be interesting to compare the theoretical value with what I measured. What is the equation for figuring this out?

The thin lens equation is:
1/f = 1/d_i + 1/d_o
where f = focal length, d_i = image distance, d_o = object distance. image distance is distance from lens to film, object distance is from lens to object. Focusing scales are usually marked with distance from film to object, which is a bit larger than d_o, but if you're focused at more than 10-20 focal lengths away this difference won't matter.

So for the max travel of your focusing ring, on the 90mm lens, the focus scale is at 3 meters, and setting d_o = 3000 mm, we get d_i = 92.8 mm. So fully rotating the focusing ring by 64mm gives 2.8mm of focus travel outward. Your focus offset was a rotation of 6mm, which is about one-tenth of the full travel, so that's an extension of approximately 0.28mm, meaning the ground glass is too far forward by ~ 0.28mm.

Another way of working this out is that the focus position is at roughly 25 meters on the scale in your picture above in post #30. That gives an image distance of 1/90 = 1/d_i + 1/25000, so d_i ~= 90.33 mm, or an extension of 0.33 mm. Obviously these two numbers are both approximate and slightly different; I certainly wouldn't trust them to the hundredth of a millimeter, but they agree at ~ 0.3 mm of extension, again in the sense that the ground glass is too close to the lens.

Roll film is usually about 0.1 mm thick so you would need a couple of layers to get the right shim; heavy paper might be closer.

These numbers agree very well with your measurement with calipers, the one caveat I have there is that you have to measure depth to the film rails not to the pressure plate, since the film + backing paper thickness means the emulsion is about 0.2mm in front of the pressure plate. Or put a dummy roll of film into the back and measure depth to that.

Dan Daniel · Aug 7, 2022

Scotch tape is a nice thin shim material. About 0.04mm per layer.

A good way to measure adhesive material thickness like tape is to measure the thickness of a flat like ground glass or a steel ruler. Then apply the tape, measure again, and subtract. This is actually good for other shim materials, adhesive or not.

Paul Ozzello · Aug 9, 2022

reddesert said:
These numbers agree very well with your measurement with calipers, the one caveat I have there is that you have to measure depth to the film rails not to the pressure plate, since the film + backing paper thickness means the emulsion is about 0.2mm in front of the pressure plate. Or put a dummy roll of film into the back and measure depth to that.

I'm always amazed at the wealth of knowledge members on this forum possess - thanks for sharing reddesert - I also learned something new - fiducial

I ended up taping two pieces of film together with double stick tape and measured 0.29mm with a micrometer and cut 2 strips to fit between the ground glass and the frame. Infinity now appears in perfect focus on the ground glass - with all my lenses including the 180mm.

Thanks to everyone who chimed in to help!

Alex Varas · Aug 9, 2022

Glad to help, is that 180mm I had in hands?

Paul Ozzello · Aug 9, 2022

Alex Varas said:
Glad to help, is that 180mm I had in hands?

I never told you I got it?! And just in time!

Alex Varas · Aug 9, 2022

Now I’m out of that market for good.
Enjoy the lens and share photos, Paul.

Paul Ozzello · Aug 9, 2022

Alex Varas said:
Now I’m out of that market for good.
Enjoy the lens and share photos, Paul.

Because of the war? I still have lots of film to scan and will post when I can Thanks to you I shot these in the spring with the 180

Alex Varas · Aug 9, 2022

Not really, it was planned before, no good place for raising a child in our opinion.
Awesome photos, Paul, I’m happy I could check lens on site for you.

Paul Ozzello · Aug 9, 2022

Alex Varas said:
Not really, it was planned before, no good place for raising a child in our opinion.
Awesome photos, Paul, I’m happy I could check lens on site for you.

Thanks Alex
Did you leave the country?

-- Alex is one of the good guys on the forum - he checked a lens for me halfway across the world that I bought site unseen ))

Alex Varas · Aug 9, 2022

Yes, we did. I miss the film community over there.
Thank you, Paul, for your kind words, it was my pleasure.

Critical focus : ground glass vs infinity

Privacy & Transparency

Privacy & Transparency