Grain focusing by moving enlarger up and down

[Bill Burk]

What inquiring minds would really like to know is if Bill had a piece of paper under the grain focuser when he did his experiments.

I had 25 sheets under the grain focuser.

 

[Sirius Glass]

I set the enlarger height at for the print size that I want. I put a sheet of paper under the grain focus, that is what I choose to do, then I very the focus control up and down to find the limits of the focus range. I check the range several times before I select the focus location. I hope this helps.

 

[Andrew O'Neill]

@Bill Burk

Should the focus point be in the middle of the stack or is there more depth of field behind the plane of sharp focus than in front? I.e., is your expectation that the focus should be in the center of the stack backed up by the math?

@Andrew O'Neill

Focusing after you've moved the enlarger head to a position you want will change the size of the image somewhat. If you're really picky about the print size or exact degree of enlargement, you need to reposition the head (i.e., negative-to-paper distance) and refocus a time or two to get what you want. If you can stand a bit of leeway when it comes the final image size, it's not such an issue.

Best,

Doremus

Got it thanks for the explanation!

 

[Bill Burk]

I set the enlarger height at for the print size that I want. I put a sheet of paper under the grain focus, that is what I choose to do, then I very the focus control up and down to find the limits of the focus range. I check the range several times before I select the focus location. I hope this helps.

That’s a good plan. How do you know the position of the lens? By feel or by scale?

 

[Sirius Glass]

That’s a good plan. How do you know the position of the lens? By feel or by scale?

When I have to change format, the lens, paper size, I have to find the height by trial and error - setting the height and focus by eye until it get the height. Then I use the grain focuser.

 

[Bill Burk]

When I have to change format, the lens, paper size, I have to find the height by trial and error - setting the height and focus by eye until it get the height. Then I use the grain focuser.

I mean as you focus up and down then to the middle how are you gauging the middle?

I just bought this and will touch it to the lens board after initial focus. Then instead of moving the height I will try focusing up and down then to the middle.

As an interesting thought. Each 0.001 inch mark on this dial at the lens probably represents about the thickness of a sheet of paper at the easel.

 

[MattKing]

FWIW, adjusting the focus always changes the magnification/image size too.
When the adjustment is small, the corresponding change is too.
And I never thought the change in focus as you move the easel up and down would be linear. I just couldn't figure whether it would be logarithmic.
Perhaps it tracks the change in image size.

 

[Bill Burk]

I will be checking, but think the final enlarger height will not be much different than the initial focus. I guess the typical height I arrive at will be about about a sixteenth inch difference from initial focus.

 

[Craig75]

Thanks, it’s plus/minus 10 mm which would add up to it. I felt like I misconstrued the plus/minus concept a couple times.

Considering all the video production problems I am surprised I made any sense at all. Originally tried making split screen but could not get Zoom to show more than one camera on the recording. Then all the phones would keep turning off whenever it took me more than a minute to do the intro.

A grain focuser does allow you to make focus adjustments more accurately than can be seen in the finished results. In that respect it exceeds expectations. It’s an excellent tool and does its job well.

dont sweat the video problems. It was a heroic effort to make a video at all about such a subject. I found it clear and easy to follow and I am sure all the darkroom community is grateful for the huge amount of work you put in to the whole process. Greg's work and then your follow up made for very interesting discussion and reading. Photrio at its very best

 

[Mr Bill]

I just bought this and will touch it to the lens board after initial focus. Then instead of moving the height I will try focusing up and down then to the middle.

Hi, that will probably be an awkward sort of tool for what you're trying to do. The working range is small and unless you have a comfortable way to mount it...

A better tool for your purpose would be a standard "dial indicator," with a plunger-style stem, and an inch or so travel.. Along with a compatible magnetic base. Now, you might not have a good place to attach the base, but if you use a C-clamp and a small steel plate I'm sure you can find a way to get it mounted. The mag bases generally come with a post plus adjustable extension arm; these let you get the indicator where you want it.

The unit you have is generally called a "test indicator," and would more typically be used for fine adjustments to a round bar chucked in a lathe, etc. (You could mount the indicator on the tool post, etc.) Or to touch the inner surface of a tube, etc. I just don't see any graceful way to touch it to the lens board, or whatever, without some further adjustable mounts.

 

[Mr Bill]

And I never thought the change in focus as you move the easel up and down would be linear.

It should be linear, depending on how you judge "focus," I guess.

Here's a simple way to look at the situation, geometrically. Say that you have a "point" on the negative that you want to project onto the paper. Now imagine a "cone of light" originating from the entire lens aperture, and coming to a point, ideally at the paper surface. Of course we can't get to a perfect point, but this is a good approximation. So if the paper is at the wrong distance then it will intersect the cone at some point, and the diameter of the cone at that point represents the smallest detail that can be on the print. So it seems clear that the size of this "blur circle" is gonna linear with the distance error at the paper.

Here's a concrete example. Say that Bill is making 11x14" prints from a 35mm film negative, using a 50 mm fl enlarging lens set to f/4. I'm from the US, so I like to work in inch units (to convert, there are exactly 25.4 mm per inch.) OK, the diameter of the aperture is about 50mm/4 =~ 12.5 mm, or very close to 1/2 inch diameter. This is the base of the light cone.

For the height of the cone, one can just measure the approximate distance from lens to paper. This should be about 16.5 inches (I just calculated it).

So now we have a cone-of-light, from lens aperture to a point near center of the image. The cone has a base diameter of 0.5 inches, and a height of 16.5 inches. One can look at this a "rise over run" case where the diameter of the the cone increases by 0.5 inches for every 16.5 inches of "run." Or 0.5/16.5 = 0.030 inch diameter increase per inch of "run."

Now for the result of an error in paper position. If the paper is 1 inch out of position, either too close or too far, the smallest possible point of detail is about 0.030 inch (or about 3/4 mm) diameter. As a long-time car guy, I see this as about equal to an old-time spark plug gap. Or the thickness of three business cards. Or, consider that the thickness of a human hair is about 2 to 3 thousandths of an inch - so this 0.030 inch "blur circle" is about ten times larger than the width of a human hair. One might ask, can I see this amount of detail? Well, personally I don't have any trouble seeing a human hair on a piece of white paper from 5 feet away. So to me, this would be a pretty unhappy print if held in my hand.

If, instead of a 1 inch paper placement error, it was only 1/2 inch error, the smallest possible point shrinks to half of that, or 0.015 inch diameter - still considerably more than the thickness of a business card. Or 5 times the width of a human hair.

So what does it take, hypothetically, to appear "sharp?" Well, sharpness is about more than just detail, but if you wanted to be able to show hair-sized details on the print you'd want to be able to lay down "blur circle" diameters of about 3 thousandths of an inch, or smaller. With our 0.030 inch diameter change per inch of paper distance error, this means about one tenth of an inch paper error. Fwiw a typical color paper print is about the same thickness as a business card, about one hundredth of an inch. So in this specific example the paper would have to be positioned within about ten paper thicknesses of ideal.

If we start looking at much finer detail being delivered to the paper then we probably have to consider the effect of diffraction. In this case (lens, aperture, and lens to paper distance), I'd guestimate the smallest possible spot (the diameter of the Airy disc) as roughly 0.001 inch, one-thousandth of an inch. (This is a rough guess, so anyone please feel free to supply the correct number). The reason for such a large Airy disc is the amount of enlargement, leading to a long lens-to-paper distance. So the lens, although set to f/4, acts as though it is f/32 with respect to diffraction from the point of view of the paper.

Anyway, this is my view of how the effect of paper position errors come into play. I'd guess that a very finicky person, either nearsighted or with a loupe, would be able to see the difference between a perfectly focused print and one that is about five paper thicknesses off. In THIS EXAMPLE, and assuming that the negative has not moved at all. With less enlargement, say to a 5x7", same lens, aperture, they'd likely be able to see a difference with perhaps a 2 or 3 paper thickness error.

Ps, thanks to Bill Burk for doing the physical tests.

 

[Bill Burk]

Thanks Mr Bill,

The first thought as I read your post was… right - an inch out of focus would be bad - 0.030 inch finest possible detail.

Then how about 1/10 inch out of focus? That would be 0.003 inch finest possible detail.

This is the vicinity of precision I am investigating.

At the lens it’s a tenth of that - so 1/100 inch accuracy is required to keep the detail of a human hair crisp on a print.

If it’s not awkward to mount, the test dial may work. Otherwise the ruler on a refrigerator door magnet is proven - and I just need to get within a tenth of an inch. (We had to replace the door seal once on our refrigerator and I kept the magnets when I threw out the old seal).

 

[Doremus Scudder]

Maybe I'm missing something here...

I don't see using a dial indicator to position focus is going to make a perceptible improvement. We focus our view cameras by using a magnifier and by manually centering focus in the middle of the range of visible sharp focus. End point for this are determined by mentally marking where the magnified image first becomes noticeably blurry on either side of the focus area. Focus is then set by turning the focus knob to a point halfway between the two end points -- by feel or tactile memory or whatever. Smaller manual-focus cameras are focused similarly, but without the aid of as much magnification. This method seems to work well enough for making sharp negatives; I know of no camera focus system that uses a meter or indicator. The margin of error inherent in the visual/manual focusing technique seems to be more than adequate.

When enlarging, we use high-quality and high-power grain magnifiers, ending up in significantly more magnification than we use focusing cameras. Still, we determine the out-of-focus end points visually and then position the focus midway between them. Whether we do this manually or with a dial indicator doesn't really change the error inherent in the visual determination of the out-of-focus points. One could argue that, even so, using an indicator to find the midway point is more accurate, but I'm not sure that is true or that, if there were an larger margin of error placing the midpoint manually, that that would make a significant difference in print focus.

Given that the depth of field at the enlarging paper is much larger than the depth of focus on the negative, and that we (at least I) can focus well enough on my negatives to see silver grains come into focus at both the top and bottom of the emulsion and then place the focus between them, i.e., in the middle of the film's emulsion somewhere, and then stopping down to optimum aperture, which increases that depth of focus at the negative (and the depth of field at the paper), I tend to doubt that more accuracy will make any important difference in the sharpness of prints.

I make 16x20 prints from 4x5 negatives and I need an 8x loupe to see the fine detail in the print. Nobody is ever going to look that closely at a print in a normal viewing situation. In other words, I've got more sharpness than I need already; why should I strive for more?

Best,

Doremus

 

[Mr Bill]

I don't see using a dial indicator to position focus is going to make a perceptible improvement.

I think what Bill B is doing is exploring how things work. And maybe he'll find some weaknesses in his setup, or maybe it will just set his mind at ease about things he is worried about. Aside from that it's just a tool to explore some of the mechanical measurements.

I make 16x20 prints from 4x5 negatives and I need an 8x loupe to see the fine detail in the print. Nobody is ever going to look that closely at a print in a normal viewing situation. In other words, I've got more sharpness than I need already; why should I strive for more?

So you're doing fine with your setup. Great! You don't need to investigate these issues. Some might ask, why did you go so far as to use an 8X loupe to inspect your prints? Surely this has not improved your photography beyond, say, a 2X loupe. Or the extra-precise camera focusing you presumably do (?)... isn't it a waste of time? Etc., etc. I think that we all get our own idiosyncrasies about what we strive for in our craft.

Personally I'm not that finicky about focus in my own "play." Something I tend to notice in photos of people is the expressions, whether they are "real" or a "fake" for-the-camera expression. (I spent a few years of full-time high volume portrait work; something like 40 to 50 thousand subjects, of all ages, under my belt. So I sort of read things into the people images I see.) Another thing that sticks out to me is how good the skin tones are on a print (I spent years doing critical tests of color films and papers, and solving lab color problems). But neither one of these things seems that important to most photographers (they tend to think they they're finicky but i see most of it differently). It's just some of my idiosyncrasies.

Here's a couple of things that someone might do with a standard sort of dial indicator on a mag base. Say a one inch travel indicator with 0.001 inch increments on the dial; typical would be one-tenth of an inch per revolution of the dial. (Bill B's "test indicator" is a different style, with a pretty limited range - that plus and minus 0.015 inch is a max range equivalent to about the thickness of 3 sheets of photo paper.)

One, you could determine if negative bulging under lamp heat is significant. How? Set the dial indicator under the movable carriage, then use your grain focuser to accurately focus on the center of the negative. Zero the dial (they generally have rotating bezel. Write down the dial reading and the time. Leave the light on, and periodically check the focus by changing enlarger height (don't use the focus knob) to make the grain focuser sharp again. Check the dial reading and write it down, along with the time. If you want, you can make a graph of time vs the necessary carriage correction. Maybe it stabilizes after a few seconds, maybe not; maybe it varies with humidity, or even with film type. Maybe you'll decide you need a glass carrier. But this tool could help you see what's happening. Of course, you could get similar results by using various known-thickness shims under the grain focuser, but it's a lot quicker with the dial. Note: if you wanted to know how much the film actually bulged you'd have to calculate it; the basic 1/fl = 1/u + 1/v should help.

Another test is to see how much you would have to change enlarger height to correct for a non-flat negative. Again use the dial under the carriage for (relative) height measurements. Make test prints/strips with the carriage at different heights to find the best height for the center of the negative vs the edges. This is essentially the same test as Bill B did using sheets of paper for shims, but moving the carriage up/down via the dial reading is much easier and gives the ability to move in finer increments. Is this useful? I dunno, but it might help you find a mid-range position to set focus at.

A third use for the dial is to look for sloppy tolerances or defects in the enlarger. If you clamp the indicator to the movable carriage, with the dial reading the lensboard height, you can see if the lens position stays fixed, or if it drifts down. Or you can wiggle it up and down to see how sloppy the bushings are (the dial needle will jump between the two extemes).

Overall a dial indicator on a mag base is just another measuring tool, with much more precision than you could get with a direct by-eye measurement. If you have a flat surface you can even use it as a comparative height gauge.

 

[Bill Burk]

It’s hard to explain why I care about finding the best focus versus a good focus that already exceeds perceptible resolution the print can reveal.

Lore has taught people to use a sheet of paper to account for a known thickness of paper which can lead to an error of 0.015 inches within a range of 5/8 inches ( 2% of tolerance ).


But the more significant source of error, focusing technique, can lead to an error of a quarter inch of the 5/8 inch range ( 40% of tolerance ).

Good focusing technique can minimize that error by bringing the focus near the center of tolerance (checking both directions and returning to the midpoint).

So what we teach should emphasize what is important.

 

[ic-racer]

I like my D5500 enlargers, they have a focus knob on the head.

Also, when doing near 1:1 with my Durst, one must always focus by moving the head (electronic fine focus motor). Focusing by moving the lens won't find any focus if you inadvertently position the negative just a few millimeters less than 4x the focal length from the paper. Whereas focus by moving the head always finds a focal point, no matter the lens to film distance.

 

[Bill Burk]

ic-racer,

I don't emphasize enough that the work I have been doing relates to approximately 10x enlargement required for enlarging 35mm to 11x14 with wide borders.

At 1:1 the geometry is different, both negative and easel have the same tolerance. Thanks for sharing your experience that once you get inside where you're supposed to be focusing by lens movement doesn't work anymore. I remember using process cameras where both lens and copyboard had scales you could read from inside the darkroom. And they were calibrated in enlargement ratios: There was a place on each scale for each enlargement you want to achieve.

 

[Sirius Glass]

I mean as you focus up and down then to the middle how are you gauging the middle?

I just bought this and will touch it to the lens board after initial focus. Then instead of moving the height I will try focusing up and down then to the middle.

As an interesting thought. Each 0.001 inch mark on this dial at the lens probably represents about the thickness of a sheet of paper at the easel.

View attachment 281766

FWIW, adjusting the focus always changes the magnification/image size too.
When the adjustment is small, the corresponding change is too.
And I never thought the change in focus as you move the easel up and down would be linear. I just couldn't figure whether it would be logarithmic.
Perhaps it tracks the change in image size.

I will be checking, but think the final enlarger height will not be much different than the initial focus. I guess the typical height I arrive at will be about about a sixteenth inch difference from initial focus.

It should be linear, depending on how you judge "focus," I guess.

Here's a simple way to look at the situation, geometrically. Say that you have a "point" on the negative that you want to project onto the paper. Now imagine a "cone of light" originating from the entire lens aperture, and coming to a point, ideally at the paper surface. Of course we can't get to a perfect point, but this is a good approximation. So if the paper is at the wrong distance then it will intersect the cone at some point, and the diameter of the cone at that point represents the smallest detail that can be on the print. So it seems clear that the size of this "blur circle" is gonna linear with the distance error at the paper.

Here's a concrete example. Say that Bill is making 11x14" prints from a 35mm film negative, using a 50 mm fl enlarging lens set to f/4. I'm from the US, so I like to work in inch units (to convert, there are exactly 25.4 mm per inch.) OK, the diameter of the aperture is about 50mm/4 =~ 12.5 mm, or very close to 1/2 inch diameter. This is the base of the light cone.

For the height of the cone, one can just measure the approximate distance from lens to paper. This should be about 16.5 inches (I just calculated it).

So now we have a cone-of-light, from lens aperture to a point near center of the image. The cone has a base diameter of 0.5 inches, and a height of 16.5 inches. One can look at this a "rise over run" case where the diameter of the the cone increases by 0.5 inches for every 16.5 inches of "run." Or 0.5/16.5 = 0.030 inch diameter increase per inch of "run."

Now for the result of an error in paper position. If the paper is 1 inch out of position, either too close or too far, the smallest possible point of detail is about 0.030 inch (or about 3/4 mm) diameter. As a long-time car guy, I see this as about equal to an old-time spark plug gap. Or the thickness of three business cards. Or, consider that the thickness of a human hair is about 2 to 3 thousandths of an inch - so this 0.030 inch "blur circle" is about ten times larger than the width of a human hair. One might ask, can I see this amount of detail? Well, personally I don't have any trouble seeing a human hair on a piece of white paper from 5 feet away. So to me, this would be a pretty unhappy print if held in my hand.

If, instead of a 1 inch paper placement error, it was only 1/2 inch error, the smallest possible point shrinks to half of that, or 0.015 inch diameter - still considerably more than the thickness of a business card. Or 5 times the width of a human hair.

So what does it take, hypothetically, to appear "sharp?" Well, sharpness is about more than just detail, but if you wanted to be able to show hair-sized details on the print you'd want to be able to lay down "blur circle" diameters of about 3 thousandths of an inch, or smaller. With our 0.030 inch diameter change per inch of paper distance error, this means about one tenth of an inch paper error. Fwiw a typical color paper print is about the same thickness as a business card, about one hundredth of an inch. So in this specific example the paper would have to be positioned within about ten paper thicknesses of ideal.

If we start looking at much finer detail being delivered to the paper then we probably have to consider the effect of diffraction. In this case (lens, aperture, and lens to paper distance), I'd guestimate the smallest possible spot (the diameter of the Airy disc) as roughly 0.001 inch, one-thousandth of an inch. (This is a rough guess, so anyone please feel free to supply the correct number). The reason for such a large Airy disc is the amount of enlargement, leading to a long lens-to-paper distance. So the lens, although set to f/4, acts as though it is f/32 with respect to diffraction from the point of view of the paper.

Anyway, this is my view of how the effect of paper position errors come into play. I'd guess that a very finicky person, either nearsighted or with a loupe, would be able to see the difference between a perfectly focused print and one that is about five paper thicknesses off. In THIS EXAMPLE, and assuming that the negative has not moved at all. With less enlargement, say to a 5x7", same lens, aperture, they'd likely be able to see a difference with perhaps a 2 or 3 paper thickness error.

Ps, thanks to Bill Burk for doing the physical tests.

It appears to be a very small linear distance, so that is why regardless of the endless discussion I insert a sheet if photographic paper. ==> to see a difference with perhaps a 2 or 3 paper thickness error.

 

[Bill Burk]

It appears to be a very small linear distance, so that is why regardless of the endless discussion I insert a sheet if photographic paper. ==> to see a difference with perhaps a 2 or 3 paper thickness error.

I don't understand your reasoning. Have you closed your mind or did you come to a different conclusion?

The lens position is the critical dimension where 2 or 3 paper thickness error makes a difference.

I will put the caliper reading 0.001 on the lens board anchored to the negative stage.

At the easel for enlargements similar to 35mm making 11x14 prints, you are looking at 10x which means 20 or 30 sheets of paper to make a difference.

 

[Sirius Glass]

I don't understand your reasoning. Have you closed your mind or did you come to a different conclusion?

The lens position is the critical dimension where 2 or 3 paper thickness error makes a difference.

I will put the caliper reading 0.001 on the lens board anchored to the negative stage.

At the easel for enlargements similar to 35mm making 11x14 prints, you are looking at 10x which means 20 or 30 sheets of paper to make a difference.

Often I find the focus range is a short distance and I choose to include the paper to eliminate one variable, even though the variable may be too small to matter. Just being careful to be consistent in minimizing any margin of error. That is why I repeatedly go over the focus range

1. to eliminate any direction bias in the focus system
2. to estimate the middle of the focus range

IMNSHO* I suspect that direction bias can on some enlargers be a much bigger factor than anything else including paper thickness.

* IMNSHO ==> In My Not So Humble Option

 

[Donald Qualls]

I suspect that direction bias can on some enlargers be a much bigger factor than anything else including paper thickness.

Any rack gear system has significant backlash, on the close order of the size of the rack teeth. Tooth design and clearance affects this, but it cannot be eliminated with a rack gear (though using a preload spring can simulate removing the backlash, all it does is ensure the directional bias is almost always the same direction).

 

[Bill Burk]

Often I find the focus range is a short distance and I choose to include the paper to eliminate one variable, even though the variable may be too small to matter. Just being careful to be consistent in minimizing any margin of error. That is why I repeatedly go over the focus range

1. to eliminate any direction bias in the focus system
2. to estimate the middle of the focus range

IMNSHO* I suspect that direction bias can on some enlargers be a much bigger factor than anything else including paper thickness.

* IMNSHO ==> In My Not So Humble Option

Ah, yes. That's the whole of my finding... if you already work to eliminate directional bias and find the center... then adding paper is the icing on the cake of refinement

 

[ic-racer]

ic-racer,

I don't emphasize enough that the work I have been doing relates to approximately 10x enlargement required for enlarging 35mm to 11x14 with wide borders.

At 1:1 the geometry is different, both negative and easel have the same tolerance. Thanks for sharing your experience that once you get inside where you're supposed to be focusing by lens movement doesn't work anymore. I remember using process cameras where both lens and copyboard had scales you could read from inside the darkroom. And they were calibrated in enlargement ratios: There was a place on each scale for each enlargement you want to achieve.

Yes, good work; keep posting your results.

Here is something related I did in 2009. I plotted out the curvature of field of a EL-Nikkor 50mm at a very large magnification. Lensboard was off slightly too. Measurements were taken on the enlarger column by grain focusing with enlarger head.

 

[Bill Burk]

I threw out an El Nikor 50mm f/2.8 because it was soft on the edges.

I didn't notice it with the lens I use now... until now.

Now I see similar on the APO Rodagon (not N) 50mm f/2.8 wide open...

I haven't used it wide open until these tests.

p.s. I have tried several times today to hold focus at the initial focus, but am consistently missing by about 1/16 inch. That's about what I expected.

The end result is that I don't have the size consistency I wanted. The mat I would need to cut to show full frame on a finished print will vary about 1/4 inch on the short dimension.

 

[Bill Burk]

The test indicator works just as I wished. Focus, arrange so the needle is zero. Focus down and back up, check dial. Focus up and back down and check dial. Set final focus in-between.

Advantages: less fussy, I get to use my standard enlarger heights.

Initially finding the focus for the same setup I have been using (50mm lens at f/2.8) ranges in the neighborhood of 0.002 to 0.003 inch. So I move the focus something like 0.001 to 0.0015 from my initial focus. Another thousandth either way is obviously fuzzy.