Shooting through glass, the right setup for it, and other questions

[eggplant]

I am looking to scan opaque diagrams/documents, SD card to 6x8 film size, at 0.8:1 to 2:1 magnification. As you can imagine this requires me to flatten them out, and so I have been using some float glass taken out of an old scanner, that's abit larger than A4.

I'm now looking to replace this as I discovered it's warped in the middle (not flattening document properly), and I don't think it's weight of 727g spread out over a much larger area than my diagrams is truly enough to flatten them.

It was my idea to use a weighty bit of glass (700-930g) of a smaller size (120-140mm diameter) to shoot through. Unfortunately that means it has to be quite thick- 20mm:



You might suggest I should weigh a much thinner bit of glass (100mm~ UV filter?) down with weights but I have found that very difficult to do evenly. If you have any advice here it would really be appreciated.

My instinct was to say it's much better to let the glass do the flattening with no extra pieces. However, taking a look around here for discussions on the impact of cover/sensor glass in places lenses don't expect, I am now more skeptical - the worst of MFT was 4mm, and this will be 15-20mm..

It's worth mentioning that I intend to stop down to f5.6 (or maybe f4.5 if the resolution is great enough) for depth of field purposes- it needs to be a bit forgiving, even at the expense of 'resolution'.

However, even then- what will be the effect of glass this thick? What alternatives should I consider? My camera is an A6000.



My current plan for good performance optics on a limited budget is through macro coupling, allowing me to work at higher effective apertures. I am not trying to 'unlock' hidden 'resolution' in these diagrams- rather, have enough sharpness that no MB of file size is wasted.

My current plan was to couple some of my high quality enlarger lenses such as Omicron-EL 50mm f2.8, to a high quality, wide aperture 'tube' lens. I have opted for an ISCO Ultra-Star HD 80mm f1.8, 35mm film projection lens (meant for Super35), which is quite high quality on a budget with no need for aperture control. I am essentially following on from the results shown here https://www.closeuphotography.com/1x-low-cost-lens-test but this was using two identical stacked lenses, whereas this is a fair bit different. The Ultra-Star HD is a Double-Gauss, but quite advanced design.

I would like to keep using the enlarger lens reversed. I do believe the Ultra-Star is more than high quality enough for a digital sensor, I am just curious how much weight I should put on modern 'telecentric' mirrorless lenses for use as tube lenses.



I do own a JML 64mm f0.85, but the working distance even coupled would be very unforgiving and I'm not sure I could light things properly. And it would be very heavy.

[enricosavazzi]

As you already suspect, 20 mm of glass is going to introduce strong spherical aberrations. It will also introduce axial chromatic aberration, although I am not sure whether the amount will be enough to notice, and lateral chromatic aberration (probably to a much larger extent). The problems can be lessened by using a lens of long focal length, so that oblique rays (which are the ones causing the most trouble) will be closer to perpendicular to the glass plate, but not completely eliminated. Also, this will force you to use a relatively slow lens, so diffraction may become a factor.

Is keeping the subject completely flat really important? At 2x and nominal f/8 the DOF is roughly 0.3-0.4 mm, which may be enough to hide the non-flatness of the subject across a 12 x 18 mm field of view (if you are shooting on full-frame at 2x).

Is it really necessary to use an A4 glass plate? At 0.8x the field of view on full frame is only about 30 x 45 mm, so a glass plate slightly larger than this is enough.

Some camera stands for reproducing documents use an inflated pillow or a layer of synthetic foam at the back of the document to press it against a glass plate. This would allow you to use a thinner glass, anchored to a frame along its edges so that it does not lift up.

[eggplant]

As you already suspect, 20 mm of glass is going to introduce strong spherical aberrations. It will also introduce axial chromatic aberration, although I am not sure whether the amount will be enough to notice, and lateral chromatic aberration (probably to a much larger extent). The problems can be lessened by using a lens of long focal length, so that oblique rays (which are the ones causing the most trouble) will be closer to perpendicular to the glass plate, but not completely eliminated. Also, this will force you to use a relatively slow lens, so diffraction may become a factor.

Well, atleast that can be avoided now

Is keeping the subject completely flat really important? At 2x and nominal f/8 the DOF is roughly 0.3-0.4 mm, which may be enough to hide the non-flatness of the subject across a 12 x 18 mm field of view (if you are shooting on full-frame at 2x).

I'm shooting on APS-C.
It is really important to keep it flat, as in my experience the increased depth of field isn't enough to bring on side into sharp relief/focus.

Is it really necessary to use an A4 glass plate? At 0.8x the field of view on full frame is only about 30 x 45 mm, so a glass plate slightly larger than this is enough.

It isn't, but it was my idea that it would flatten it out properly. However as you highlight even at 0.8:1 mag on full frame, you would have a small field of view, and at higher magnifications an even smaller portion of the diagram.

Some camera stands for reproducing documents use an inflated pillow or a layer of synthetic foam at the back of the document to press it against a glass plate. This would allow you to use a thinner glass, anchored to a frame along its edges so that it does not lift up.

This is an interesting idea, but I have never seen it before so will need to research to see what I'm meant to be recreating.

Im guessing the weight of the glass pushes the document against the foam to flatten out the difficult parts?

[rjlittlefield]

Some camera stands for reproducing documents use an inflated pillow or a layer of synthetic foam at the back of the document to press it against a glass plate. This would allow you to use a thinner glass, anchored to a frame along its edges so that it does not lift up.

This is an interesting idea, but I have never seen it before so will need to research to see what I'm meant to be recreating.

Im guessing the weight of the glass pushes the document against the foam to flatten out the difficult parts?

It's the rigidity, not the weight. The frame around the glass provides the front force that counters the pillow/foam on the back. This is just an upside-down version of using a piece of foam to press a document onto the glass of a normal flatbed scanner.

Regarding aberrations with thick glass, there's an old thread about shooting through thick medium that is relevant here. Bottom line is that for your application the big problem would be aberrations away from center if the ray cones are not oriented perpendicular to the surface. See https://www.photomacrography.net/forum/viewtopic.php?p=102475#p102475 and the rest of that thread. So, you would want at least a long lens, or better a combo that tends telecentric on the subject side. Telecentricity of the tube lens, on the sensor side, is no matter.

--Rik

[eggplant]

Some camera stands for reproducing documents use an inflated pillow or a layer of synthetic foam at the back of the document to press it against a glass plate. This would allow you to use a thinner glass, anchored to a frame along its edges so that it does not lift up.

This is an interesting idea, but I have never seen it before so will need to research to see what I'm meant to be recreating.

Im guessing the weight of the glass pushes the document against the foam to flatten out the difficult parts?

It's the rigidity, not the weight. The frame around the glass provides the front force that counters the pillow/foam on the back. This is just an upside-down version of using a piece of foam to press a document onto the glass of a normal flatbed scanner.

Ah, yes, I did think it was going to be like a scanner, just didn't mention it.

Regarding aberrations with thick glass, there's an old thread about shooting through thick medium that is relevant here. Bottom line is that for your application the big problem would be aberrations away from center if the ray cones are not oriented perpendicular to the surface. See https://www.photomacrography.net/forum/viewtopic.php?p=102475#p102475 and the rest of that thread. So, you would want at least a long lens, or better a combo that tends telecentric on the subject side. Telecentricity of the tube lens, on the sensor side, is no matter.

--Rik

Very useful thread, thanks, particularly using the clear filter as the window.

That fully answers my question, and I think in my position it would be better to try ways which minimise glass thickness.

Now I think the inevitable recommendation is to just use a thin bit of float glass like anyone else in a copy stand/scanner etc, but that's still going to be too big.

I am tempted to make use of two 'flat' diving weights such as these, that are 500g each:


They are much closer in size to my diagram, and being flat shouldn't disrupt the lighting setup.

Im considering using these either to weigh down a large UV filter, or preferably to 'tense' a page flat with no glass used, by putting them on top of the diagram and then pushing them in opposite direction to each other.

Either way, something appropriate to the size of my diagram is required.

[colohank]

This is perhaps more bother than it would be worth for a few photos, but if you have a continuing need to do flat copy work, have you considered making a vacuum table?

Drill a pattern of small holes 1/16" to 1/8" diameter on a 1/2-inch grid in some flat, rigid material like 3/4"-thick medium density fiberboard (MDF). Consider using a double thickness of MDF for a very large table. 3/4" MDF is often available in various sizes up to 4 X 8-foot sheets at any home improvement store. Construct a relatively air-tight chamber beneath the MDF using 1/4-inch Masonite or whatever and drill a hole in one side large enough for a decent friction-fit of a vacuum cleaner hose. Turn on the vacuum cleaner, and the vacuum will suck the document flat against the MDF. If the document is smaller than the pattern of holes, simple cover exposed parts of the grid with scrap paper to maintain the vacuum. Of course, the camera mount would have to be isolated from the vacuum cleaner and vacuum table to eliminate any transfer of vibration.

No concerns about aberrations or the need to tame reflections.

I hope the attached hastily drawn and admittedly crude drawing helps you visualize the project.
Dimensions aren't critical.

[lothman]

This is perhaps more bother than it would be worth for a few photos, but if you have a continuing need to do flat copy work, have you considered making a vacuum table?

That also would have been my recommendation. For the lens I think having AF is benefit and the new Sigma 70/2,8 lens is very affordable and optically excellent.

I think there are different slightly sticky materials available which leaves no residue on your objects and which could be coated to a flat board. This would avoid vibration of a vacuum pump/vacuum cleaner.
https://www.aslanfolien.de/en/products/ ... 11224.html
I think in english language it is called "mounting film".

Edit:
just found a product from 3M called 9415:
https://www.3m.co.uk/3M/en_GB/p/d/b40070350/