How to use a step wedge with roll film?

[grainyvision]

I want to get a bit more consistent in testing some custom film developer formulations and a step wedge is the most obvious way to do this. However, I typically like testing as many films in a developer as possible per run, and prefer to not use expensive sheet film. So, using a roll film like 120 or 35mm would be greatly preferred. However, exposing these aren't so simple in my tests under an enlarger. Film curl is a big annoyance, as well as the unpredictable ways in which enlarger exposure works. My enlarger timer does have 0.1s resolution, but I don't trust that to be too consistent since it doesn't use a regulated power supply or anything like that.

I'm considering just shooting a step wedge that is backlit and using the same rig as I have for DSLR scanning, but with a 35mm SLR and macro lens (if needed for 4x5). I understand lenses will have some effect on this, specifically some lenses are lower or higher in contrast due to coatings. My bigger concern is edge fall off which would change the step wedge exposure results, but that can be solved by using a good lens and leaving a bit of room in the frame around the step wedge. Anyway, does anyone have any better methods of exposing a step wedge onto roll film? Also, how would I meter in a situation like this?

 

[RalphLambrecht]

I want to get a bit more consistent in testing some custom film developer formulations and a step wedge is the most obvious way to do this. However, I typically like testing as many films in a developer as possible per run, and prefer to not use expensive sheet film. So, using a roll film like 120 or 35mm would be greatly preferred. However, exposing these aren't so simple in my tests under an enlarger. Film curl is a big annoyance, as well as the unpredictable ways in which enlarger exposure works. My enlarger timer does have 0.1s resolution, but I don't trust that to be too consistent since it doesn't use a regulated power supply or anything like that.

I'm considering just shooting a step wedge that is backlit and using the same rig as I have for DSLR scanning, but with a 35mm SLR and macro lens (if needed for 4x5). I understand lenses will have some effect on this, specifically, some lenses are lower or higher in contrast due to coatings. My bigger concern is edge fall off which would change the step wedge exposure results, but that can be solved by using a good lens and leaving a bit of room in the frame around the step wedge. Anyway, does anyone have any better methods of exposing a step wedge onto roll film? Also, how would I meter in a situation like this?

macro shot and average metering for the whole step wedge.

 

[Mr Bill]

Hi, I think it is fundamentally a bad idea to rely on photographing a step wedge - the "shadow" parts of your curve are gonna be contaminated, by flare, to various extents, depending on how bright the brightest parts get, as well as your lens. I have personally tried variations of this, way back. Just as a demonstration, here's something you might try: set up and photograph a full-step wedge; then block off the brightest (most clear on the backlit wedge) and photograph it again. You may be surprised how much effect there is on the result, presumably the effect of lens flare. (If it's not objectionable to you, great!)

If you wanna study the film/developer characteristics, only, with no lens effect, I think you wanna stick with contact prints onto the film.

At the outfit where I used to work I used to oversee some in-depth film testing, so we had a couple of sensitometers; this is by far the best way to do it, in my opinion, IF one has an appropriate sensitometer. Without a formal sensitometer I would suggest rigging up a system. If I were gonna do it I'd probably base it on an electronic flash in a darkroom, maybe a hot-shoe style flash, firing at full power to keep the exposure as long as possible. If you fire it from a distance, to minimize side-to-side fall-off, and use some sort of diffuser over the front to make sure that no optical pattern is projected, I think this ought to work pretty well. Now, I have not actually done this, so there may be something I'm overlooking (does it have enough power, etc.).

I wouldn't initially worry about a "speed rating," using the contact print via flash mainly to make a relative characteristic curve (assuming that your step wedge has even steps, density-wise, you should get good results). As a later step you could actually shoot something through an actual camera, then correlate this actual metered exposure back to your exposed wedges (I don't knock exactly how, offhand, but it ought to be easy enough to figure out a way.)

For the mechanism of handling the film, you might use 35 mm film, winding from a fresh cassette into an empty one (designed for reloading, etc.). So you would only need to have about 6 or 8 inches exposed at once - whatever you need to hold your step wedge. If you've never done this before, it can be difficult to find a reference point on the wedge, so you might want to put several reference points on the edge of the step wedge - a little bit of black tape cut to a triangle point, or whatever, works.

Best of luck with the testing.

Ps, as one last check on your system, you might try exposing the wedge in both directions, just in case your light is uneven. And at the processing end, also try a set of opposing wedges - this will detect something likely a flow pattern problem.

 

[grainyvision]

The flare issue is definitely my biggest concern with trying to just photograph the step wedge. I know from my DSLR scanning that my setup is definitely far from flare free with modern coated digital lenses. I can't imagine an older film lens being better. However, it means the results would be sub-par but not necessarily useless. I'm not intending to do density measurements or anything like that with these tests. I primarily only intend to use these as a method of comparing two identical films shot using the same setup, and developed in different developers. As long as flare etc isn't bad enough to completely obscure shadows, that's still usable to some extent.

I've also considered if there is some method of mounting a step wedge inside a camera (smallest size I have is a 120 step wedge) while being able to use the lens aperture and shutter speed of the camera for exposure. This seems pretty risk to test out though as it could scratch or ruin my step wedge or even the camera itself, if the mounting method doesn't work well enough. Alternatively, I have a 3D printer and I could build some kind of device to hold the film flat etc (I've printed with enough precision to make a good negative holder for DSLR scanning) and have a slot or something that the step wedge goes into, then expose using the enlarger. I don't necessarily need precision in this, but rather consistency though is my worry. A non-regulated enlarger might be a bit risky for sub-second exposures.

edit: beyond a step wedge, if the flare and glare is not too bad for the photographing method, I could even setup a series of carefully made/chosen slide film exposures to help judge things more objectively, though I think setting up a proper 100% controlled environment with various texture and tonal examples would be more effective than that. Hell, it might be a lot easier than trying to work with a step wedge even, though the step wedge having numbers is of course very nice

 

[Kino]

Without a formal sensitometer I would suggest rigging up a system. If I were gonna do it I'd probably base it on an electronic flash in a darkroom, maybe a hot-shoe style flash, firing at full power to keep the exposure as long as possible. If you fire it from a distance, to minimize side-to-side fall-off, and use some sort of diffuser over the front to make sure that no optical pattern is projected, I think this ought to work pretty well. Now, I have not actually done this, so there may be something I'm overlooking (does it have enough power, etc.).

.

This is the basis for the EG&G PH-10 Mark VI Sensitometer. Basically a strobe tube in a box pointing up, with a platen that holds the step wedge behind a glass plate. The hinged cover has foam on the underside to insure flatness of the film being tested and when you lower the cover and press lightly down on the handle, it closes a momentary contact switch and exposes the wedge onto the film. The EG&G unit has a push-button selector for intensity, probably a capacitor network of varying power, but this can be simulated by building a slot in the box and making an ND frame to slide over the flash head or by varying the output power of the strobe itself.

Here is an example of one on evilbay: https://www.ebay.com/itm/EG-G-Sensi...239036?hash=item23b9556dfc:g:YcoAAOSwWiJcnloY

 

[ic-racer]

The best way to make control strips is with a sensitometer. Last time I checked on ebay it indicated:

91 results for sensitometer

 

[ic-racer]

Actually on the EG&G the different capacitors were primarily to alter the exposure time. As you suggested, a series of ND filters are used over the flash tube to alter exposure.

 

[Mr Bill]

Actually on the EG&G the different capacitors were primarily to alter the exposure time.

Yep, at the time there were apparently still issues with film still subject to reciprocity failure due to electronic flash duration, so three (or four) exposure durations allowed testing for this. The machines were originally supply with attenuators to equalize the exposures; they looked similar to ic-racer's example.

 

[grainyvision]

I ended up buying a cheap used X-Rite 334 Dual Color Sensitometer off ebay for $15. I figure best case it works well for this, worst case I have yet another tool for other purposes. The way that'd be nice to use this kind of device I think, is to work in a darkroom and pull out enough film from a 35mm canister (or 120 if needed) to completely cover the device area, expose using the device, then cut the film and do it again so that I can basically just make a dark bag full of presensitized strips of a certain film type that I can quickly load into a reel and tank and test out a new developer formulations. Might also come in handy for paper occasionally. Unsure how to make too much sense of these strips directly other than by comparison to other developers, at least without a densitometer (been looking for one, but no good deals yet on one not requiring a complex computer interface). Either way, I figure it'd be a good reference to have, especially for developer aging if nothing else. This method looks more simple than trying to figure out some 3d printing rig at least.

 

[ic-racer]

I think you are on the right track buying a commercial sensitometer. The hardest part of home-made sensitometer is getting the light even and constant. That part of even a 'broken' sensitometer should still be workable and the light could be replaced or modified.

I tested flash, incandescent and LED (green and blue) against each other and they all worked essentially the same to make control strips to monitor one's process.
https://www.photrio.com/forum/threads/the-great-sensitometer-shootout.95837/

 

[Mr Bill]

Unsure how to make too much sense of these strips directly other than by comparison to other developers, at least without a densitometer...

Well, one useful way is to make contact prints of your self-processed film wedges onto photo paper (perhaps a soft grade). So this roughly mimics the result of actually printing: you could even expose through the actual enlarger for a better result (it's the Callier coefficient thing, or whatever they call it now). To compare results you can butt the printed wedges side-by-side to see how individual steps vary. You might find, for example, that one developer gives gentler contrast in the shadow areas, but overtakes another film in the midtones, or whatever. I don't know that this is useful on its own, but... if you find that you really like a certain film-developer combination for something, portraits for example, you can study how that response differs from those you don't like. Even if not immediately useful you'll at least developed a sort of understanding of what's happening, contrast-wise.

Fwiw the film exposed in your sensitometer essentially covers all film exposure variations; this happens automatically as a result of the original step wedge being in equal density steps. So if you get to wondering about camera exposure, for example (what if I increased my camera exposure by an f-stop on portraits with this particular film?), no need to repeat the film exposure. Just take the one film test you have, and reprint it. If you print it darker this is equivalent to having increased exposure on a portrait subject - the printing is showing results higher up on the film response curve. So again, no immediate revelations from this, on its own, but if you already know what you like, for some particular purpose, you can do the contrast comparisons to narrow down other possibly useful film/developer/exposure combinations to try out.

If you had all sorts of spare money and some assistants you might simply try every possible combination in the real world. So the main benefit, in my view, of making the step wedges, is to help you narrow down the variations that you want to study further.

Fwiw, what camera flare will do to real world film should be exactly equivalent (I think) to a slight fog exposure on the film - it will effectively give the film a longer toe. So you can probably mimic this with a very weak darkroom "flashing exposure" on your film wedges. As a warning it can get like a disease so don't forget about your first priorities as you dig into this.

 

[RalphLambrecht]

Hi, I think it is fundamentally a bad idea to rely on photographing a step wedge - the "shadow" parts of your curve are gonna be contaminated, by flare, to various extents, depending on how bright the brightest parts get, as well as your lens. I have personally tried variations of this, way back. Just as a demonstration, here's something you might try: set up and photograph a full-step wedge; then block off the brightest (most clear on the backlit wedge) and photograph it again. You may be surprised how much effect there is on the result, presumably the effect of lens flare. (If it's not objectionable to you, great!)

If you wanna study the film/developer characteristics, only, with no lens effect, I think you wanna stick with contact prints onto the film.

At the outfit where I used to work I used to oversee some in-depth film testing, so we had a couple of sensitometers; this is by far the best way to do it, in my opinion, IF one has an appropriate sensitometer. Without a formal sensitometer I would suggest rigging up a system. If I were gonna do it I'd probably base it on an electronic flash in a darkroom, maybe a hot-shoe style flash, firing at full power to keep the exposure as long as possible. If you fire it from a distance, to minimize side-to-side fall-off, and use some sort of diffuser over the front to make sure that no optical pattern is projected, I think this ought to work pretty well. Now, I have not actually done this, so there may be something I'm overlooking (does it have enough power, etc.).

I wouldn't initially worry about a "speed rating," using the contact print via flash mainly to make a relative characteristic curve (assuming that your step wedge has even steps, density-wise, you should get good results). As a later step you could actually shoot something through an actual camera, then correlate this actual metered exposure back to your exposed wedges (I don't knock exactly how, offhand, but it ought to be easy enough to figure out a way.)

For the mechanism of handling the film, you might use 35 mm film, winding from a fresh cassette into an empty one (designed for reloading, etc.). So you would only need to have about 6 or 8 inches exposed at once - whatever you need to hold your step wedge. If you've never done this before, it can be difficult to find a reference point on the wedge, so you might want to put several reference points on the edge of the step wedge - a little bit of black tape cut to a triangle point, or whatever, works.

Best of luck with the testing.

Ps, as one last check on your system, you might try exposing the wedge in both directions, just in case your light is uneven. And at the processing end, also try a set of opposing wedges - this will detect something likely a flow pattern problem.

Flare is indeed an issue with the purchase, however, they solve many of the exposure issues with other methods.

 

[grainyvision]

Well, one useful way is to make contact prints of your self-processed film wedges onto photo paper (perhaps a soft grade). So this roughly mimics the result of actually printing: you could even expose through the actual enlarger for a better result (it's the Callier coefficient thing, or whatever they call it now). To compare results you can butt the printed wedges side-by-side to see how individual steps vary. You might find, for example, that one developer gives gentler contrast in the shadow areas, but overtakes another film in the midtones, or whatever. I don't know that this is useful on its own, but... if you find that you really like a certain film-developer combination for something, portraits for example, you can study how that response differs from those you don't like. Even if not immediately useful you'll at least developed a sort of understanding of what's happening, contrast-wise.

Fwiw the film exposed in your sensitometer essentially covers all film exposure variations; this happens automatically as a result of the original step wedge being in equal density steps. So if you get to wondering about camera exposure, for example (what if I increased my camera exposure by an f-stop on portraits with this particular film?), no need to repeat the film exposure. Just take the one film test you have, and reprint it. If you print it darker this is equivalent to having increased exposure on a portrait subject - the printing is showing results higher up on the film response curve. So again, no immediate revelations from this, on its own, but if you already know what you like, for some particular purpose, you can do the contrast comparisons to narrow down other possibly useful film/developer/exposure combinations to try out.

If you had all sorts of spare money and some assistants you might simply try every possible combination in the real world. So the main benefit, in my view, of making the step wedges, is to help you narrow down the variations that you want to study further.

Fwiw, what camera flare will do to real world film should be exactly equivalent (I think) to a slight fog exposure on the film - it will effectively give the film a longer toe. So you can probably mimic this with a very weak darkroom "flashing exposure" on your film wedges. As a warning it can get like a disease so don't forget about your first priorities as you dig into this.

The biggest concern I've had is that developer formulation tests are already quite wasteful by necessity, but my goal is to be able to pinpoint and answer at least one part of the question "how does my developer compared to D-76" etc. Or "how does my developer fair when I double the dilution" or even "what happens when I increase development time by 20%". These are very tough questions to answer with much confidence without either step wedge exposures, or a perfectly controlled and consistent environment to do test shots. Figuring out a good test environment seems a lot more complex (need a good amount of space, great lighting, many small subjects, ensure time of day makes no difference, etc) than doing step wedge exposures. Step wedge of course won't meaningful answer the other characteristics like how is grain size or sharpness, but I tend to think the contrast behavior is the most important thing about a developer. Also the question of developer aging has been a huge concern as it has had a direct impact on real non-test photos in the past. I don't exactly expect to do super precise densitometer measurements, but even going by the bare eye it's pretty easy to figure out when things have changed significantly due to a developer stock aging poorly.

In short, I have a goal for formulating a few different developer types that I've not seen delivered in formulation in the way I want or that come with other compromises. My three big things progressing toward this has been EXG1, a normal contrast rodinal-type developer with better speed characteristics which works especially well for pushing. EXG2 which is an extremely compensating lower contrast developer with moderate grain and normal or very slightly increased speed. And EXG3, a high contrast developer explicitly designed to be as minimally compensating as possible and letting highlights separate significantly from shadows, but with still good prevention from clipping into dmax if properly exposed, and exhibiting surprisingly fine grain. None of these are really suitable to me if the stock solutions don't last more than a week without significant changes in contrast. and in tweaking these developers I need to ensure I'm still maintaining the properties I want out of them and what actual tangible differences can be seen by these modifications in terms of contrast and speed, especially the ability to objectively measure speed loss.

 

[ic-racer]

Related thread that may be of some help:

https://www.photrio.com/forum/threads/testing-some-new-film-shanghai-8x10-how-i-do-it.171836/

 

[ic-racer]

Just to emphasize that when I tested all my sensitometers, irrespective of the light source, they all gave the same answer to relative changes in contrast with changes in development TIME.

I think repeating the test with changes in development TEMP or development CHEMISTRY would give the same results, but could do those tests if someone had concerns using a inexpensive green or blue sensitometer for CHEMISTRY tests.

 

[bendytwin]

I want to get a bit more consistent in testing some custom film developer formulations and a step wedge is the most obvious way to do this. However, I typically like testing as many films in a developer as possible per run, and prefer to not use expensive sheet film. So, using a roll film like 120 or 35mm would be greatly preferred. However, exposing these aren't so simple in my tests under an enlarger. Film curl is a big annoyance, as well as the unpredictable ways in which enlarger exposure works. My enlarger timer does have 0.1s resolution, but I don't trust that to be too consistent since it doesn't use a regulated power supply or anything like that.

I'm considering just shooting a step wedge that is backlit and using the same rig as I have for DSLR scanning, but with a 35mm SLR and macro lens (if needed for 4x5). I understand lenses will have some effect on this, specifically some lenses are lower or higher in contrast due to coatings. My bigger concern is edge fall off which would change the step wedge exposure results, but that can be solved by using a good lens and leaving a bit of room in the frame around the step wedge. Anyway, does anyone have any better methods of exposing a step wedge onto roll film? Also, how would I meter in a situation like this?

I found this video from The Naked Photographer (posted just a few weeks ago!) to be an extremely practical and easy way to expose a 21-step tablet on roll film:



Incidentally, I'm wondering if anyone has disassembled an X-rite 334? I've never used one -- is the light source in it LED or something else through a filter? If it's easy enough to modify, they're so cheap on fleaBay that I want to buy two, then yank out and replace the required bits to keep one "BLUE - GREEN" but make another "BLUE - RED." Is that a pipe dream?

 

[Donald Qualls]

Incidentally, I'm wondering if anyone has disassembled an X-rite 334? I've never used one -- is the light source in it LED or something else through a filter? If it's easy enough to modify, they're so cheap on fleaBay that I want to buy two, then yank out and replace the required bits to keep one "BLUE - GREEN" but make another "BLUE - RED." Is that a pipe dream?

Actually, it's Part 1 (the video immediately before that one) that shows exposing the step wedge, and then reading the film on the densitometer after processing it in the development under test. The video linked above is about using the data from Part 1.

What would be the advantage of a "blue-red" densitometer? Seems to me that "blue-green" lets you read densities for both layers of multigrade paper. What would you gain with a red option? And why not make it a tri-color, if you're going to modify it?

Further on this, however -- I'm pretty sure those predate the invention of blue LEDs, so they're most likely incandescent bulbs...

 

[bendytwin]

My mistake! Thanks for the correction on the link -- I'd had both tabs open rather late last night, picked the wrong one.

I haven't fully thought out the possibilities of having a cheap tri-color sensitometer, but I have boatloads of expired 16 and 35mm color film stock and wanted a simple, repeatable, comparable way to measure RGB response between stocks of the same type but of different age/storage. I imagine an RGB sensitometer would be especially useful when trying to compensate experimentally for color casts by using filters or in development or with a dichroic source in making film prints, like it gets done with Laboratory Aim Densities. Though I suppose just shooting a GretagMacbeth chart or reading for disparity between RGB densities on a shot of an 18% gray card works fine too.

Assuming it's incandescents inside, they'd have to be filtered right? It seems obvious that the small microswitch control on the outside of the X-rite 334 between Blue and Green is an electronic contact and not mechanical (i.e., it's not physically sliding one filter out for another or something like that). If so changing out the filter should be doable but adding a third + bulb sounds like too much work -- hence my suggestion to just buy two and mod one. I should probably just spend the $15 + shipping to experiment myself.

By the way thanks for your reply in general Donald -- I appreciate the wealth of knowledge you and others in this thread have shared that has been such a great help to me over the years (I searched my notes just now and found posts of yours I'd saved as far back as 2004 from Photo.net and LFPF about everything from time/temp/dev dilution for ImageLink microfilm to lens collimation to chromogenic film in B&W chemistry!)

 

[Donald Qualls]

You're very welcome. My main qualifications are a pretty good memory and an inability to keep my figurative mouth shut.

Yep, incandescents would have filters. I suspect the switch just swaps between two bulbs, likely with individual filters and a diffuser. Making it tri-color could be as "simple" as replacing the switch and converting the light source (and power supply) to a set of tricolor LEDs with identical lumen outputs. For color, you might be ahead to change to a single continuous spectrum white light source (incandescent with a correction filter?) and have tricolor filters mounted on the sensor head. Easier mod to make and install. Seems likely a 3D printer would come in handy...