View Full Version : Collodion is only sensitive to visible light, not UV

Tim Meisburger
24-Jan-2016, 00:22
Yesterday we shot a lot of plates under constantly changing skies, from direct sun to heavily obscured by clouds. It was a pain in the butt, so this morning I have been looking at UV meters to try to determine if it would be useful to get one for wet plate. My (hour of) internet research has enabled to be definitively state that UV is not relevant for wet plate, as wetplate is only sensitive to visible light. Here is why I can can confidently state this with a level of certainty only the internet can provide:

There are three types of UV: A, B, and C.
UV C cannot penetrate the atmosphere.
UV B is 290 320 nanometer, but does not penetrate glass so cannot get through your lens to the collodion.
UV A is 320 400 nanometer, and can get through glass (unless it has a UV coating). 95% of the UV that reaches earth is UV A
According to Lund Photographic the spectral sensitivity of collodion is 403 to 521 nanometers. According to the internet violet (color) begins at 380 nanometers and extends to 450. This means that there is a disagreement between the description of UV A extending to 400 nanometers and the description of violet beginning at 380 nanometers, but both would agree that by the time you get to 403 nanometers you are in the visible spectrum.
So, collodion is sensitive to almost all of violet, all of blue, and half of green, so a lightmeter that measured only this section of the visible light spectrum could be accurately used to predict wet plate exposure.

I rest my case!

24-Jan-2016, 06:27
Good analysis Tim, you're probably right. I remember when I started, a lot of people figured using blacklight bulbs would be the Next Big Thing to allow short indoor exposures. But I noticed it never caught on, and those that tried it didn't get the fast exposures they were expecting. Probably most of the exposure was from the visible violet, like you say. But you should confirm the sensitivity of Collodion. Just because Lund is saying that's the range, they may have just pulled those numbers out of the air.

Drew Bedo
24-Jan-2016, 06:36
A UV filter on the lens would eliminate any exposure from UV-A. This would allow you to use a regular light meter and simplify the process.

Am I missing something?

24-Jan-2016, 07:10
You are right, that would make it where you don't have to factor for UV. But you don't really use a meter anyway. Wetplate collodion changes in sensitivity daily, going from possibly ISO 5 when fresh, to ISO 1 or slower as it ages over weeks and months. The aging isn't predictably either, it's based on temperature, light, and the particular iodizing salts you are using. Using a light meter would only get you close, which most people can do with an educated eye, and no meter. I've taught wetplate to people, and they can reliably calculate the exposure after a few shots, even in diminishing light.

Tim may be onto somthing, but I do suspect collodion is somewhat sensitive to intensities we cannot see. And our eyes cannot detect when one day has more "blue" light and the next day less, due to atmospheric conditions. But your exposures will change, based on time of day (It seems just as bright at 4PM as 11AM, but it's not) and the particular day, even at the same hour!

Tim Meisburger
24-Jan-2016, 15:21
Yes, I agree with you Garrett that a meter will never account for the changing nature of collodion, but it might have helped us in the variably cloudy conditions over four hours (or maybe not). But, it was pointless to make tests, as by the time you developed and were ready to shoot again the light had changed. One of the people who joined us kept trying to use her meter on and for the last shot (a group picture, in both subject and execution) recommended f/16 and six seconds, with predictable results. I poured another plate, looked at the sky, spit int he wind, and stated--oracle-like--32 seconds at f/8. Spot on!

24-Jan-2016, 15:26
That cloudy overcast got me one time too, shooting at a Paris photo fair with a friend, Quinn Jacobson. Every shot was radically different, 8 seconds overexposed. Then 4 seconds extremely dark. Then back to 8, dark again. Then to 15, good exposure, then the next one overexposed at that. We finally figured that there was a low ceiling, but different thicker clouds blowing over every few minutes. You couldn't tell below, it just looked uniformly fairly bright grey. We finally move away from several green trees that were heavily in leaf, and faced the sitter north. That helped. Stay out of the trees/jungle, remember I told you trees can do weird things too.

25-Jan-2016, 04:23
This has happened to me a couple of times, most notably probably late summer during days when overcast light is constantly changing, between patches of sunny sky.

I found for me the best way to overcome it was to practice developing by eye, which I'm still trying to get myself (hard under a red light)

I made a test plate in the overcast light and it gave me as expected 1 second exp bang on with 16 secs of development.

The proper plate I made about 15 minutes later under development showed to be massively underexposed, mids and shadows were very slow coming in so I took the risk and pushed development into the late 20 seconds until it looked good under the red light, plate came out perfect. I thought it may have had something to do with the light but I wasn't sure (light was changing) if it was that or my poor technique

This is an interesting thread thankyou for the information.... I'll keep an eye on this :) great to know.

Fr. Mark
25-Jan-2016, 16:09
I am skeptical. Nice hypothesis, and in a perfect world theory (or hypothesis) would agree perfectly with experiments. But, frequently theory and practice disagree.

There has to be some upper limit of the energy at which photons interact with collodion suspended silver halides. For example, most XRay photons would pass right through it. However, so much effort has been expended figuring out how increase both film speed and spectral range the last 150 years or so, and the energy differences are so slight (I think we'll with an order of magnitude between UV-A and infra-red) that I remain skeptical that there's a super sharp cut off excluding UV with collodion/Ag Halides. One Getty PDF looking at conservation properties lists a UV transmission of cellulose nitrate polymer (no silver halides) at 313 nm.

A Schott glass (no pun intended, it's a real Company) tech paper PDF suggests that for an inch of glass depending on glass the UV cutoff could be as low as 250 nm, but I think high 300's is more common. UV transmission of glass is dependent on the kind of glass and its thickness. Many lenses have less than an inch of glass so ymmv.

Honestly, I'm inclined to believe you are right, that UV may not matter for wet plate. But, the scientist in me would like to see some data.

I suspect that exposure guidelines like in old 35mm film packaging i.e. If it's open shade do this... Combined with experience and good notes (lens, conditions, age and kind of emulsion, how you timed the exposure, etc) will be more reliable than a meter.

I admire y'all wet plate folks. I'm not quite ready to play with such highly energetic materials nor tote a darkroom everywhere. I think if I'm going to make "film" it's going to be gelatin emulsions. I would be curious to know if anyone's made Ortho chromatic or even pan film versions of wet plate. I strongly suspect both could be done since it has been done with gelatin emulsions even by do it yourself-ers. I greatly prefer the look of at least Ortho films.

Tim Meisburger
25-Jan-2016, 16:54
I simply don't know, as the information I have been able to access so far is inconclusive or contradictory. I would be surprised if no one has ever looked at this in a serious way, so I'm still hoping to track down more conclusive info.

It would be interesting to look at uv transmission of various lens designs. Should it be possible to set a few examples on a sheet of cyanotype and expose through them with a UV light source (maybe the sun?) and compare the density of the various lens circles with the surrounding paper.

25-Jan-2016, 18:15
I also believe at least some of the invisible UV spectrum is sensitive. One thing that we wetplaters today discuss, and they did in the old days, is the "actinic light." We use Northern light, because it has more of it. I think some UV gets through.

Fr. Mark
25-Jan-2016, 18:23
First step is to see what a UV filter does in terms of blocking UV. Then look at lens/cyanotype paper with and without the filter in front. And due to different kinds of glass each lens will likely be different. I'd also suggest keeping track of exposure time and if you live where I do, use a repeatable UV source not the sun.

Tim Meisburger
26-Jan-2016, 01:13
Well, I don't think I will do it, because I don't have a UV source, but if someone does, they might want to give it a try.

Struan Gray
27-Jan-2016, 07:24
Wet plate collodion was used to take spectra from individual stars as early as the 1870s. Henry Draper is usually given priority, with his spectrum of Vega in 1892, which showed what would today be considered UV lines.

Bare silver chloride is also most certainly exposed by UV light - it's how Johann Ritter discovered ultraviolet light in 1801.

Whether any of this has practical significance for wet plate photography is another matter, but early journals are full of discussions of the difference between actinic and visual focus, so I'd say yes, in general. UV-only or UV-rich illumination for taking photographs does not seem necessary, but that doesn't mean there are no effects - enhanced haze in alpine scenes, for example.

Struan Gray
27-Jan-2016, 09:46
Oops. Spectrum of Vega in 1872.

Fr. Mark
27-Jan-2016, 18:06
1872 spectrometry! Amazing. It was several decades after that before a fairly firm theoretical basis for those lines was developed.

I'm also amazed that there was enough open time with collodion to get a bright enough image to collect.

One last thought, a couple summers ago I took a photo of my brother with an Ortho X-ray film and his skin in the photo was blotchy in a way that was not obvious to just look at him. I'm suspicious it was because of sunscreen and uv absorption.

Can't prove it. Was a huge projector lens with 2 or three inches of glass in the optical path.

Maybe some day I will deliberately test this notion.

Fr. Mark
27-Jan-2016, 18:07
Can you re-wet collodion plates with ether and alcohol they've dried out? Or does it come
Off the plate?

Struan Gray
28-Jan-2016, 05:04
Fr. Mark: they did switch to dry plates as soon as they became available/known of. The speed increase was the major reason, but Vogel discovered dye sensitisers in 1873 and although it took another decade or so before good sensitisers were available for gelatin emulsions, there was a great deal if interest in the possibility of tuning spectral sensitivity among spectroscopists - astronomers and others.

Blue-sensitive emulsions show up skin damage much more clearly than red or IR responsive ones. Old scars and sunburn patterns show up nicely, even without a UV component. Some forensic photography uses near UV photography to show up bruising and other trauma which are not evident in a colour or red-sensitive B+W shot. Pure UV makes things even more dramatic:


Easy way to check the influence of UV: grab the UV blocking gel from a set of printing filters and take one pic with, one without. There are more expensive solutions too :-)

Tim Meisburger
28-Jan-2016, 07:05
Does a UV blocking gel (whatever that is) only block UV, with no effect on near UV violet? Someone else said that the proof UV affects collodion is that you can seen sunscreen, but I would not be surprised if sunscreen also blocks light in the near UV.

28-Jan-2016, 10:00
With regards to the sunscreen, you should check the effects under a violet coloured filter with your eyes, and then compare to the effect captured on your film.

But if you really want to test spectral sensitivity, then setting up an actual spectrometer pointing at something that emits an even and full spectrum light would be a rather good option. The equipment isn't that complex if you don't need super reliable accuracy and precision, and are just looking for a more ballpark kind of figure.

But you do have to remember to account for issues such as different lenses transmitting and absorbing different wavelengths of light, and it is one of the things that makes wide spectrum data collection 'interesting', as it can be awkward to calibrate if you have multiple lenses in use, and made more awkward if the lens happens to cause radical changes in focus for different ranges of non-visible light.

A project I worked on several years ago had multiple optics in play, and our telephoto lens let far less UV through than the wide angle lens we used, which made for some very confusing experiments due to a mix up in a spec sheet that had the wrong transmission values. Nothing like trying to figure out why the robot keeps missing target when pointing the second camera at things when no one has yet realized the second camera can't even actually see it.

Struan Gray
28-Jan-2016, 10:47
Tim, the UV printing gels I had in my Ilfochrome/Cibachrome filter packs were fairly clear, with no real colour cast. Perhaps a tiny touch of yellow, which would suppress some near-UV blues and indigoes, but nothing I would call a tint.

I think A_Tabor is right, you're asking questions that can only really be answered once you are in possession of some spectra. Homemade UV-vis spectrometers can be quite good, but calibration can be an issue with unknown lens transmissions and grating reflectivities.

Tim Meisburger
28-Jan-2016, 16:41
That sounds reasonable. I don't have the equipment (or brains) to do this sort of thing, but I think that this kind of objective testing is clearly what is lacking. I'm not sure I know enough about chemistry, optics or physics to pose intelligent questions, but this is what comes to mind:

- spectral sensitivity of fresh collodion
- change in spectral sensitivity of collodion over time (older collodion has less speed, but does its sensitivity to various spectra change as well?)
- spectral transmission of lens designs (long/wide, triplets, etc) and coated vs. uncoated.

Garrett and others have pointed out that its easiest to just shoot a plate and get on with it, and I agree. But I've been made curious by the contradictory assertions, and limited data that seems to suggest that the common belief that UV is important in wet plate may be unfounded. Or maybe not. Just a bit o' fun, like a crossword, with limited practical implications.

28-Jan-2016, 17:45
The road to science is all about asking dumb questions and finding the answers to them till you learn enough to start asking better questions and figuring out the harder answers.

I will admit that I am highly biased because I'm a geek and I enjoy optics, but I would encourage you to spend time reading up about some of the early works and experiments done by the pioneers of the field. Science is fun and interesting, and the more you learn about something then the more questions you can ask and things you can poke at or play with. And photography is all about light, so the more you understand it then the more you can control it and compose images exactly to what you want them to be.

If you do want to get into experimenting with this kind of thing and doing your own tests, then getting a fairly simple and cheap triangular prism is probably a good place to start. Cutting edge labs might have billions of dollars worth of gear and gadgets, but much of the early work that we base all our current knowledge on was done with shockingly simple tools like what a $5 prism is today.

Get yourself a few dispersive prisms, and play around with them. Shouldn't be too terribly hard to rig up something to throw a rainbow pattern into a dark box and onto a white sheet you could observe, then angle the pattern so that you have the violet and where you expect the UV to be. Carefully mark the visible parts and where it ends, and then without moving the prism or box replace the target paper with the light sensitive media you want to test. Develop, and compare to your marked target and see how far into the UV things go.

Testing through an actual lens would require a little more setup. But one thing to remember when testing this kind of thing is that you really only need narrow strips rather than using full plates at a time.

Hopefully others can chime in with ideas and suggestions. I might sketch something up this weekend if others feel it would be useful to the discussion.

Remember: Science is fun. (At least till things catch fire, then they're exciting.)

Tim Meisburger
29-Jan-2016, 01:15
What a great idea! I'm not sure where I can buy a prism here, but I'm sure they would be available somewhere.

Struan Gray
29-Jan-2016, 10:12
Transmission gratings are reasonably cheap on eBay, some with highly UV-friendly substrates such as fused silica. A dispersing prism would also work, but their transmission can often drop off fast in the UV or IR (that's why they disperse :-)

Personally though, I would make a bar of LEDs of varying wavelength and photograph that. Make it small enough to contact print, and you can avoid lens transmission issues if you're interested in the emulsion only. With some UV LEDs you get fluorescence from the plastic packaging, so if you're a purist you could add a UV-only filter for the LED part.

Fr. Mark
29-Jan-2016, 21:34
Science is fun...til something catches on fire...then it is EXCITING!

How true, how true. Been there done that, thankfully, lived to tell about it. Perhaps the most dramatic incident:

Saw a professor flash burn off his eyebrows once in a lab "oops" that thankfully did not ignite the 30 gallons of ethanol/KOH cleaning bath next to the sink or we might have been front page news...there was a tiny amount of sodium metal a smidgen of water for ignition and maybe a 1/4 cup of acetone involved as fuel and maybe some residual isopropanol. I had my back to it, heard a whoomp sound, felt a flash of heat and just managed to see the flames licking the 14 foot high ceiling before it was over. Total elapsed time, maybe three seconds.

I think I mentioned that some of our favorite chemicals are highly energetic. Wear safety glasses and gloves, be careful and have fun.

2-Feb-2016, 03:14
Just a thought: might the 1872 spectragraphs have been taken with a reflecting telescope, with no glass in the way to eat any UV? Seems a reasonable assumption; most astronomy uses reflective optics.


Struan Gray
2-Feb-2016, 04:57
Draper used a reflector for the Vega spectrum, but switched to a refractor later.