Garrett
flickr galleries
Why are so many people who shoot wet plate arrogant jerks who assume that not one else actually shoots?
Yes, I actually shoot wet plate. I will admit that I can only afford to do it one day a week at this point, but that still means I'm making a couple dozen plates a month. I also realize that daylight studios exist. I don't happen to own one. Nor, in this part of Ohio, would it be a guaranteed success as we have full sun just 67 days a year. High overcast would be fine as well but unfortunately we have a ton of days where the direct sunlight (well, through the clouds, I just mean not in the shade) exposure has gone as long as 45 seconds at f/2.8 for me. That's crap for shooting people.
Having a bad day, sorry. But don't like people telling me to 'be serious', when I was being serious. It was offensive, so I came back at him.
Garrett
flickr galleries
Roughly 4000-4800Ws (additive exposure from 2x200Ws strobes, so possibly limited through reciprocity failure) is what I use for close-lit (1-3ft.) still life collodion negatives at around f/8-f/11 - so with roughly 1 stop loss in bellows extension. Note that negatives require ca. 2 stops more exposure for pleasant tonality than positives. For positives and half-body portraiture, I get away with about 2x200Ws (so two separate strobes) @ f/4.5, strobes at about 3ft, with very little loss in bellows extension (<0.5 stop).
With strobes, things get complicated if (1) you want to stop down to f/8-f/11 or (2) you want to use light modifiers (e.g. softbox) or (3) you want to shoot negatives (like I do) instead of positives. Under these constraints, there's no life below 2000Ws I find.
When working with models, a continuous light source will in my opinion always be problematic when doing wet plate. Best case scenario (LOTS!! of light) you may get your exposures down to a second, which is still long. I'll settle with strobes.
Btw: I checked the spectral sensitivity thing; I took the band gap data of AgBr and AgI (the halides present in a sensitized collodion emulsion) and calculated to which cutoff frequencies they correspond, which are 477nm (AgBr) and 517nm (AgI). This means that a collodion emulsion will not have any sensitivity beyond 517nm (which is in the blue part of the spectrum), corresponding quite nicely with the Lund spectral sensitivity charts available online (specifically this one here). The peak sensitivity around 475nm is also explained by these band gaps, as at that point, both AgI and AgBr are responsive.
Now, one might expect that below 475nm the sensitvity should remain high - after all, smaller wavelength photons have higher energies and therefore UV photons should be able to excite AgX grains as well as ones around the optimal wavelength. However, some problems emerge at the lower wavelengths: firstly, energy efficiency decreases - while close to the band gap, essentially all photon energy is used effectively, while at smaller wavelengths, part of the energy of each photon is wasted - excitation of an AgX kernel is essentially a binary thing. Secondly, the smaller the wavelength, the higher the attenuation in a medium. Lund used a prism to create the spectrum used in his test and since a prism is quite a hefty chunk of glass, particularly the lower UV wavelengths will be significantly attenuated before they reach the collodion plate (the same is true for the glass of the lens used). Even window glass (which is only a few mm thick) will block 90% of the radiation below 300nm; a thicker chunk of glass (prism and/or lens) will already significantly attenuate light around 350nm. Additionally, since Lund used a solar spectrum, attenuation of UV by the atmosphere plays a role as well, as much of the UV is blocked, the extent to which depending on atmospheric conditions. This implies that the Lund test is not entirely linear and the response curve is biased by the measurement method.
In conclusion, maximum photon efficiency is achieved around 475nm, which is a deep blue. Sensitivity below this wavelength is essentially the same given the same overall photon flux, but attenuation reduces efficiency as the wavelength gets lower. It would be interesting to test a monochromatic light source around 475nm and one around 400nm (at the edge of UV), both with the same photon flux. The sensitivity should in fact be comparable, as there is no reason to suspect that the sensitivity of collodion would be limited at lower wavelengths. However, there is also no benefit to be expected in using UV vs. deep blue light and overall energy efficiency will in fact be lower.
For practical purposes, this means that for continuous/'hot' lights, a light monochromatic light source of 475nm would be an ideal compromise: it's efficient, doesn't carry the risk of UV damage to subjects/sitters, its light quality can be judged visually and availability of high-power LEDs at this wavelength is good.
For strobes, the situation is a little different, as strobes are of course not monochromatic (discounting pulsed laser); a Xenon flash (unfiltered!) can output a significant part of its energy in the UV spectrum, depending on the power at which is driven and glass envelopes/filtration used (particularly relevant for the sub-350nm spectrum). This means that in a flash exposure on collodion, a significant part of the image is actually formed by UV in addition to the blue light exposure. Due to the short distance, atmospheric attenuation doesn't play an important role as is the case with a solar exposure. The implication is that the quality/nature of the light will be different from a pure blue exposure (as in blue LEDs), and it will also be more unpredictable due to our inability to directly see UV.
TL;DR: blue light at 475nm is ideal, but UV is not wasted on collodion.
Last edited by koraks; 29-Jun-2018 at 02:44.
hi folks,
currently working with 4 of these at the opera in zurich I wonder if I should get one of these
https://www.youtube.com/watch?v=3Fvn62luJFc
my boss told me we also have a cheaper option to the Arri ones... and that i could borrow them for a test... which i will do for sure.
my only concern is to built a octabox around it that fits somehow...
but first i will try this bugger...
my other Idea is t either use a Arri 2,5KW HMI or maybe give it a try with HQI as both of them have a high amount of UV output... but to not waste money and energy the LED version is best for sure...
what do you think about the LED solution for a romantic indoor session
Farid,
All of those surely put out the right kind of light for wet plate but at significant expense and without knowing how long your exposures will be without testing. Arri is one of the top brands of light you can buy in the filmmaking world but there are flash setups out there you can get for 1/5th or less the price for wet plate. Look into Speedotron or maybe even Norman 2400 or 4800w/s power packs and strobes.
If you're still thinking Arri, I would suggest renting one of each light type and testing them before buying.
3 secs @ F9 using 2x 100watt uv LED spot lights that cost $60 each (same as the types they use in night clubs etc ). Have lights connected to a enlarger timer triggered by my tow. No lag, open shutter with bulb in dimly red lit room just before and close shutter once lights go out.
They are cheap not very bright compared to other UV's Ive been using but very effective
Should be able to get 1.5secs with F5.6, could light the room just before to get eyes better accustomed.
Just need to better set the lights and fix all the other wet plate variables
.
Paul
Tin Can
Bookmarks