Rollei IR400 - ghetto DIY hillbilly reciprocity testing - check my work?

[M Carter]

Rollei states their IR400 film needs no correction up to 1/2 second, and provides no data beyond that. I'm going to be testing it in a 4x5 pinhole and a 4x5 view camera with various filters and expecting some long exposure times. (I've used it extensively with just a red filter, but want to try a full IR filter and also the pinhole).

So, not being a densitometer owner, I set up a constant-lit set with some seamless paper, so I had half the frame on zone III and half on zone V (I know, I should test for a lower zone, but I tend to get my shadows closer to zone III than IV). I had a roll of 35 around, so I did this:

Shot reference frames at 1/2 second at all f-stops, so I'd have a strip of film showing how the densities look in 1/2 stop increments. So just a strip of film fading to blank. Just so i could visually "see" what a half stop looks like - I figured i could align this with the recip. frames for a visual idea.

Then I shot some frames to show reciprocity - used the aperture to control exposure, and shot at 1/2, 1, 2, 4, 15, and 30 seconds. So if you'd done this with, say, Acros, all the frames should appear the same (I used an N90s which I assume has a fairly accurate shutter) - any differences in density would be down to shutter or lens inaccuracies on a film with no reciprocity corrections needed out to 30 seconds - right?

I processed that part of the roll. The 2nd strip (where I changed exposure times) - sure enough, those frames fade to near-transparent (film base + fog), with the 30 second time being very faint. So the recip. is very obvious visually.

By eyeballing on a light box with uniform brightness, I visually compared to the control strip to get a rough idea of how many stops the film was losing due to reciprocity - I came up with this:

2 secs 2.5 secs
4 secs 6 secs
8 secs 14.5 secs
15 secs 30 secs
30 secs 75 secs

But - I'm guessing this isn't the correct correction, since correcting 15 seconds metered exposure to 30 seconds per my tests wouldn't take into account that a 30 second exposure needs more exposure to correct for the extra reciprocity, if that makes sense?

What's my next step? I know there's some sort of formula, but I've never seen instructions on how to use that data - and translate that formula to a pocket calculator… or…whatever?

 

[Andrew O'Neill]

I ran a test a few years ago. 1 sec give 1.5x more; 2 sec give 2x; 4 sec give 2.5x; 8 sec give 3x; 16 sec give 4x; 32 give 5x; 64 give 7x... 256 sec give 15x

The curves generated also show that contrast is consistent throughout. I don't bother with any development compensation.
I've use this data successfully in the field.

 

[M Carter]

Thank you, I'll give that a shot. Though now I'm kind of wondering how I get to useable numbers based on the numbers from my first test...

 

[DWThomas]

Hmmm -- to begin with I consider pinhole to be an inexact science so as to rein in my expectations -- and adding IR, which has its own hazards; as eyes and meters do not see what the film sees really invites an adventure. So I would plan on a lot of bracketing! Andrew's numbers sound like a reasonable starting point for some empirical tests. The problem with IR is the percentage of it in relation to visible light appears to vary widely with atmosphere, season of the year, sun angle, and whatever, so there is not much substitute for shooting plus or minus a stop or two and picking the best. One memorable attempt of mine involved ripples in a river in the foreground surrounded by trees on the banks, etc. I tried twice over a wide range of exposures (and this with a lens and shorter exposures, so less reciprocity involved) and still there was one portion of that river that acted like a black hole -- frustrating!

I have noted in the past that with pinhole and substantial reciprocity correction a folding lawn chair is useful equipment. Pinhole Designer can output a table with the compensated exposures for a number of films and with my typical f/250 to f/300 pinholes the default table gets into hours alarmingly quickly.