I've been experimenting with Infrared film recently - specifically Rollei 400 and Ilford's SFX 200. I've shot quite a few rolls through my Mamiya RB67 to try and decide which film is best and have ended up with some wild and varied results from the same scene.
A little reading up later, and I discovered reciprocity failure which typically affects long exposures. What (very simply) happens is that the light takes longer to affect the film surface and thus for exposures greater than 1 second you often need to compensate for this with longer than normal exposure times.
I use a pair of Zomei 720nm IR filters for my film work, which is an extremely deep red filter that blocks around 10 stops of light. Yes. I'm going to be using the words "around" and "roughly" quite a bit. This isn't a scientific test, IR photography depends on the amount of IR light available. And unfortunately, not being able to see IR light makes it difficult to expose accurately.
There is also precious little information on the internet. The Ilford datasheet for SFX suggests taking your metered exposure time and raising it to the power of 1.43 to determine the actual exposure. So a 60 second exposure would actually be 348 seconds (5 minutes!!). This is quite a scary increase, and it gets worse the longer your exposure is.
However - remember that whilst 1 minute to 5 minutes sounds like a massive increase, it's actually just over 2 stops (and nowhere near 3!) and this is pretty much the crux of this blog post.
I took a bunch of test shots.
On the left is the base shot (in this case 1/250sec). Then next to it is 10 stops -1, then 10+0 then 10+1 then 10+2 then 10+3, which in this case is 2, 4, 8, 15 & 30 seconds. As you can (perhaps) see, the +1 exposure is actually closer to the base exposure.
So then we repeat for a bunch of base exposure times...
The contact sheet above is really quite difficult to represent well on the internet, and to determine precise results, I pulled everything into Lightroom and got comparing base vs each of the shots.
The most interesting find was that the highlights are the first to recover. Often, when (for example) the highlights are back to base exposure at +1, the shadows don't recover until +2. This means more sacrifice at longer exposures. Do you choose to expose for the shadows, or to preserve your highlights. Because of this, the exposure overall feels more contrasty which might be good or bad depending on what you're doing.
The result of all of this, is the following. If your base unfiltered shutter speed is...
1/250sec, +10 stops = 4 sec, Rollei 400 IR needs 8sec (+1 stop).
1/125sec, +10 stops = 8 sec, Rollei 400 IR needs 20sec (+1-2 stops).
1/60sec, +10 stops = 15 sec, Rollei 400 IR needs 45sec (+1-2 stops).
1/30sec, +10 stops = 30 sec, Rollei 400 IR needs 120sec (+2 stops).
1/15sec, +10 stops = 60 sec, Rollei 400 IR needs 360sec (+2.5 stops).
If your base unfiltered shutter speed is faster than 1/250sec then the exposure is so close to a second that reciprocity failure isn't going to affect your image. And if it's slower... Well that's longer than I'm prepared to hang around, but I'd have a good guess at around 15-20 minutes for a base unfiltered 1/8sec exposure. Fortunately, to be shooting at 1/8sec on 400 speed film means either really bad light or super narrow aperture. And with my lenses, I can achieve infinity focussing relatively easily at f16-22 so it's extremely rare for me to get down that low.
Also of interest is that these times are remarkably similar to the SFX times which makes working with either film much easier. Also, with the Rollei 400 I can see lots of latitude in the exposure which means guessing will work pretty well out in the field.
So that's the result of 2 rolls of Rollei 400. I'll try SFX when I get the impetus again, but it's exceedingly dull to do these tests and easy to get mixed up as you go... You don't need to memorise these numbers. If I was out and about without my books or just couldn't be bothered, and had deep shadows and a contrasty scene, I'd probably add 2 stops. If I was desperate to protect the highlights, I'd probably add just 1.
I can do science me.
As to which film I prefer? The Rollei. My initial tests showed it to have nicer grain, better contrast and better overall sharpness. The sharpness could be down to the fact that SFX is really curled and thus is getting scanned out of focus (the Rollei is cheap and doesn't curl), or it could be down to just a characteristic of the film.
Being less than half the price of the SFX though has sold it for me. I'd still be inclined to bracket for +1 and +2 to be sure, and on 10 shots per roll that eats through film quick. And that's not even thinking about the Tomiyama which gives me just 4 exposures on a roll of 120.
Phew! I think an arty post next time.
Thanks for reading!