Ilford PAN-F + Reciprocity

[lensman_nh]

I come to collective in search of wisdom.

PAN-F is one of my favorite films, and I love both the grain and tonal quality.

But today I was using some fairly dense ND filters to smooth out water motion. The metered exposure (hand held spot) was around 1 second at f22. With a 3 stop ND filter that takes me to (2, 4, 8) 8 seconds.

The data sheet gives a reciprocity curve that says the exposure should be about 20 seconds. Which is what I gave. The frames are way overexposed, so much so it looks like 8 seconds would probably have been right. Has anyone else noticed that the given reciprocity curve seems to way overexpose? Perhaps it works better at very low EVs (nighttime) rather than very heavy ND filters, though that doesn't seem to make sense.

The other "Normal" exposures on the roll (1s or less) are just fine, so it's not a systematic failure.

I'm planning to conduct a test of my filters with metered + filter factor, then an additional +1 and +2 stops to figure out my own reciprocity curve.

Any wisdom here to impart?

J.

 

[Disconnekt]

You would have to use the XY button (same as the YX button) on a calulator to find the new time.
You would put in 10XY1.33, so the new time would be 15.88 seconds, rounded up would be 16 seconds

 

[lensman_nh]

Urrm, what XY button where? I don't follow your logic here.

1second + 3 stops is 8 seconds (1 doubles to 2, 2 doubles to 4, 4 doubles to 8 - 3 stops). If you want a log scale, it's an ND 0.9.

Then take reciprocity into account using this handy, dandy graph in the datasheet for PAN-F, and you get about 25 seconds, give or take. But 25 seconds overexposes, so the datasheet seems iffy.

J.

 

[Disconnekt]

Google "online calculater", and when it pops up enter "8" then click the XY button (next to the 0), then enter 1.33 (the time Ilford gives for Pan F+) and the time comes out to 15.88 seconds.

Ilford gives more info on it on their Reciprocity Failure Compinsation pdf for all their films.

 

[lensman_nh]

Where are you getting the magic 1.33 number?

 

[Disconnekt]

From the pdf I attached in my last post, straight from Ilford.

 

[lensman_nh]

Ah got it.

Interestingly the graph I posted gives about 25 seconds, from the PAN-F documentation, so the Ilford documentation is not consistent.

That said, the frames are more than 1 stop overexposed (25 seconds is about 2/3 stop), so even that 1.33 factor appears way too high, unless something else is in play. They look to be 2 stops over.

The exposures of 1s or less are fine.

I'll run a controlled test and report back.

 

[MattKing]

When reciprocity failure is in play, metering technique can be challenging.
For example, the shadows in your scene may reflect so little light that proper exposure puts the film into reciprocity failure, while at the same time the mid-tones and highlights in your scene may reflect enough light that proper exposure doesn't involve any reciprocity failure.
So if you take a meter reading from a shadow area in order to determine your base exposure, that may lead you into correcting for a reciprocity failure not needed for the majority of the image.

 

[Poisson Du Jour]

Urrm, what XY button where? I don't follow your logic here.

1second + 3 stops is 8 seconds (1 doubles to 2, 2 doubles to 4, 4 doubles to 8 - 3 stops). If you want a log scale, it's an ND 0.9.

Then take reciprocity into account using this handy, dandy graph in the datasheet for PAN-F, and you get about 25 seconds, give or take. But 25 seconds overexposes, so the datasheet seems iffy.

J.

Xy is also intechangeable with Xn on Sci calculators..

 

[GLS]

Just be sure when using the above factor of 1.33 that you input the measured exposure in seconds, otherwise you will get an erroneous answer.

These newer published reciprocity factors from Ilford are far more accurate than the adjustment curve that is used in the data sheets. As far as I understand they publish the same curve for all their films as an averaged, middle-of-the-road solution, which to my mind is just too oversimplified and makes no sense. The real factors can be quite different from one another, and can therefore give very large differences in the accurately-adjusted times, especially for long exposures.

Long story short: use the new factor table. It's easy to do with any smartphone.

 

[Tim V]

Going by the Reciprocity Timer app on my iPhone, 8 seconds metered with Pan F+ extends to 12 seconds when accounting for reciprocity.