This is related to Bob's question below. It seems most workers rate B&W films at less than the manufacturers published speed, unless they push. Many people suggest calibrating your system using transparancy film, as exposure is so much more critical, yet I don't see people making huge changes from the manufacturers published speeds on that. Is the measurement of speed different for the different materials? Or is it that transparancies are metered more for the highlights, whereas negs are metered more for the shadows? Or is 0.1 over b+f for b&w films really a valid criteria at all? Equipment accuracy today is pretty good, yet it seems that accurate exposure is as personal now as it was 50 years ago!

-- Conrad Hoffman (choffman@rpa.net), February 13, 2002

Answers

Hi Conrad. Yes, there are different speed methods for a variety of materials. I have only worked specifically with color negative materials, but am familiar with the B&W method and somewhat with color transparency (haven't seen ANSI or ISO docs on that, however). The generally good book, "Basic Photographic Processes and Materials" really only talks about the standard B&W method, but doesn't make this clear; thus one might assume that all materials use this method. Below is a summary of the common film types; I can elaborate more on this if anyone is interested.

Very roughly, B&W continuous tone neg materials use a speed method where a specified developer is used (they give the formula and aim pH of both developer and fixer as well as agitation method); development time is varied until you get a result such that when one exposure gives 0.10 density above base+fog and an exposure that is 1.30 log(10) greater (this is a factor of 20X more exposure) results in a density 0.80 higher than the lower density. When this condition is achieved, the exposure that produced 0.10 above base + fog is used to get the ANSI/ISO speed.

For color neg materials, the processing is specified by the material manufacturer. So presumably Kodak materials would be via C-41 process and Fuji would be via their equivalent process (I don't remember the name). They have a different method of finding the speed point; it is based on an exposure down low on the characteristic curve (similar to the B&W material); since the three color dye layers don't necessarily hit this at the same exposure level, there is a weighted calculation method. I have read comments by some people that color is not considered in this; they are wrong!

Color transparency materials use a different method based on an exposure roughly midway up the characteristic curve. So it would seem that these should generally correlate better to an exposure meter reading than the neg materials.

-- Bill C (bcarriel@cpicorp.com), February 13, 2002.

I could probably find this on the web in 10 minutes of searching, but I think the answer might be of more general interest.
Does anyone know, offhand, what the relationship is between the exposure in Lux-seconds to get 0.1D over base + fog, and the ISO rating? That is, there should be a simple multiplier or constant to arrive at the ISO speed from the 0.1D+B+F exposure figure. Anyone know exactly what it is?
From rough estimation it appears to be somewhere in the region of 250000, but I'm suffering from a hangover this morning, so that figure might easily be a mile off!

-- Pete Andrews (p.l.andrews@bham.ac.uk), February 14, 2002.

Yep! That hangover's certainly playing havoc with the old brain cells.
The formula appears to be 1/Exposure*K. Where K is around 1.5. I'd still like the exact figure, please.

250000? Where the heck did I get that from?

-- Pete Andrews (p.l.andrews@bham.ac.uk), February 14, 2002.

Pete, this site's pretty good- http://johnlind.tripod.com/science/scienceexposure.html

-- Conrad Hoffman (choffman@rpa.net), February 14, 2002.

Thanks Conrad. It seems that everything was on that page, except what I wanted to know!
I did a more accurate analysis of manufacturer's sensitometry curves, last night, and it seems my guesstimate of K=1.5 might be pretty close.

I converted the Log H value of exposure which gave 0.1D +B+F on the makers curves to linear lux-seconds. Then I applied the formula 1/(lux-seconds*1.5), and this gave me realistic EI figures for most of the films I looked at.
For example: T-max100 in D-76 (normal time) was giving an EI of 80; Fuji neopan 400 in D-76 was reading EI 325; and TMY400 in D-76 was giving an EI of 300.
All these figures are very believable, and tally pretty well with my own speed ratings, and those that others have posted on this forum.

It's my belief that by using the maker's published curves, it's possible to get more realistic speed ratings than the ISO figure. Ignore anything it says on the box, and just take the exposure required to get 0.1D above base and fog from the relevant published film/developer curve.
This doesn't work with Ilford films though, unfortunately, since they still label the exposure axis in the antiquated Relative-log form, and not in absolute Log Lux-seconds.
If David Carper is reading this, Ilford please note!!

-- Pete Andrews (p.l.andrews@bham.ac.uk), February 15, 2002.

Pete, how about k = 1.25? I don't know what the basis of this is, but using your formula it seems to produce the correct answer.

Way, way back doing speed testing, I just used the tables that are part of the standard. Their method for ending up with only the allowed speeds (100, 125, 160, etc) was to establish exposure ranges for each allowed speed.

For the logarithmic speed, try speed = 1 + 10* log(0.8/lux_seconds) (using a base 10 log).

-- Bill C (bcarriel@cpicorp.com), February 15, 2002.

Pete,

After Multigrade IV, it seems that no one I know of (including myself) could reproduce Ilford's published curves for papers. I don't know if they are using strange measurement settings, or making silent changes, but I think published data are untrustworthy unless proven otherwise!! In fairness to Ilford, TMX data sheet has two major types of curve responses, one straight with T-MAX developer, another shouldered with D-76 on the contrary to John Hicks saying TMX is impossible to put shoulder. I am curious if he found similar descrepancies for other films...

-- Ryuji Suzuki (rsuzuki@rs.cncdsl.com), February 17, 2002.

TMX can have a little shoulder but it's so high up the scale it's of no practical use.

-- John Hicks (jhicks31@bellsouth.net), February 17, 2002.

Thanks everyone.
I found the answer in a textbook that I hadn't looked at for years.
For future reference, the official ASA rating was derived by using K = 1.25, or 0.8/(lux-seconds). Bill C was bang on.
That's the 1962 ASA standard, which I think still holds as the current ISO standard.

It does seem that the recommended ISO development gives a curve that most people would find too steep for their liking. In effect, it appears that the ISO speed is derived from a slightly pushed development.
It also means that colour negative film can't be given a proper ISO rating, since its development is standardised away from the ISO recommendation.

-- Pete Andrews (p.l.andrews@bham.ac.uk), February 18, 2002.