Turtle
Veteran
If 1/3 of a stop never matters to you, you would appear to be consistently overexposing and so can afford to lose it 
I find myself very mobile in EI too, but it depends not only on the light but how I am able to do the metering. If I am not able to reliabily meter shadows I will set a much lower thena normal film speed if I am in bright contrastyn light - maybe 200 and work from there, but as soon as I hit a good shadow to work from, I will use 400 to determine whether that leaves me.
My conclusion is that there is no simple rule, only a good understanding of light and your materials...
I find myself very mobile in EI too, but it depends not only on the light but how I am able to do the metering. If I am not able to reliabily meter shadows I will set a much lower thena normal film speed if I am in bright contrastyn light - maybe 200 and work from there, but as soon as I hit a good shadow to work from, I will use 400 to determine whether that leaves me.
My conclusion is that there is no simple rule, only a good understanding of light and your materials...
Stuart John
Well-known
I tend to just dial in about a stop of extra exposure or be generous if I am using sunny 16 rules and then pull back on the development. Enjoyed Tom's car photos very much.
Turtle
Veteran
I suppose it matters on how one interprets the question:
1. 1/3 stop does not matter to me because I don't care about that level of accuracy or//
2. 1/3 stop just does not matter because of the other variables.
If its the first point, then I can't argue with that, but in the case of the second its just not correct.
1/3 of a stop makes 1/3 of a stop of difference. No more, no less and 1/3 of a stop can make a huge difference when you need it in a neg, but don't have it. Those who make serious prints will know how when you have perfectly exposed neg, you may have quite a bit of wiggle room, but when you are right on the edge of critical underexposure, 1/3 stop can be make make the difference between 'acceptable' and 'awful.'
Sure there can be bigger variables, but if you are not centering these errors around the most accurate EI you care capable of (lets say you don't bother to make a 1/3 stop adjustment that you should be making to make your average exposure more accurate), your overall error will always be 'other errors' + '1/3 stop' i.e. a third stop worse off.
If you consistently underexpose by 1/3 stop, for your methods, then this will bite you when you make another underexposure error on top of this for a given scene. When you consistently overexpose it causes lesser, but real negative effects overall, by increasing grain, causing you to use slower shutter speeds that reduce sharpness at the lower end, increasing darkroom exposures, causing DMax issue scanning etc. Not major, but real.
Not knowing how to determine an EI that is best for you and being bamboozled by the other variables is a different matter, as is just not being worried about exposing from the most accurate starting point as possible.
This is as much an academic point as anything. Nobody has to care, but there is a factual aspect to this discussion too!
1. 1/3 stop does not matter to me because I don't care about that level of accuracy or//
2. 1/3 stop just does not matter because of the other variables.
If its the first point, then I can't argue with that, but in the case of the second its just not correct.
1/3 of a stop makes 1/3 of a stop of difference. No more, no less and 1/3 of a stop can make a huge difference when you need it in a neg, but don't have it. Those who make serious prints will know how when you have perfectly exposed neg, you may have quite a bit of wiggle room, but when you are right on the edge of critical underexposure, 1/3 stop can be make make the difference between 'acceptable' and 'awful.'
Sure there can be bigger variables, but if you are not centering these errors around the most accurate EI you care capable of (lets say you don't bother to make a 1/3 stop adjustment that you should be making to make your average exposure more accurate), your overall error will always be 'other errors' + '1/3 stop' i.e. a third stop worse off.
If you consistently underexpose by 1/3 stop, for your methods, then this will bite you when you make another underexposure error on top of this for a given scene. When you consistently overexpose it causes lesser, but real negative effects overall, by increasing grain, causing you to use slower shutter speeds that reduce sharpness at the lower end, increasing darkroom exposures, causing DMax issue scanning etc. Not major, but real.
Not knowing how to determine an EI that is best for you and being bamboozled by the other variables is a different matter, as is just not being worried about exposing from the most accurate starting point as possible.
This is as much an academic point as anything. Nobody has to care, but there is a factual aspect to this discussion too!
Last edited:
Steve M.
Veteran
Ansel Adams has some wonderful info on this subject. Like Roger said, it depends on a lot of things. Subject, lighting, developer, etc. Having said that, I shoot Tri-X at 400 ISO and it works fine for me. In the end, if you want ideal negs (whatever that means) you'd be smart to do some testing.
Gabriel M.A.
My Red Dot Glows For You
I'm sure it can be conclusively demonstrated that "true ISO ration 1/3 stop" exists. Therefore, it is a strong candidate for believableness.
mdarnton
Well-known
In the end, as you say, NB23, what matters are the results.
For my preferences, the way I meter, the accuracy of my camera, the way I develop, and the kind of negatives I like to see, 250 works better for me than Tri-X's 400, so I use 250. It took me a while to arrive at that, and my opinion changed as the look of what I wanted changed. I actually got pushed to that conclusion, by David Vestal in a workshop--I didn't arrive at it on my own.
For all I know, someone else could be getting identical results setting their meter at 800, bu that doesn't matter.
For my preferences, the way I meter, the accuracy of my camera, the way I develop, and the kind of negatives I like to see, 250 works better for me than Tri-X's 400, so I use 250. It took me a while to arrive at that, and my opinion changed as the look of what I wanted changed. I actually got pushed to that conclusion, by David Vestal in a workshop--I didn't arrive at it on my own.
For all I know, someone else could be getting identical results setting their meter at 800, bu that doesn't matter.
BobYIL
Well-known
The longest gradations I have achieved with the HP5 (and eventually Tri-X too) is between 250-320, with development time cut accordingly. This was since the HP3 days..
Tom A
RFF Sponsor
I try to be consistent in areas of developing, always the same times, always the same agitation for a specific film and "look" that I want. It is one area that I can control easily.
As for exposure etc, I shoot about 60-70% of my black/white according to "sunny f16" and even though I am pretty good at guessing exposure - there are a lot of variables in doing so.
My 320 iso rating is based on the two main types of film I am using, Kodak XX and TriX/Arista 400 - the XX can take a slight underexposure with no problem and the TriX/Arista can handle the slight (very slight 1/3 stop) over exposure without any problem.
I do keep track of what cameras are used and if any of them have shutter speed variances. Mechanical speeds do vary - even from the factory you are looking at +/- 10 % variations in speeds.
Film is surprisingly flexible - exposure latitude is 1/2 to 2/3 f-stop without really affecting image quality (with 35 mm). I can't be bothered to be more precise - as what I get works for me. When I am shooting 120 I am more careful and in the days of 4x5 and 8x10 I spot metered with the best of them and processed with almost clinical precision.
As for exposure etc, I shoot about 60-70% of my black/white according to "sunny f16" and even though I am pretty good at guessing exposure - there are a lot of variables in doing so.
My 320 iso rating is based on the two main types of film I am using, Kodak XX and TriX/Arista 400 - the XX can take a slight underexposure with no problem and the TriX/Arista can handle the slight (very slight 1/3 stop) over exposure without any problem.
I do keep track of what cameras are used and if any of them have shutter speed variances. Mechanical speeds do vary - even from the factory you are looking at +/- 10 % variations in speeds.
Film is surprisingly flexible - exposure latitude is 1/2 to 2/3 f-stop without really affecting image quality (with 35 mm). I can't be bothered to be more precise - as what I get works for me. When I am shooting 120 I am more careful and in the days of 4x5 and 8x10 I spot metered with the best of them and processed with almost clinical precision.
Chriscrawfordphoto
Real Men Shoot Film.
I just exposed two rolls of Tri-X at 200. I'm developing them in PMK, which produces inadequate shadow detail at 400 or 320. The developer produces gorgeous images, which makes the speed loss worth it.
V-12
Well-known
Setting the meter for 320 ISO (if that is the true film speed) is not so much about the fanatical search for accuracy but not unnecessarily adding further inaccuracy to the aggregation of variables in film processing. If everything is kept within the bounds where one accurate choice (like true film speed) can happily cancel out another inaccurate choice (like a fraction over on the temperature) then the world keeps going around. But get to the stage when your choices are all wilfully slightly inaccurate and sooner or later it will go tits up. And you won't know what caused it because you lost your datum points a long time ago.
Photo_Smith
Well-known
I thought speed was a function of the range of tones you want to record. When they (Kodak) originally tested films shooting an average scene in the 1930's they shot a film at a given speed, developed in a standard MQ developer.
They then plotted the emergence point on a graph and marked a line 1.5 Dlog E to the right if that point .
If that point is on a linear portion of the resulting curve then the film has the correct rating (this was eventually adopted as the ASA rating).
In the 1960's they re-visited the data and found that few average scenes in the Northern hemisphere measured as wide as DLog 1.5 so the standard was revised to DLog 1.3 giving films a boost of a stop speed in 'average' shooting conditions.
The upshot is that the ASA/ISO ratings are just a guide based upon the original scene Kodak used (a bridge in bright sunlight with trees and shadow) your scene may look very different and that's why we adjust the exposure and development for fine tonal control.
So Tri-x is a 400ISO film based upon the criterion of the average scene (in a standard MQ developer) if your scene is different then your rating will be also.
So don't just down rate out of belief, think about the brightness range in the scene.
They then plotted the emergence point on a graph and marked a line 1.5 Dlog E to the right if that point .
If that point is on a linear portion of the resulting curve then the film has the correct rating (this was eventually adopted as the ASA rating).
In the 1960's they re-visited the data and found that few average scenes in the Northern hemisphere measured as wide as DLog 1.5 so the standard was revised to DLog 1.3 giving films a boost of a stop speed in 'average' shooting conditions.
The upshot is that the ASA/ISO ratings are just a guide based upon the original scene Kodak used (a bridge in bright sunlight with trees and shadow) your scene may look very different and that's why we adjust the exposure and development for fine tonal control.
So Tri-x is a 400ISO film based upon the criterion of the average scene (in a standard MQ developer) if your scene is different then your rating will be also.
So don't just down rate out of belief, think about the brightness range in the scene.
Roger Hicks
Veteran
I don't think you're quite right about speeds. The Kodak system was based on the 'first excellent print' (5x7 inch contacts). From http://www.rogerandfrances.com/subscription/ps expo neg.html
A wide range of people were asked to choose the 'first excellent print' from each series of 'typical' subjects. With the early prints in a series, each increase in negative exposure made for an obviously better print. Then, fairly suddenly, the quality would stabilize and then remain good for quite a while until it began to deteriorate again. The point at which the quality stabilized was taken as the 'first excellent print', corresponding to the minimum desirable exposure. This in turn formed the basis for Kodak film speeds, which later became first ASA and then ISO(free module).
The trouble is, the original 'first excellent print' research was done with 5x7 inch contact prints, uncoated lenses and an audience that did not consist entirely of keen photographers with a full awareness of the maximum possibilities of the photographic process: Kodak was quite understandably after the mass market. The result of this was that the 'first excellent print' may not always have been quite as excellent as the one made from the next negative that had received more exposure again. To this day, some photographers prefer to exposure their negatives a little more than others do: usually, anything from a third of a stop to a stop more.
They then found that the best way to describe what made an 'acceptable' print was a fractional gradient criterion. Unfortunately the fractional gradient was hard to measure, and when the ASA and ISO standards were pretty much consolidated they switched to a fixed density criterion:
From http://www.rogerandfrances.com/subscription/ps iso speeds.html
The current ISO standard could hardly be simpler. Film speeds are based on the exposure required to give a log density of 0.10 above film base plus fog (fb+f). This requires only a bare minimum of explanation, but if you are not familiar with log densities you would do well to look at the free module on the subject.
There's more about contrast later in the module. As for brightnesss ranges, look at http://www.rogerandfrances.com/subscription/ps neg development 1.html
Film or developer manufacturers' recommendations are based on average outdoor subjects, usually for the 128:1 brightness range that obtains around Rochester, New York. If you live further north -- and London is as far north as Edmonton, Alberta -- you may find that you are happier with development times that are longer than in upstate New York, while if you live further south -- California, Greece, southern Japan -- you may prefer shorter times. Indeed the Japanese contingent on the ISO standards committee has lobbied for a lower ISO standard contrast, which would of course mean reduced development times.
Cheers,
R.
RObert Budding
D'oh!
I always add a little extra exposure to color neg film to ensure that I capture shadow detail in high-contrast scenes. B&W film I base my exposures on film tests since I have complete control of the end-to-end process.
Photo_Smith
Well-known
Ah I took my information from the Theory book by James and Mees. They seem to suggest film speed is 0.10 + bf then make a point they call (A) then measure 1.3 DLog E units to the right and mark the point (B)
If your point B obtains 0.8 density units above (A) and is on the linear part of the curve then your film can handle the 128:1 SBR of the scene in Rochester park under bright conditions.
The speed being 0.8 divided by the point at where the density is above 0.10+bf? This is what it says in the Theory of Photography, my copy being 1980's so things may have changed...
The 1.5 DLog E from the earlier version came from Dr Jones essay on the speed of photographic films 1944.
I'm aware of the prints and the judgement by jury method, (and have seen the scene of the bridge and trees)
This method was instrumental in the choice of 1.5 DLog E which was later shortened in the 1960's.
Possibly all my information is out of date, but it would be strange that using a single data point of .10 above film fog being the only metric for speed, basically the manufacturing company can put any speed they wish on the box and call it ISO if that is the case!
Roger Hicks
Veteran
Well, that and contrast. As I noted, "There's more about contrast later in the module." I give it below, only with the gamma calculation already done instead of the 1.3-to-the-right
The reason fixed contrast is important is obvious to anyone who has ever 'pushed' a film, increasing its development (and therefore its contrast) in order to get more speed. The ISO standard specifies a contrast equivalent to a gamma of about 0.62: the higher the gamma, the higher the contrast. For present purposes it does not really matter very much what gamma is but again you may want to check the free module which explains it more. There is a great deal more about all of this in our book Perfect Exposure.
Many photographers find a gamma of 0.62 or so perfectly satisfactory, but many don't. A lot depends on subject matter, lighting conditions, lens contrast, and enlarger type. Some prefer a higher contrast, typically up to about 0.70. Rather more prefer a lower gamma, 0.56 or lower. Understandably, those who prefer the higher gamma get more speed than those who prefer the lower gamma.
Astonishingly, the original DIN was 0.10 above fb+f at gamma infinity
My Kodak Reference Handbook, November 1940, pp 206-207, refers to the 'new' film speed system with the fractional gradient (0,3 average gamma -- actually CI later) and, as you say, a log exposure range of 1.5; I don't know exactly when it was shortened to 1,3. My 1965 Focal Encyclopedia has 1,3, so it had clearly happened before then.
The magic of fractional gradient is that it is surprisingly independent of gamma, unlike fixed density points. My suspicion is that the change to 1,3 reduced still further the effect of development, and maybe got it closer to the DIN standard for more films, but that's substantially guesswork. Another possibility (quite convincing) is that when the fixed density criterion was adopted, 1.3 came out a lot closer to the existing best possible negative.
Cheers,
R.
Photo_Smith
Well-known
They changed it in 1963 because they re-visited the data of the original 1930's experiments taking out 'irregularities' the DLog 1.3 was considered to be a more realistic target. I have some of the information from the 1963 paper if you like I'll try to dig it out...
fotomeow
name under my name
Do you really believe in exposing at iso 320?
I dont think it makes much difference if you are not changing any other variables.
That said, I was taught in a B&W 2 or 3 film class that it does make a difference, insofar as you modify your OTHER variables in like.
And when we did the experiment, you could definitely SEE a difference.
We took pics of blank white walls, at different ISOs, then at different development times and concentrations for the same developer.
And also took pics of real life situations, with same method.
We compared different tonal ranges (Zone System) from the pics with different all those different variables.
Then we did density tests on the negs.
In the end, we could confidently say that for:
1. this particular film,
2. shot at this particular ISO
3. on this ONE camera (b/c shutter speeds DO vary)
4. in THIS particular developer,
5. at THIS particular time,
6. at THIS particular concentration,
Will give you maximum tonal range gradation and density.
We did the experiment. It made sense to me. I could SEE my own results, and those of my classmates.
The scientific method seemed to work.
Now, if you change any of the above variables, then you can easily get a different result. Thats why we see fotogs excel with one camera, one lens, one film, one developer, one paper. However, I am no longer that disciplined!
I dont think it makes much difference if you are not changing any other variables.
That said, I was taught in a B&W 2 or 3 film class that it does make a difference, insofar as you modify your OTHER variables in like.
And when we did the experiment, you could definitely SEE a difference.
We took pics of blank white walls, at different ISOs, then at different development times and concentrations for the same developer.
And also took pics of real life situations, with same method.
We compared different tonal ranges (Zone System) from the pics with different all those different variables.
Then we did density tests on the negs.
In the end, we could confidently say that for:
1. this particular film,
2. shot at this particular ISO
3. on this ONE camera (b/c shutter speeds DO vary)
4. in THIS particular developer,
5. at THIS particular time,
6. at THIS particular concentration,
Will give you maximum tonal range gradation and density.
We did the experiment. It made sense to me. I could SEE my own results, and those of my classmates.
The scientific method seemed to work.
Now, if you change any of the above variables, then you can easily get a different result. Thats why we see fotogs excel with one camera, one lens, one film, one developer, one paper. However, I am no longer that disciplined!
Roger Hicks
Veteran
Thanks for clarifying the date, but there's no need to dig out the paper: I think I've reached the limits of how excited I can get about this, at least for today. I can't help wondering, though, if the 'irregularities' were not, as I suggested, caused by the switch to the fixed density point which happened at around the same time (a little earlier).
Cheers,
R.
Roger Hicks
Veteran
Photo_Smith
Well-known
The paper(s) are B.S.I 1380 part 1 (1962) A.S.A PH 2.5 1060 (1963)
Prior to that point the best method was the fractional gradient method given by log E value corresponding to a point at the foot of the curve where the gradient is 0.3 of the average gradient over a range of 1.5 to the right of that point.
The new ASA ratings required all films to be developed to the same contrast and uses a fixed density above fog (0.10) and that the film be developed so the density is 0.9 at 1.3 log units to the right of fb+f
There is still controversy about the change as it seemed to inflate film speed without changing emulsions