matti
Established
After some escapades into new film/developer territory (Ilford fp4+, efke and Diafine, Calbe R09 and D76) lately I am start to understand that I am BAD at two things:
1. SEEING on the negatives or on the scanned image what went wrong.
2. Exposing the film. If I can not do 1, I can not learn this.
(I was bad at doing repeatable things with the chemicals when developing due to using the oven thermometer and not being prudent ennough with times and such. But my spouse, soon to be phd in chemistry, helped me a bit with that.)
So, now I decided to setup some tests to learn how to see what is overexposed, underdeveloped and things like that. I mean, I got some good pointers here for fixing things but messed up when I tried to follow them.
How should I set up the test best? I went out and bought this book Creative Black and White Photography by Suess that describes one way of exposing a series of five test shots with recommended exposure and over and under exposure and then develop each series differently and see what is best.
Sounds like a good place to start but:
1. I am not so confident that I can look at the negatives and locate the best one.
2. I havn't found a way to make Vuescan behave in a sort of uncalibrated way, that can be the same for each scan now and with every new film I want to test.
3. How do I locate a good consistent subject? Isn't it the whole point that different subjects need different exposure/development for different contrast? How do I choose a subject that is not too contrasty or flat?
The goal is to learn to SEE what is wrong with the negatives and scans and be able to fine tune things even more.
/matti
1. SEEING on the negatives or on the scanned image what went wrong.
2. Exposing the film. If I can not do 1, I can not learn this.
(I was bad at doing repeatable things with the chemicals when developing due to using the oven thermometer and not being prudent ennough with times and such. But my spouse, soon to be phd in chemistry, helped me a bit with that.)
So, now I decided to setup some tests to learn how to see what is overexposed, underdeveloped and things like that. I mean, I got some good pointers here for fixing things but messed up when I tried to follow them.
How should I set up the test best? I went out and bought this book Creative Black and White Photography by Suess that describes one way of exposing a series of five test shots with recommended exposure and over and under exposure and then develop each series differently and see what is best.
Sounds like a good place to start but:
1. I am not so confident that I can look at the negatives and locate the best one.
2. I havn't found a way to make Vuescan behave in a sort of uncalibrated way, that can be the same for each scan now and with every new film I want to test.
3. How do I locate a good consistent subject? Isn't it the whole point that different subjects need different exposure/development for different contrast? How do I choose a subject that is not too contrasty or flat?
The goal is to learn to SEE what is wrong with the negatives and scans and be able to fine tune things even more.
/matti
lushd
Donald
matti said:
Rush to your nearest branch of Amazon.com and get a copy of The Negative by Ansel
Adams and do everything he tells you. Then you will see, grasshopper.
kaiyen
local man of mystery
Hey Matti,
Here are some of my methods. First off, it sounds like you'd be willing to set up a test scene that you can use for these experiments, rather than using "real-world" subjects. Is that right? In that case things will be much easier...
What I do when I need a test scene is to use some towels. A brown or dark blue towel is just about 2 stops darker than middle grey, and a light beige or white towel is about 2 stops brighter. Set up a scene with those, plus a grey card in the middle. Try to find some controlled lighting (natural, though) and meter them to make sure they are right.
Now, shoot a whole roll of that scene but varying your film speed. So if you're shooting Tri-X, start at like 100 and go up by 1/3 stops (or 1/2 if that's all you can do) until...let's say 800 (unless you are using Diafine, too - in that case, go to 1600). If you do it right, you should be able to shoot 2 entire sets of those exposures on a single roll. That way, you can clip the roll in half and use 2 different developers.
Now, soup the negatives in whatever developer you want, with controlled, repeatable methods. So standardize on how many inversions, at what speed, at what frequency. Get your water temperature right (doesn't have to be perfect, but gotta be close), etc.
Next, put those negatives on a light box or hold them up to the light. Look at the area where the dark towel is. Even with the naked eye, you should be able to zero in on which ones do and do not have enough shadow detail. If it is _absolutely_ clear, then it's underexposed. If it looks like there is a whole lot of stuff there, it's probably over exposed. You will like have several frames that look "pretty good." That's fine.
If you want, take a 50mm lens, open it up all the way and put it front-side-up on top of the negative, over a light box. It's a make-shift loupe. Look even closer at the shadow areas to see which ones are good candidates.
Now, on to your scanner. The problem with doing these tests on scanners is that it's almost impossible to get a good baseline. Almost everything is relative. But, you can get close. What I do is:
1 - lock exposure (1,1,1)
2 - lock the film base (look at Vuescan help, under Advanced Workflow)
3 - under color, select Kodak TMX as film and D76 CI = .55 as the profile.
This is a nice, natural looking curve that should work for most situations. You could standardize on a different one but the key is to stick with one.
Now, do a scan of the first candidate. Does the shadow detail look good to you? Scan the others. Which one looks "best?"
The truth is that several might look good. For instance, when I tested FP4 in Rodinal I felt that 64 and 80 both looked pretty darn close. I preferred the speed, so went with 80.
Now, for development. Let's say you picked the one for EI 250 for Tri-X. Look again at that scan. Look at the historgram, too. Is the right edge off the histogram? If so, what I do is to then reduce development. Shoot an entire roll at 250 and cut it into 2 or even 3 sections, then develop for 10%, 20%, and 30% less than before. Does one of them bring that right side back in? Try to get it so that it's just back in - don't compress it too much. When you find that one, you now have your "normal" contrast development time.
Similarly, if the right edge is really far awy from the side, you might increase development a bit. I prefer to calibrate such that the histogram covers as much of the range as possible - it maximizes the tones available in a film's dynamic range.
Instead of changing dev time you could change the film profile and CI setting in Vuescan. Try CI = .7 or .8 and you'll see the right edge move over. However, you've also changed the mid tones and perhaps even pushed your shadows off the left edge. I use those only if I happen to have a poorly-developed negative. I prefer to handle things in development if I can.
So...that's what I do. I hope that helps...
allan
Here are some of my methods. First off, it sounds like you'd be willing to set up a test scene that you can use for these experiments, rather than using "real-world" subjects. Is that right? In that case things will be much easier...
What I do when I need a test scene is to use some towels. A brown or dark blue towel is just about 2 stops darker than middle grey, and a light beige or white towel is about 2 stops brighter. Set up a scene with those, plus a grey card in the middle. Try to find some controlled lighting (natural, though) and meter them to make sure they are right.
Now, shoot a whole roll of that scene but varying your film speed. So if you're shooting Tri-X, start at like 100 and go up by 1/3 stops (or 1/2 if that's all you can do) until...let's say 800 (unless you are using Diafine, too - in that case, go to 1600). If you do it right, you should be able to shoot 2 entire sets of those exposures on a single roll. That way, you can clip the roll in half and use 2 different developers.
Now, soup the negatives in whatever developer you want, with controlled, repeatable methods. So standardize on how many inversions, at what speed, at what frequency. Get your water temperature right (doesn't have to be perfect, but gotta be close), etc.
Next, put those negatives on a light box or hold them up to the light. Look at the area where the dark towel is. Even with the naked eye, you should be able to zero in on which ones do and do not have enough shadow detail. If it is _absolutely_ clear, then it's underexposed. If it looks like there is a whole lot of stuff there, it's probably over exposed. You will like have several frames that look "pretty good." That's fine.
If you want, take a 50mm lens, open it up all the way and put it front-side-up on top of the negative, over a light box. It's a make-shift loupe. Look even closer at the shadow areas to see which ones are good candidates.
Now, on to your scanner. The problem with doing these tests on scanners is that it's almost impossible to get a good baseline. Almost everything is relative. But, you can get close. What I do is:
1 - lock exposure (1,1,1)
2 - lock the film base (look at Vuescan help, under Advanced Workflow)
3 - under color, select Kodak TMX as film and D76 CI = .55 as the profile.
This is a nice, natural looking curve that should work for most situations. You could standardize on a different one but the key is to stick with one.
Now, do a scan of the first candidate. Does the shadow detail look good to you? Scan the others. Which one looks "best?"
The truth is that several might look good. For instance, when I tested FP4 in Rodinal I felt that 64 and 80 both looked pretty darn close. I preferred the speed, so went with 80.
Now, for development. Let's say you picked the one for EI 250 for Tri-X. Look again at that scan. Look at the historgram, too. Is the right edge off the histogram? If so, what I do is to then reduce development. Shoot an entire roll at 250 and cut it into 2 or even 3 sections, then develop for 10%, 20%, and 30% less than before. Does one of them bring that right side back in? Try to get it so that it's just back in - don't compress it too much. When you find that one, you now have your "normal" contrast development time.
Similarly, if the right edge is really far awy from the side, you might increase development a bit. I prefer to calibrate such that the histogram covers as much of the range as possible - it maximizes the tones available in a film's dynamic range.
Instead of changing dev time you could change the film profile and CI setting in Vuescan. Try CI = .7 or .8 and you'll see the right edge move over. However, you've also changed the mid tones and perhaps even pushed your shadows off the left edge. I use those only if I happen to have a poorly-developed negative. I prefer to handle things in development if I can.
So...that's what I do. I hope that helps...
allan
matti said:
matti
Established
Allan, thanks. I will try your way with first setting the dark point and then developement time. About the natural lightning, if I would shoot this in shadow outside, would I get repeatable results? Even when the weather or season is different, if I measure the light on the grey card? Or would the different lightning conditions change the contrast of the scene too much?
/Matti
/Matti
T
tedwhite
Guest
I can't help you here, as I've always used a wet darkroom. If a negative was under or overexposed I compensated by reducing or increasing exposure times and/or changing filter grades if using polycontrast paper - or, if using graded papers, selecting the grade based on negative density and contrast. Plus, of course, dodging and burning.
I have a neg scanner, and Vuescan, but my learning curve in this arena has just begun.
Sorry.
Ted
I have a neg scanner, and Vuescan, but my learning curve in this arena has just begun.
Sorry.
Ted
kaiyen
local man of mystery
Matti,
As long as the dark towel (in my example) is metered to be 2 stops darker than middle grey, and the light towel 2 stops brighter, it's fine. However, you may find that it'll be hard to get a light towel that bright in the shade.
Maybe a part of your house with a window that gets nice even light for a bit of time?
If you get a scene set up, just shoot like 5 rolls all at once
Good luck. Keep us updated.
allan
As long as the dark towel (in my example) is metered to be 2 stops darker than middle grey, and the light towel 2 stops brighter, it's fine. However, you may find that it'll be hard to get a light towel that bright in the shade.
Maybe a part of your house with a window that gets nice even light for a bit of time?
If you get a scene set up, just shoot like 5 rolls all at once
Good luck. Keep us updated.
allan
matti
Established
Hi Allan,
thanks again. You are right, I should test several different films at once and maybe like two of the same sort, so I can develop in four different ways if I want to, if I cut them in half. (Someone at home will think I am crazy for sure...)
Donald, I actually have that book, and even read it. But I find it hard to use that mathod. It seems like it is directed more towards Large format and as I remember it you don't shoot a "scene" but one single grey shade at the time. Then I don't see how I can learn how to look at the negatives and neutral scans and actually predict how my exposure was? Maybe I don't remember correctly, as I don't have the book in front of me. But of course I do respect Adams workflow, even if the pictures are a bit... (shhh don't tell anyone I said this) ... boring.
/matti
thanks again. You are right, I should test several different films at once and maybe like two of the same sort, so I can develop in four different ways if I want to, if I cut them in half. (Someone at home will think I am crazy for sure...)
Donald, I actually have that book, and even read it. But I find it hard to use that mathod. It seems like it is directed more towards Large format and as I remember it you don't shoot a "scene" but one single grey shade at the time. Then I don't see how I can learn how to look at the negatives and neutral scans and actually predict how my exposure was? Maybe I don't remember correctly, as I don't have the book in front of me. But of course I do respect Adams workflow, even if the pictures are a bit... (shhh don't tell anyone I said this) ... boring.
/matti