It's frequently addressed on these forums because most people aren't well informed when it comes to flare. Sometimes things have to be repeated until they sink in.
Originally Posted by Leigh B
When we talk about flare, we aren't discussing excessive, sun streak kind of flare. We are discussing veiling flare and any optical system has some. Because the amount of flare varies depending on the scene's luminance range was well as the distribution of values within that range and because it is near impossible to measure flare, the goal is to eliminate it in film testing. The basic idea in any test is to eliminate as many unwanted variables as possible. The variables can later be incorporated back in when interpreting the test data.
By eliminating flare from the test, the film curve only represents how the material responds to exposure and development. As flare basically acts like flashing the film, including it in the test would make it difficult to determine the shape of the toe and the placement of the speed point.
Now it's just as important to remember to factor flare back into the data when applying the data for shooting purposes. This mistake probably causes more confusion about the photographic process than anything else I can think of. Average flare is approximately 1 to 1 1/3 stops for the average scene luminance range of 2.20. Normal processing for a seven stop scene is really for a 6 stop log-H range. Don't believe me? Look at Kodak's aim CI for Normal. There's no way to arrive at that number using their aim negative density range without factoring in flare.
B&W ISO speed point is one stop higher than where the shadow exposure should fall for an average scene. This is because speed is determine using a no flare test but needs to factor in real world conditions. Flare effectively doubles the film speed.
The concept of flare makes almost everything about photography make more sense. Obsession? No. Awareness.
Thanks for the explanation. That makes some sense. I'll have to think about it further.
Back to my original objection to using an enlarger and factoring in the flare comments...
Why not expose the film in a camera using a uniform source of illumination rigidly attached to the lens mount, without a lens?
This would get exposure times back in the range where they belong.
“Wise men talk because they have something to say; fools, because they have to say something.” - Plato
Leigh B, an enlarger is just a poor man's sensitometer. It's a light source with a timer. It's just not as accurate as a real sensitometer because it wasn't built to be one. Otherwise if certain steps are taken, I don't see anything wrong with using one. Well, someone else using one. When something is important to me, I don't like to compromise, so I bought a sensitometer to avoid all the frustration.
I don't see how your example of using a camera without a lens is any different from the basic enlarger set-up. Maybe I'm not following you.
Other than flare, lenses create additional problems if you are shooting a step tablet or grey scale. There's the question of shutter efficiency, which I've recently discussed. Also lenses are brighter in the center and have light fall off as you move further off axis. The variable "q" in the classic exposure equation is a product of various light loss factors in an optical system. One of the elements that makes up q is the angle of image point off axis. The equation uses a point 12 degrees off axis which has an average value of 0.916. In other words, while the middle grey of the tablet or grey scale are representative, the highlight and shadow values will be darker than they should be. This is not a problem with pictorial photography, but it is in testing.
Well, I spent the day looking at charts and have come to the conclusion that the EG&G is a pretty stable sensitometer.
Although it came with a calibration sheet that says its light output is 765 mcs, I estimate its output is 178 mcs.
Out of a dozen TMY-2 tests, five were close enough to be worth pretending they might represent a sample that achieved full film speed. Working under that assumption, I calculated what the light output would have been. The estimated light output averaged 172 mcs with standard deviation 24.
I don't think it varies pop-to-pop by 24 mcs - I think it's dead on every time. But I believe its actual light output is between 154 and 202 mcs and I'm pretty sure it's close to 178. I give the greatest weight to the TMY-2 that hit the ASA triangle exactly.
This brings me back to Panatomic-X and my story about why I get higher than rated speed.
I think I know what the deal is. Time has not degraded its toe sensitivity. It has not given the film any fog. Time has reduced the contrast!
When I developed for 7 minutes which is what the sheet says to do... I get CI 0.48 and speed 32.
The time-CI chart shows 8 minutes for 0.6 CI and 10 minutes for 0.7 CI.
When I developed for 9 minutes I got CI 0.51 and speed 40... The last test where I developed for 10 minutes, I got 0.62 CI and hit the ASA triangle. The speed is right between 50 and 64.
So instead of 7 minutes developing, I am developing for 10 minutes. This restores the contrast. But the additional development has the unintended consequence of giving me between 2/3 to 1 stop additional speed.
I just cannot believe that the value in mcs that you fund differs so much from the stated value that EG&G gives. That is a huge difference for such a reputable instrument.
Sponsored Ad. (Subscribers to APUG have the option to remove this ad.)
That calibration was for my sensitometer! I sold you my spare. The calibration sheet was supposed to give you an idea of the difference that could be found between the values in the manual and those from the individual unit. I've had two different units calibrated. Mine has 765 mcs at 10^-2 and the other had 681 mcs at the same setting. But at 10^-4 one had 60mcs and the other 59.4mcs. Almost identical.
Originally Posted by Bill Burk
That was years ago. God knows where mine is now.
As for consistency, when I was working in the lab business, I would run a sensitometric strip along with a control strip. We were running tons of film at the time and the control plots were solid as was the sensitometric strip.
Stephen gave part of the answer, well 681 is what it should have been.
I just assumed since it has been over 20 years since calibration a capacitor lost some capacity.
The next speed up (1/1000th second) says 3278 mcs and it is interesting that it hits really bright!
I puzzled over why they chose to make the faster speed so bright. But I found a practical way to take advantage of it. I do the film tests at 1/100th second because the brightness is suitable for film (with a 1.84 ND filter, just gelatin, in the well).
While the next speed is bright enough to be suitable for making paper test strips.
Interesting. Stephen sent it to me with a 1.84 ND filter in the well, which makes the 400 speed film speed point graph very nicely. It falls off just a little below the last step. Somebody knew it wasn't 681 mcs - otherwise they would have had a different filter in the well.
Like PE, I can't hardly believe the flash power would fall to 1/4, unless a severe fault existed. I'm not an electronics guy, but I spent years working with guys who repaired studio flashes. Typically, a failure to that extent meant that 3 out of 4 caps had failed, or something along those lines.
I think my next step would be to expose at both the 681 and 3278 mcs conditions (reciprocity failure should be minimal in the associated time durations) to see if that nominal 5X difference still exists. Or if it's now ~20X, per your guesstimate of ~178 mcs.
If the original ~5X ratio still exists, I'd question your evaluation of actual power. Previously, I said that I'd use a color negative film for this, not B&W. Mainly, because I have a lot of experience testing color neg film; I know that Kodak pro films, at least the 160 speed, have been extraordinarily consistent and stable. Also, the development stage can be confirmed with process control strips.
B&W might be as good for testing - I just don't know for sure. Did I mention I have a lot of past experience testing Kodak pro color neg films?
We know I got the wrong calibration sheet.
It has been stable at ~178 mcs for two years. Maybe it was customized.
This is what Bill is referring to, the odd progression of light outputs of the 3 flash speeds (1/100, 1/1000 and 1/10,000) in meter-candle-seconds:
Originally Posted by Bill Burk
10-2 light output = 800
10-3 light output = 4000
10-4 light output = 70
10-2 and 10-3 circuits are each powered by the same type of capacitor. 250uf/475v charged to the same voltage and connected to a single tube with a fixed resistance provided by the gas.
The 10-2 circuit (slower/dimmer) dumps its 250uf charge through a larger value resistor on the way to the tube. This prolongs the duration of the flash based on the formula:
Flash Duration = 1/2 * R * C
R is the resistance of the circuit
C is the total charge on the capacitor (not current)
The resistor, however, also decreases the voltage on the tube:
Voltage lost to resistor during discharge (not avaiable to the tube) = I * R
I = current
R = Resistance
Decreased voltage on the tube makes a flash dimmer as illustrated below. That is why 10-2 is dimmer than 10-3.
So, what about the 10-4 circuit (the fastest one). Shouldn't that one be even brighter than 10-3? It turns out that circuit uses a smaller capacitor (10uf) and dumps its charge directly to the tube with no intervening resistor. Since there is less resistance, and less charge the flash duration is the shortest. However, there is another relationsip: the smaller the capacitor the dimmer the light. I don't have an equation for that one, just the empirical relationship illustrated below:
(Graphs are generic, not EG&G data).
Last edited by ic-racer; 07-15-2012 at 04:38 PM. Click to view previous post history.