PDA

View Full Version : Kodak Characteristic Curves expalined ?



percepts
27-Nov-2009, 10:49
I have never understood the notation for Kodak characteristic curves. Why do they use Lux Seconds and what is the significance of zero lux seconds on the x axis of all their film curves? And what is the significance of -3.0 lux seconds?

p.s. I know what a characteristic is but why do kodak seem to use something different than we all do when working out sensitometry?

gteeter
27-Nov-2009, 11:04
I have never understood the notation for Kodak characteristic curves. Why do they use Lux Seconds and what is the significance of zero lux seconds on the x axis of all their film curves? And what is the significance of -3.0 lux seconds?

p.s. I know what a characteristic is but why do kodak seem to use something different than we all do when working out sensitometry?

I think Kodak and other manufacturers prefer to work in absolute units. I presume that manufacturer's labs use calibrated light sources to make this possible. For an exposure the standard is lux seconds. Lux is the intensity, and multiplying that by exposure time gives the exposure--the time-integrated flux of photons onto the film. There must also be standards relating to how spectral respone is folded into the analysis.

Photographers, on the other hand, tend to use calibrated or uncalibrated light meters that give relative light intensities, I guess because it's more convenient and it allows the light meter to become a part of the testing procedure.

For a particular light meter it should be possible to directly convert reading into lux seconds, as long as the manufacturer is willing to share this information.

percepts
27-Nov-2009, 12:41
but there is a point on their curves where the log lux seconds is zero which I assume means 1 lux second. But then there is also log -3.0 lux seconds which must mean 0.001 lux seconds (1/1000th of a second). Is that correct?

gteeter
27-Nov-2009, 12:45
but there is a point on their curves where the log lux seconds is zero which I assume means 1 lux second. But then there is also log -3.0 lux seconds which must mean 0.001 lux seconds (1/1000th of a second). Is that correct?

That sounds right to me. Of course, to get an exposure of 0.001 lux seconds doesn't necessarily require an exp. time of 0.001s--that would only be the case if the light intensity were 1 lux.

ic-racer
27-Nov-2009, 13:10
Yes, decimals will show up as negative numbers after the log.

Here is a quote from Kodak on the subject:

Now we have to get the log exposure for each
step so that we can graph the characteristic curve for our
film sample. We can measure the amount of light that the
sensitometer is putting out by using a lux meter (a light
meter that measures lux. Lux and metre-candles are
the same thing, but lux is the preferred term). So we try
this and we get a reading of 100,000 millilux. The
instruction manual for the sensitometer tells us that the
exposure time is 1/5 of a second.
Exposure = Illuminance (in millilux) x Time (in sec)
Exposure = 100,000 x 1/5
Exposure = 20,000 millilux-seconds
Now we take the log of our exposure, 20,000
millilux-seconds, and get an answer of 4.3. Our log
exposure is 4.3.

percepts
27-Nov-2009, 13:10
The final question I think is: what fraction of a lux second corresponds to a 1 stop difference in exposure? I guess each doubling of time from any point is a 1 stop difference as normal.

ic-racer
27-Nov-2009, 13:11
The final question I think is: what fraction of a lux second corresponds to a 1 stop difference in exposure?

Well 1/2 or double, depending on if you are going up or down. That is why they convert it to logs, because in that case it is easy, just 0.3 when going up or down.

percepts
27-Nov-2009, 13:14
0.3 on the log scale.

That figures. Now I understand their charts. Thanks all...

ic-racer
27-Nov-2009, 13:20
This is a good document. Everyone interested in film curves should read it. It really explains it in an understandable manner.
http://motion.kodak.com/motion/uploadedFiles/US_plugins_acrobat_en_motion_education_sensitometry_workbook.pdf

Nathan Potter
27-Nov-2009, 17:57
Lux is a measure of light intensity as measured with a spot meter. You can find conversion tables here and there. but for reference here are some common equivalents between lux and EV.

EV lux
-1.0 1.25
0.0 2.5
1.0 5.0
5.0 80
10 2600
15 81,900
18 656,000

Exposure dose is lux times seconds. As mentioned lux-sec. is often given in log units for ease of use.
More strictly lux is a measure of illuminance and is a lumen per meter squared. But these photometric units can drive one crazy so best to not get into too much detail except as necessary.

Nate Potter, Austin TX.