Bill, my only point (perhaps not made all that clear) was disconnection of lab environment in meter testing vs. real world use. Which ultimately led me to the conclusion that keeping the meter readings consistent is MORE important to good exposure than having it perfect on some unrealistic test target. In other words my meter may well be a stop off against yours, but right on for my total procedure.
Witold
simplest solutions are usually the most difficult ...
Marc, last try, not sure if your not understanding what I was trying to accomplish, or just defending your earlier post.
I am NOT trying to get two meters to read identical. I am trying to have two meters that read accurate lighting levels under all color temperture light. Having two Sekonics alone, will not accomplish this task. They will give me two identical wrong readings under all color temp light except 3400K light. If one goes through all the trouble of doing the tests described above, you can easily make two different "corrections" charts, one for each meter, regardless if they are the same make. Although having two identical meters has other benefits, so I am not ruling that out, but its certainly not the solution to the issue which started this thread.
Witold, your point is well taken. In spot meter readings, your point is VERY valid, as it's not just the color temp of the light source, it also considers the color of the subject (target). The number variables are just too great to control, as a previous poster also mentioned. However, in incident mode, which only measures the light source, if I can deliver varying color temperture light to the meter during the testing phase, then use the same color temp meter in the field, I would think these should correlate well and produce more accurate light level readings. But until I test all this, I just don't know myself how error prone all this is.
In a different thread, I read about the processing E6, and a comment about adding 16% ? of time to the process when using Fuji film. This seems to suggest Fuji film processes to a different standard then other e6 films? I am curious how this can effect the outcome of processing. It's the first time I ever read of E6 films treated differently based on the maker? Any darkroom gurus care to comment?
Bill, you need one of these:
http://www.oceanoptics.com/products/usb2000uvvis.asp
You'll have to also buy a short optical fiber and one of their cosine illuminance probes for sampling your ground glass. With this, and a handlheld PC, you can take a spectrum at each point on your ground glass.
Writing the lookup tables to determine the photographic exposure index for individual films is left as an exercise for the reader.
If you're feeling even lazier, and richer, there are devices called 'imaging spectrometers' which combine a line scan image with spectral resolution. Rotating the instrument, or taking successive datasets as your satellite moves relative to the ground, gives you an image with an associated optical spectrum at each pixel.
I've chosen to stick with one film and learn it's characteristics :-)
Struan, interesting devise, but surely a bit extreme for what I am trying to accomplish. I am simply trying to see how far off the light meter reading is under varying color temp light sources. I felt film was the best method of this, but involved, indeed.
For a simple cross check, I can put gels over a light source and then subtract the gel value of the pre gel light reading... make sense? Any other quick cross check ideas come to mind?
My point was that if you don't want to make any assumptions about the colour temperature of your light or the spectral reflectivity of your object, there are tools that do the job, and at some point it's not worth fussing about with photographic equipment: you might as well go to the vendors who have solved exactly the problem you have posed.
I don't think quick and dirty tests of the meter have much validity, unless you have a very well-controlled light source. If you know you are dealing with fairly continuous-spectrum lights then correlating meter readings with a colour temperature meter could prove useful, but then you'll need to take the colour temperature meter with you (note: this is what most architectural photographers do). With line or narrow-spectrum emitters (and they're getting more and more common as incandescent lamps are made obselete) a bracketed test shot probably is the best practical and scientific solution.
I see a use for a calibrated digicam, with good temperature stability and the ability to read out the exposure values post-capture. Perhaps the next Picker will be selling modified Cybershots.
Struan, yes, I always have the color meter with me anyway....
What do you mean by a well controlled light source? What is the problem with putting gels over any light source and reading the temp of the light, then doin the cross check as I described? I was not aware that light sources, even incnadescent was not continuous spectrum?
Incandescent bulbs are fairly good black-body radiators at their rated voltage and current. If you start, say, changing the current to get a lower or higher colour temperature, the brightness also changes so you're back to worrying about your meter's linearity before you can measure colour response. If you pulse-modulate the current, with a dimmer switch, you'll get a time-averaged mix of colour temperatures.
Basically, if you lack a good calibrated light meter of some sort, you end up trying to pull yourself up by your bootstraps.
Most of the households I've visited in recent years have a mix of incandescent, compact fluorescent, fluorescent and even LED lighting. All except the good old Edison light bulb have spikey spectra that can in principle cause gotchas with some films and not with others. A colour meter makes too many assumptions when converting the spikey spectrum to a simple colour temperature plus mired reading. My standard colour film is Portra NC, because I love what it does in daylight, but I sometimes get oddly saturated casts near unknown lighting fittings.
This is why optical workshops have an 'artificial sun', a model black body radiator that can be reliably tuned in both colour temperature and brightness. I only know of the optics research grade gear, which is pricey but has a calibration trail back to a National Standards body like the NIST in the US. I don't know how the equipment in a typical camera repair shop compares.
Bill, you're speaking about photography "under all color temps" in the field etc. Experience is your best bet - after all, the temps are not so impredictable. You can learn with your MF format to know how the film reacts at for ex. the morning or the evening light. It's not a rocket science, fortunately. All the fiddle with corrections charts will just lead you away from the intuitive way you can learn from frequent shooting. I shoot commercially and I have to say that measuring the exposure is for me so automatic that sometimes I don't even pay attention to it at all. It's very rare I have the exposure wrong (and then mostly because the situation was impossible anyway - I shoot chromes). Just my humble opinion.
Struan... my testing method is much more simplfied......
Take single non fluorescent bulb that delivers 5500K light, set x distance from incident meter diffuser dome. Take light reading, put color gels in front of light source, take color temp reading, then light reading. In theory, the light meter should read the last reading, less the light loss of the gel (which the maker gives us). Then prepare a carrections sheet. This was the rough test I was referring to? What do you think?
Now do the same at different distances to vary the light intensity hitting the meter? Have a gel set up to maybe test 3200 4500 5500 6500 7500 9500 light. Then prepare a corrections sheet. Then based on this, I could always integreate a film test, but I was considering this first?
I could also set this test up in a box on a sunny day with no cloud cover... I can continously monitor the sunlight, but in the amount of time it would take to complete this test it shouldn't vary much? This light source would more easily duplicate real world situations, since I shoot with landscapes. Your thoughts?
On my way to bed.....
How do you make 5500 K light into all those different colour temperatures in a way that is accurate to, say, 1/6 of a stop? Trust the gel manufacturer's filter factor? But then you have to know which spectral sensitivity they assumed when they came up with the filter factor.
I'm not trying to be argumentative. Just pointing out that you can very easily end up going round and round in circles. Either you accept that good enough is good enough (in which case there is a strong argument for test shots on film and a $1 notebook), or you invest in a precision instrument with a known and stable calibration.
Some measurements are inherently stable with time. A good mechanical balance will still be a good mechanical balance after ten years in a cupboard. Sadly, even the best optical sensors need regular calibration: for lab instruments there is a whole industry dedicated to periodic recalibration, with an audit trail back to rigorous standards.
Posting Permissions
Reply With Quote
Bookmarks