Light meter spectral sensitivity and sensitometry

[BrianShaw]

,,.

Brian - this has nothing to do with meter to meter variation in terms of manufacturer quality control. I've had a number of unmodified Pentax spot meters which identically matched over their full range, and even precisely matched the response of a Minolta Spotmeter F. But the Zone VI modification was exactly that - an internal modification to render a different kind of spectral response. Some people liked it, some didn't. Those can still be serviced by Richard Ritter; but otherwise, they are an evolutionary dead end.

Right. Post 17, which I was responding to, didn't seem to even imply sample variance or QC variance; It was comparing meter brands. Compariong meter brands is probably too generic for a meaningful discussion.

 

[chriscrawfordphoto]

I actually did an extensive series of tests of several spotmeters' spectral sensitivities a few years ago. There were pretty significant differences, especially under incandescent light sources, which have a lot of IR in them.

Spotmeter Color Sensitivity

crawfordphotoschool.com

 

[BrianShaw]

I remember that now, Chris. Great report. For me, though, I stopped reading after this:

“The Results of the Color Sensitivity Test
If you look at the numbers on the chart, you'll see that there really isn't a whole lot of difference between the meters when reading different colors under daylight-balanced lights. On many of the colors, all of the meters read the same, or very nearly so. On most of the colors, the four meters are all within 1⁄3 of a stop.”

I’ll have to reread, especially about incandescent even though that’s not much of an issue for me.

 

[DREW WILEY]

Chris - In terms of methodology, I seriously question the use of that so-so LED panel for any kind of serious color value comparison.
And did you even check it's true K temp against the alleged setting or rating using a good color temp meter? Those panels can be way off at times, and rarely reach true 5500K daylight output. Generally, they don't even take into account the yellowing effect of the overlying plastic diffuser (same problem as with light boxes except for the most expensive).

But since any skewing of results seems to have affected all the meters being tested in a similar manner, that probably didn't matter so much in this particular test case. It might with respect to critical color photographic applications, if people take the trouble. The MacBeth Color Checker Chart is a wonderful tool; but colors in nature often have a mind of their own.

Per flare, one of the Zone VI modifications was to apply flat black paint internally to the housing, to reduce flare reaching the sensor. I just use a threaded on rubber lens shade, just like on a camera lens, and perhaps hand shade it too, when taking a reading. That should be routine, common sense procedure for everyone.

 

[ic-racer]

I actually did an extensive series of tests of several spotmeters' spectral sensitivities a few years ago. There were pretty significant differences, especially under incandescent light sources, which have a lot of IR in them.

Spotmeter Color Sensitivity

crawfordphotoschool.com

Very nice!

 

[tanner]

So, I opened my L-308DC and the diode looks very much like this one.

https://www.hamamatsu.com/content/d...9_SALES_LIBRARY/ssd/s16839-01ms_kspd1093e.pdf

There are no markings, but the package is the same shape and the filter looks to be the same, judging by the color. There are other filtered diodes in this manufacturer's product range, but this is the only one that's magenta, so I reckon the spectral response is "analogous to CIE spectral luminous efficiency", as it says in the document, even if it's not actually a Hamamatsu part.

https://en.wikipedia.org/wiki/Luminous_efficiency_function

In the case of this meter the spectral response is meant to be very close to that of the human eye, as far as brightness is concerned. I still don't understand why that needs to be the case, since light meters are supposed to meter light for an emulsion or a digital sensor and the interpretation of the image is done by the human at a later stage. Most spectral response charts for panchromatic BW film that I've seen are pretty flat across the visible spectrum, although I did see one for a digital sensor and it might have been more pointy at green. I'll have to investigate further. Exposure vs density curves also have lux.s on the exposure axis, which is basically spectral power density weighted by the luminous efficiency function. Why?

With film one can only use filters to change the response: blue filter with white flash bulbs for daylight film and orange filters for tungsten film in daylight.
Now with digital cameras have the capability to set to White Balance to: Sunny Bright, Daylight, Cloudy, ... and alter the response in post processing well beyond the adjustments with film in the past. Therefore the manufacturers probably do not feel a need to supply the sensors' response curves.

None of these change the spectral response of the film or sensor. It's a physical property of the recording media.

 

[DREW WILEY]

Why ideally the same spectral response as the human eye? For sake of color film and other color recording devices, which, with the exception of specialty color films like IR, are intended to yield results look "natural" to humans. For decades, the standard for reproduction work was based on positive chrome images selected atop a light box, evaluated by natural vision. Color negative film had to be inverted through printing to do that, a whole extra step. Scanners weren't around. Today, we still intuitively see things the same way; the human eye itself the most crucial instrument of all.

Black and white films differ somewhat from one another. Even panchromatic films have their own distinction. And certainly NONE of them are anywhere near "flat" across their own spectral sensitivities. That's why you need to understand the specific filter factors for each type black and white film you might be using. Yes, one particular meter line was modified for sake of alleged pan film sensitivity, rather than color film, which was the Zone VI modification of the Pentax; but it was optimized for Tri-X, before other kinds of sheet films like TMax with a somewhat different spectral response became dominant.

 

[BrianShaw]

That has all been written before, Drew and I thank you for writing it again so clearly!

 

[ced]

One of Sekonic's latest does it all...

 

[Alan Edward Klein]

Chris - In terms of methodology, I seriously question the use of that so-so LED panel for any kind of serious color value comparison.
And did you even check it's true K temp against the alleged setting or rating using a good color temp meter? Those panels can be way off at times, and rarely reach true 5500K daylight output. Generally, they don't even take into account the yellowing effect of the overlying plastic diffuser (same problem as with light boxes except for the most expensive).

But since any skewing of results seems to have affected all the meters being tested in a similar manner, that probably didn't matter so much in this particular test case. It might with respect to critical color photographic applications, if people take the trouble. The MacBeth Color Checker Chart is a wonderful tool; but colors in nature often have a mind of their own.

Per flare, one of the Zone VI modifications was to apply flat black paint internally to the housing, to reduce flare reaching the sensor. I just use a threaded on rubber lens shade, just like on a camera lens, and perhaps hand shade it too, when taking a reading. That should be routine, common sense procedure for everyone.

Drew, I used a Spectraview II (NEC) light calibrator puck to check my Kaiser LED panel. Here are the results.

 

[DREW WILEY]

What is the listed rating of that Kaiser unit, Alan - 5000? Kaiser is a pretty reputable brand. But the actual CRI of the LED version isn't going to be as high as one with fluorescent color matching tubes in them, which cost more.

 

[Bill Burk]

K in this context is Kelvin color temperature. I’m using a light source that is basically a Beseler 23 enlarger bulb with power resisters and rheostats. B. D. 92, near the top of its range, creates a 2700 K source.

100 foot lamberts, by my reckoning (correct me if I am wrong), should be 32 candles per square foot (the scale on several Weston Master meters. For meters with an arbitrary scale, the “exposure formula” would suggest exposure 1/30 second at f/10 for ASA 100. I can get some Weston Master meters to read close to 32 when my Sekonic L758-DR says the light is 32. (ISO 100, 1/30 sec f/8.0.7).

I can get my source near 4700 K by adding 80A and 82A filters in the path. But when I do that the Weston’s do not reach 32 when the Sekonic says it is.

I am convinced by this experiment that Weston’s were calibrated to 2800 K and it is driving me crazy trying to calibrate them at 4700K.

Heck, it’s insane trying to calibrate them at 2700K.

 

[Stephen Benskin]

There is no ISO standard “K” value. Meters don’t all read the same.

You could say there is a standard value for K. It is 12.5. It tends to incorporate the standard value of q and P (K`1) with the exposure meter variables being neutral or 1.0. Most exposure meters have a K of 12.5 or 14.0. This difference can be attributed to the spectral sensitivity of the photo cell which in the equation of K has represented by r. The range of r and its affect on the value of K breaks down as:

Selenium Cells –
r = 1.0 to 1.2
K = 1.16 to 1.40 cd/ft2
K = 12.5 to 15 cd/m2

Cadmium sulfide cells –
r = 0.8 to 1.0
K = 0.93 to 1.16 cd/ft2
K = 10 TO 12.5 cd/m2

r = 1.0: K = 12.5
r = 1.1: K = 14.0