Understanding our Light Meters

dmr said:
The one eternal problem I have on this subject is how to check or assure the accuracy of a light meter, whether in the camera or a handheld meter. There seems to be no reliable and convenient method of doing this.

For about every other physical measurement we have something that's easy and convenient and more than accurate for everyday use. For length we have rulers, for volume we have measuring cups, for mass/weight we have gram weights, for temperature, we have the freezing/boiling point of H2O, for time we can measure the high point of the sun if we have to.

For luminance or illuminance, we really don't seem to have anything more precise available to the lay person other than "Sunny 16" which I think has too many variables and gotchas.

There really isn't anything like a standard candle or anything traceable to a known standard.

I have several cameras, and I know they are somewhere in the ballpark, but for some reason it frustrates me that all I can do is compare one with another.

Any comments on this angle?
Click to expand...

A good handheld meter should be pretty much "on", but just to be certain, you can have it checked out for accuracy. Then, you can use it as a reference vis-a-vis your in-camera meters, but you need to be careful in terms of hat you're really comparing. A lot of people who insis on using handhelds value them most for the one thing in-camera meters aren't good at: incident-light metering. Unless you set up your handheld for reflective metering, you wind up, to quote the Brits, comparing chalk to cheese.

Trickier still is dealing with several different cameras whose meters have differing metering patterns. A camera with spot or tight center-weighted metering will respond differently from a camera (particularly an older camera) with plain old "average" metering. Throw in age, abuse, and, inthe case of certain older cameras, having to substitute no-no mercury cells for a more eco-friendly susbstitute, and you can almost wind up chasing your own tail.

Anyway, just start with a good handheld of known quantity. Have it checked by a repaor center if you're feeling a bit paranoid. Find someone else with a handheld and compare hers/his with yours. If it checks out, you can more or less take it on your own from there.


- Barrett
 
1. Handheld Gossen Luna Pro, CDS...
is a good all around incident meter.
I take a few readings in very bright light,
and find this meter excellent in low light.

2. Center weighted meter in 30 year old OM-1,
is also CDS, and gives readout quite similar
to the Gossen.
Ciao, mike
 
rxmd said:
But since reciprocicity failure is known (either through manufacturer's data sheets on the Schwarzschild factor, or through experimentation), a meter that is sensible in low still helps in getting accurate exporures at night.

Philipp
Click to expand...
Yes, I'd agree with you, hence I didn't say such a meter is of *no* use. However, it still requires you to have a table of corrections, either in your head or on paper, for the film you're using and the lighting situation. There are many variables involved and some degree of compromise, hence the meter's absolute accuracy is less important. Personally, I've taken shots (on transparency) at 20 and at 30 seconds, both of which came out quite acceptable.
 
dmr said:
The one eternal problem I have on this subject is how to check or assure the accuracy of a light meter, whether in the camera or a handheld meter. There seems to be no reliable and convenient method of doing this.

For about every other physical measurement we have something that's easy and convenient and more than accurate for everyday use. For length we have rulers, for volume we have measuring cups, for mass/weight we have gram weights, for temperature, we have the freezing/boiling point of H2O, for time we can measure the high point of the sun if we have to.

For luminance or illuminance, we really don't seem to have anything more precise available to the lay person other than "Sunny 16" which I think has too many variables and gotchas.

There really isn't anything like a standard candle or anything traceable to a known standard.

I have several cameras, and I know they are somewhere in the ballpark, but for some reason it frustrates me that all I can do is compare one with another.

Any comments on this angle?
Click to expand...
Actually there *is* a standard definition of illuminance and a "standard candle" derived from it. Look it up on Wikipedia or some such site.

You may have rulers for length, cups for volume and so on but they are no more traceable to national standards than your light meter. Time is one of the few "standards" accessible to Joe public but then again there are several methods of measurement...Sidereal, solar, fixed-star, atomic etc.
 
Ok, let me bounce this idea off of the collective wisdom here. I was talking about this to a couple co-workers and one of them noted that they make bright white LEDs in known brightness intensities. Quickly looking through a supply book, we found 1000 and 2600 millicandela LEDs (1 and 2.6 candela).

Now I'm sure that there's quite a bit of tolerance, and that these are minimum light values, but we all agree that one of these should be very consistent in light output as long as you operate it from a regulated power supply and at a reasonably constant temperature (like ordinary comfortable room temperature).

So, the project is to make kind of a "shadow box" of a known length with a LED at one end and a window for a camera or light meter at the other end. Then set this up with a 12 volt plug-in power supply, a 5 volt regulator chip, and a known series resistor to run the LED from 5 volts. Then do one of the following:

A. Do a whole bunch of very messy calculations to convert the 1.0 or 2.6 candela light output and the length of the box to the expected EV you should expect.

OR

B. Measure the light output at the end of the box with a known-good light meter.

I like plan B better. :)

From that point on, any camera or light meter should read the expected EV when positioned at the end of the box.

This is really a more high-tech refinement of my current method of measuring one camera to another on a blank bathroom wall with the lights on dimmers.

Anybody see any major problems with this idea?
 
A cloudless blue sky with your back to the sun will give a meter reading of sunny sixteen. You can check your meters with it. Of course you have to wait for a sunny day.
 
if it is digital or analogue, it's only a matter of interface. Some people definitely like the needle type interface better.
None of the two is generally better.
 
dmr said:
I have several cameras, and I know they are somewhere in the ballpark, but for some reason it frustrates me that all I can do is compare one with another.

Any comments on this angle?
Click to expand...

If this bothers you so much, dmr, i really advise you to buy a new and totally reliable meter. ONE, not more. Nothing to compare.
 
dmr said:
Ok, let me bounce this idea off of the collective wisdom here. I was talking about this to a couple co-workers and one of them noted that they make bright white LEDs in known brightness intensities. Quickly looking through a supply book, we found 1000 and 2600 millicandela LEDs (1 and 2.6 candela).



Anybody see any major problems with this idea?
Click to expand...
Yes, I see one MAJOR problem. If you look at the specs for white LEDs (or any other colour for that matter), you'll see that there's a huge tolerance, which renders your plan A to be built on sand. For plan B, using them as a comparison means, there's no issue but for absolutes they're useless. The outputs quoted are usually "typical" but the tolerance is often plus or minus 50% or so.

The sun provides a reasonably steady and known output (fortunately for us!) so it can be used as a calibration source of sufficient accuracy...if you're lucky enough to get sunny days!
 
The reason Sunny-16 is so accurate is that the sun IS a constant light source.

Incident meters are most accurate because they measure the source of light without factoring in the reflectance of objects being metered.
 
You must log in or register to reply here.
Back
Top Bottom