How accurate are exposure meters? Not very....

[bglick]

I have been experimenting with light meters, again. Many years ago, we had a long thread on this subject, it was very interesting, lots of good contributions. Light meters are calibrated at one color temperature. So how far off is the readings when the color temperature changes from the Kelvin range it was calibrated at ?

For those who don't shoot chrome film, and are NOT ultra sensitive about nailing exposure, please stop reading here. :-)

Anyway, I took my Gossen 3 Color meter, which has RGB sensors... it produces a Lux value for light intensity based on 3 color sensors. In the old thread, it was suggested that an ideal light meter would read the 3 colors of light, and then produce an EV value... discussions then went into having the meter calibrated for each of the color films...OK, that was really dreamin considering film was dying then, and now.... well.....

I used the Gossen 3 Lux reading as a baseline reading. I converted Lux to EV via a table such as this on Sekonics web site....

http://www.sekonic.com/support/support_2.asp

I took incident meter readings with 4 meters and compared them with the Gossen Color 3 EV value... the 4 meters were...

Sekonic L 608
Sekonic L 208 (very small manual)
Gossen DigiPro F
Gossen DigiFlash (mini meter, new model)

the color temperatures I tested ranged from 2300 to 10k Kelvin. I had readings at almost every 1k color temp interval.

I used the Gossen 3 meter Lux / EV reading as the basis, and compared all the other readings to it. I sorted the results by color temp. and by EV. A value of 0 means the incident meter matched the Gossen 3 Lux value. A value of -.4 means the incident meter read .4 stops below the Gossen 3 Lux / EV value. +.4 = incident meter read .4 stops above the Gossen Lux EV value.

As expected, the results vary tremendously, not just against the Gossen 3 Lux/EV but also vs. each other. IMO this demonstrates something I have always felt...... these meters are at best, "fair instruments", certainly not ideal tools if you shoot under varying light and color temperatures. These test results are all from outdoor lighting, so landscape shooters beware. I consider a .4 stop error quite large from a meter, for chrome film. Add in some errors of your own, processing, etc. Its not hard for your exposure to be off .75 of a stop. Then, as you can see, there is often errors near one stop. (all based on the Lux/EV value)

I was hoping the results would yield some compensation values I could apply to a given meter, when knowing the color temp, but that is not the case, as its a mix of color temp AND EV values that trip up the readings. Of course, maybe the color temp meter provides the most accurate exposure readings. Not sure of this either..... any thoughts?


Regardless of whether the Gossen 3 color meter provides the most accurate exposure readings, the sad finding of this test was.... the meters themselves vary tremendously from each other... often a full stop variance....you can read across the lines, -4 to +6 = 1.0 stop variance.

[Drew Wiley]

I have three Pentax digital spotmeters. One is kept as a reference and never used in
the field. All three give exactly the same readings over a wide range. If one differs,
I get it recalibrated. I also had a Minolta spotmeter; it eventually was stolen, but
gave virtually identical readings to the Pentax ones. My internal meter in my Nikon
FM3a is very close to my current Pentax readings. What I discovered was rarely
consistent, however, was gray card values from one brand or even batch of card
to another. That is why spectrophotometers are always calibrated to an individual
ceramic tile which doesn't fade. But as far as the meters I personaly use, I'm
convinced they're very reliable, and I have to recalibrate them less than once a
decade. I have far more accurate custom meters for lab use per se, actually too sensitive for field use; but this would be overkill for general photography anyway.
If there's going to be an error it's a lot more likely to be due to poor judgment on
my part or possibly flare than due to any inconsistency with my light meters.
Sorry if your meters aren't so satisfactory.

[bglick]

> I have three Pentax digital spotmeters. One is kept as a reference and never used in the field. All three give exactly the same readings over a wide range.


Drew, I have other Sekonic meters similar to the 608....guess what, they all read identical to the 608. What does this prove? You are making a wild assumption that if 3 identical meters, give identical readings, than all 3 produce accurate readings. What is the basis of this? Your meters, just like the meters I showed, can only be calibrated at ONE color temperature. So of course, they will all read very similar, they are the EXACT same meter, calibrated to the EXACT same benchmark..... ??? This demonstrates nothing about accuracy over a wide range of color temps and light intensities. That was the finding from this test.



> But as far as the meters I personaly use, I'm convinced they're very reliable,


All my meters are very reliable also.... but the test I performed, demonstrated they all can't be accurate....and if yours was in the mix, it too would be all over the map vs. the Gossen 3 which reads all 3 colors before calc. light intensity. That was the point of the post....

[Bill_1856]

My Pentax digital spot meter came from the factory one stop off (I don't remember if it was high or low -- never got around to having it adusted). Instead, I just just a little Kentucky windage for my Weston Euromaster.

[Oren Grad]

So that's half the equation. Does the effective speed of your film change at different color temperatures?

[Kirk Gittings]

How accurate are exposure meters?

Accurate enough obviously.

[Vlad Soare]

I don't need a perfectly accurate lightmeter. I only need one that's good enough for the job. And from this point of view my Gossen has never disappointed me. As long as I get perfectly exposed slides with it, I couldn't care less about its theoretical accuracy.

The same goes for shutters (is 1/125 really 1/125? No, but it's close enough), for apertures (is the amount of light passing through two different lenses at the same f-stop really the same? No, but it's close enough), for the film position (is the film really placed at the exact same spot as the ground glass? No, but it's close enough), and so on.

So that's half the equation. Does the effective speed of your film change at different color temperatures?

Good point. Yes, I think it does.

[Ivan J. Eberle]

I can remember hardly ever shooting a bad chrome with either of my Nikon F5s, which camera had arguably the most sophisticated RGB meter of it's era. Downside was that it ate battery charge like nobody's business. (On a stay-alive remote just sitting there waiting to fire with the meter on but doing nothing else, I never got more than 7 hours out of 8 AAs).

Now after nearly a year of toting around a DSLR to use as a spot meter for my LF color transparency work I'm tempted to go the C41 route to simplify. With NPS 160NC or Portra NC and Sunny 16 I figure my GE DW-58 Selenium footcandle meter would pretty much handle everything between fifteen to dawn and a quarter past dusk.

[Scott Knowles]

Interesting, but it makes the assumption the Gossen 3F is "the" standard for light measurement, which in reality, it's just another light meter, albeit an excellent one, but still relative to reality. If you want a real test, then use a scientific standard for light measurement against all five light meters. Then maybe the results have merit more than a comparative test, and I wouldn't be surprised if even the Gossen meter has an error factor.

[bglick]

> Accurate enough obviously.


For some, yes.... for others shooting in very tight exposure latitude scenes, "obviously", there is a lot of room for error.... If I shot neg film, I wouldn't care the least bit.




> I don't need a perfectly accurate lightmeter. I only need one that's good enough for the job.


Hence why I wrote this below, in the 2nd line of my post....
> For those who don't shoot chrome film, and are NOT ultra sensitive about nailing exposure, please stop reading here. :-)





> The same goes for shutters (is 1/125 really 1/125? No, but it's close enough), for apertures (is the amount of light passing through two different lenses at the same f-stop really the same? No, but it's close enough),


HUH? Are you serious? I am showing how exposure values are off by .3 - 1 stop.... and you think the 10% max. a shutter or apt. is off any given exposure is its equal? .1 stop, vs. 1 stop? 10x difference? I don't see the logic. BTW, I tested 15 of my LF lenses with a shutter tester, and the worst error was 3%, not 10% and that was for one lens, and one f stop only...thats how accurate most shutters are. So I don't really follow your logic..... Whatever......




> for the film position (is the film really placed at the exact same spot as the ground glass? No, but it's close enough), and so on.


GG alignment? Exposure? ?? Anyway, we have something called Depth of Focus at the film plane to protect against small gg/film alignments. We have no equal when it comes to exposure. But this discussion is not about focus....




> Interesting, but it makes the assumption the Gossen 3F is "the" standard for light measurement, which in reality, it's just another light meter, albeit an excellent one, but still relative to reality.


Scott, even if you eliminate the Gossen 3 as the "standard", look at the errors between the meters themselves, that was the point. If they all had very similar errors vs. the Gossen 3, then I would agree, but its obvious, the errors against each other are extreme IMO. This demonstrates at different EV levels and different color temps, the different makers of meters have errors that are not consistent with each other.




> If you want a real test, then use a scientific standard for light measurement against all five light meters. Then maybe the results have merit more than a comparative test, and I wouldn't be surprised if even the Gossen meter has an error factor.


A more scientific test would be nice.... but if you take two new different makers meters, and they read 1 stop different than each other...... the evidence is obvious...they both can't be accurate. Even if the error falls right in the middle, thats .5 stops off per meter.... that's the best case scenario. That's the point.... no need to go to the next level unless you want to track each of the errors vs. a standard. This test identified the meters have much larger errors than most think, period.



And yes, the Nikon did have a "color" meter to evaluate exposure IIRC..... good point... I think that was the the only camera that incorporated such...