What I see as potentially useful information from your tests is that, if you photograph under varying color temps (e.g. indoor incandescent, outdoor daylight, etc.) then it would be a good idea to calibrate your process for each of those general scenarios.
.
Thanks for being the "voice of reason" Jim, yep, that's the sensible conclusion to draw from this....or at least be aware of such.
Of course if you shoot neg film , you have greater exposure error capacity, and it might be overlooked, specially if you bracket....
But regardless, a spritz of common sense sure was refreshing here, thanks!
I wonder what the George of Quality Light Metric would say to all this, besides "Your meters need to be calibrated. Send them to me."
Bill there's ample evidence that many meters are delivered off calibration. And I've never got a iused meter that was trustworthy. OTOH, the on-board meters in my humble low-end Nikons have only rarely steered me wrong.
I once took a used Lunasix to the guys in Springfield, NJ, who explained to me what's required to test a meter properly. Bill, I'm not sure you have it. I know I don't. I also know that when both were freshly calibrated, my Lunasix 3 (Bogen did it) and Master 5 (George at QLM did it) didn't often agree closely, whether reading incident or reflected. Funny thing is that both gave me ok E-6s. They still don't agree closely. And when I don't present them with a difficult situation both still give me good E-6s, be it at high noon or at dawn/dusk when the light is dim and cold.
Cheers,
Dan
> Bill there's ample evidence that many meters are delivered off calibration.
And I've never got a iused meter that was trustworthy.
The 608 was sent in for calibration recently, it was nutz-on.... the other meters were all brand new.... So none "used" except the 608 which I bought new and had rechecked.
> I once took a used Lunasix to the guys in Springfield, NJ, who explained to me what's required to test a meter properly. Bill, I'm not sure you have it.
Dan, once again, I did not "test" these meters as you suggest, instead, what I performed was a "relative comparison" between the meters, nothing more. The results demonstrate, the errors between meters vary greatly based on the type and manufacturer over a wide range of color temp and EV. It's that simple. If you threw some more money at this, we would learn more..... I spent 15 minutes on this test, and about an hour so far trying to explain the results on this forum. So far, Jim, Mike and Rob, "get it".... consider this post a "be aware of" type post. Don't consider this post a "hardcore scientific study that will be published in peer reviewed journals". It's amazing I have to even mention this.... It's no surprise many great contributors that used to be on this forum dropped out long ago.
> Funny thing is that both gave me ok E-6s. They still don't agree closely. And when I don't present them with a difficult situation both still give me good E-6s, be it at high noon or at dawn/dusk when the light is dim and cold.
I think you summed it up good here.... when the meters are used close to their calibration color temps, they are accurate enough, agreed. You can see this from test results at 5370 color temp....all meters were within 1/3 of a stop, not bad. Its when the meter is used outside of their calibration zone, the error variances becomes much greater. Notice the 7750 color temp reading, 1.3 stop variance. Of course, this is just the variance, the true error can be greater than this....we don't know without a reliable standard.
bg -
How do you print your chromes after processing?
Your numbers don't seem so far off (color) that some minor tweaks in PS would not remedy. What am I missing?
As far as the amount of light read I see more concern.
When my 13 year old Sekonic final fell apart I exchanged it with an upgrade for a new (at the time) L-508. I tested it against every make and model I came across in everything from ambient daylight, to fluorescent to studio flash. Generally I found a difference ranging from -0.4 - +0.8. The current setting on my meter is biased @ -0.8.
Did you normalize for (or even checked for) differences in K factor between the meters... or is there a reason why nobody addressed this known source of variance, even after I mentioned it earlier?Originally Posted by bglick
Additional question: does anyone know off-hand what ISO, NIST (or any other) standard specifies photographic light meter calibration? I know that scientific photometers and industrial light meters are calibrated to some standard but don't know if they are the same or what the acceptable range mignt be.
Let me clarify. I check all three meters simultaneously, pointed at different gray values
AND different hues of a Macbeath chart, under identical lighting, using a Minolta color temp meter is necessary. They match. Does that mean the linearity with all hues is
a straight curve. Of course not. You have to understand the spectral sensitivity of the
sensor in the particular brand of meter. But all three meters are predictable in exactly the same manner. If not, it's time to recalibrate. I also have a lab meter about hundred times more accurate than these Pentax meters. It is programmed to read in density units rather than EV values, so isn't for camera use - but if I know how to program and work with something like this I damn well understand what I'm talking about with a handheld meter. I can't speak for the quality control of Pentax meters in current production, since I haven't bought one in the last 15 years; but I can state that when Quality Light Metric recalibrated the ones I have used, the result precisely matched the linearity and EV value of my unused control meter at its original factory settings.
> but if I know how to program and work with something like this I damn well understand what I'm talking about with a handheld meter.
Drew, what you have demonstrated is that 3 identical meters, give identical readings when exposed to the same color temp and EV Range......right? How is that relevant here? I still don't understand your point as it relates to this post. My post is about different meters under different lighting conditions. Your post is about identical meters under different lighting conditions?
As I mentioned, I have another digital Sekonic that is nutz-on vs. the 608, never more than .1 stop difference. That demonstrates to me, that Sekonic meters of the same type (digital) have tight manufacturing tolerances and all perform relatively close under most all lighting conditions. It gives no indication the readings are accurate vs. the other meters I tested. does this make sense?
Brian, if you describe your K factor more clearly, I will try to respond.... and there is no industry standard for photo meters... at one point, the german meters were calibrated at 5500K daylight, and the Japanese calibrated to ~3000k ...... but, supposedly, now, Sekonics are calibrated to daylight as well.... not sure this is true, its based on speaking with Mamiya USA rep, can't find anything in writing....
NewBearings, sometimes I don't print my chromes, the chrome itself is the final product....other times they are drum scanned and corrected. Interesting how you bias your meter by .8 stop..... One of the reasons I wanted to run this test, was to see if there was a simplified biasing method that would help compensate.....but as you can see, as color temp changes, the bias would have to change directions as well. Again, this is based on the Lux value readings of the Gossen 3, which we all agree is not a bona fide standard, but regardless, i was curious of there was any trend I could latch on to, then maybe I would have ran another test....
Keep in mind, unless you are doing copy work, "accuracy" may be a subjective determination. Most gear requires some "calibration" and fine tuning. But this is often based on your own system and -subjective- needs.
Somewhere buried in my archives is a graph by Pentax showing the spectral sensitivity
graph of their silicon sensor. It peaks close to the maximum sensitivity of human vision. Therefore a reading from a green surface of grass, forest, etc easily
substitutes for a gray card reading or gray rock etc. In the real world you simply get
used to your meter. My very first camera was a very early Pentax with a exterior
coupled CDS meter, and I don't think I ever botched an exposure, even with Kodachrome. My apparent luck at the moment is that the sensors in my manual Nikons
seem to behave very similarly to the cells in my handheld Pentax meters, apparently due to a similar mentality and similar silicon cells in use (also similar vintage Japanese
mfg). There obviously has been an attempt with the modified Zone VI meter to correct for certain color inadequacies, which has been discussed many times on this forum. But color temp per se has far less to do with this than spectral response and what kind of standard you take as a reference. Sensitometry is a subject in its own right. And it's impossible to make a meter that's completely color-blind; and I don't
even believe it would be smart to make all meters behave exactly the same, because
they are engineering for a different range of applications. Choice is important. But
to call them unreliable for color readings is absurd. Color tranny film is fussy stuff;
we'd all go broke if our meters were unintelligently designed.
Posting Permissions
Reply With Quote
Bookmarks