PDA

View Full Version : 3 meters read differently from my dslr... Why???



Heespharm
26-Jun-2011, 09:03
So here's the setup.... Gossen Luna pro, soligor digital 1* spot and sekonic 400 series 1* spot... Canon t2i dlsr VS 18% grey card under even light


I use the reflective mode on the gossen and the two spot meters on the middle of the grey card... All reads EV 9 then I take a picture of the grey card under evaluative and spot metering with the dslr and the equivalent shutter speed and aperture readings equals EV8 (histogram shows peak in the middle of the middle zone)

What gives???? I think I did the tests pretty scientifically and either all my meters are wrong or the dslr reading are based on a different light scale?

engl
26-Jun-2011, 09:19
The look of the histogram in the camera is going to depend on the JPEG settings used, as well as the tone curve inherent to the camera. ISO is also not very well specified for digital cameras, or rather, there are several conflicting specifications. One is related to the light needed to saturate the sensor, while one is based on the resulting output levels after processing (for some setting).

In short, if you want to use a digital camera as your meter (which I do), you are going to have to experiment and find how it relates to the rest of your workflow (film, developing etc.).

Mark Woods
26-Jun-2011, 09:28
Great response Stefan.

Ivan J. Eberle
26-Jun-2011, 10:34
Was the angle of reflectance of the light the same off the card for each? Spot metering v. incidence has that variable.

Too, what color temperature was the ambient lighting? There was a long recent thread here about how many modern meters are tuned for a particular color temperature with results varying widely outside that window. Not having a Canon Rebel I can't speak to that, but my Nikon DSLRs and F5 have RGB sensor array meters which don't suffer this, they're color-temperature agnostic. These and my Pentax 645N spot meter are extremely accurate, and agree closely with one another (and my old General Electric DKW-58 selenium meter usually agrees with them. The DKW-58 is probably all I'd ever need for shooting color negative films from dawn to dusk outdoors, incidentally).

Ignore histograms for film metering purposes--except as quick and dirty indicators of subject brightness ratios. (You won't know or see what's being clipped by the .jpg algorithm, and if exposure is on the edge of blowing out a transparency film highlight, you can too easily miss it looking at a histogram.)

Or shoot color negative film and don't sweat occasionally being a stop over or under. (It's what the newsies gravitated to before abandoning scanning film for DSLRs en masse.)

Leigh
26-Jun-2011, 10:36
Metering systems in dSLRs are NOT "light meters". They're analytic systems that use complex algorithms in an attempt to determine what kind of scene you're shooting, and what exposure is most likely to result in an acceptable image.

The algorithms are designed around the characteristics of the camera sensor, not the characteristics of film.

Since a gray card is not in the scene library, they're not likely to interpret it correctly.

You have three real meters that agree. Believe them.

- Leigh

engl
26-Jun-2011, 11:06
Metering systems in dSLRs are NOT "light meters". They're analytic systems that use complex algorithms in an attempt to determine what kind of scene you're shooting, and what exposure is most likely to result in an acceptable image.

The algorithms are designed around the characteristics of the camera sensor, not the characteristics of film.

Since a gray card is not in the scene library, they're not likely to interpret it correctly.


The analytic system part only applies to matrix metering, most DSLRs also have "dumb" center weighted and spot metering. However, as you say, the metering systems are in all modes tweaked for the characteristics of the sensor and the processing used by the camera.

Heespharm
26-Jun-2011, 12:13
Angle of reflectance was same height aimed from me standing withe meter at eye level aimed at direct center of grey card....

I don't shoot much slide film but I do have 115 sheets of rvp50 4x5 so good to know...

I'll just trust my light meters... And the zone system...

I guess the camera settings in dslr are just for the dslr and doesn't actually measure luminance in an objective way... More just in relation to the sensor??? Am I thinking correctly??

Jack Dahlgren
26-Jun-2011, 13:08
Angle of reflectance was same height aimed from me standing withe meter at eye level aimed at direct center of grey card....

I don't shoot much slide film but I do have 115 sheets of rvp50 4x5 so good to know...

I'll just trust my light meters... And the zone system...

I guess the camera settings in dslr are just for the dslr and doesn't actually measure luminance in an objective way... More just in relation to the sensor??? Am I thinking correctly??

The dlsr is reading light at the sensor plane - so there is some loss through the lens and perhaps a bit of bellows compensation required unless you are focusing at infinity. Both of those together are probably not a full stop, but they are definitely non-zero.

Tony Evans
26-Jun-2011, 13:54
My Spot Meter reading of the Grey Card on the Pentax K20D varies with which lens I have on (view adjusted to cover Grey Card). Same for a big uniform lit wall. For analogue shooting, I always go with my Soligor Digital 1 degree Spot, which I get calibrated every 1 -2 years. I`m surprised by the EV 8. Most DSLRs under-expose by about one- to two-third stops, presumably to prevent overblown highlights.

Ivan J. Eberle
26-Jun-2011, 18:34
Using consumer-grade DSLR model to use as a spot meter begs the question of whether you were trying to use it in a manual metering mode with a kit lens-- perhaps one not having a constant aperture? Some of these drop lose two full stops from one end of the zoom range to the other.

Heespharm
26-Jun-2011, 20:13
Using consumer-grade DSLR model to use as a spot meter begs the question of whether you were trying to use it in a manual metering mode with a kit lens-- perhaps one not having a constant aperture? Some of these drop lose two full stops from one end of the zoom range to the other.

70-200 f2.8L lens so um no.. Plus I was using both matrix and spot metering... It was consistently off

Ivan J. Eberle
26-Jun-2011, 21:17
Since you've taken the other variables out of the equation and are getting a result with the spot meter that doesn't agree with 3 other meters that do agree--it's a safe bet it's the camera's fault. One could always try "matching the needle" by moving the ISO or exposure compensation dial up to be consistent with the other meters to see if it was possible to get results consistent with the other meters across the board under different lighting conditions. But a whole stop off seems pretty odd. Strange enough that this particularly camera wouldn't be my first line, go-to metering solution (as my Nikon DSLRs most often are for me).

Since Matrix Metering was mentioned earlier, I'll also note Nikon first used it for film, and quite effectively, too. With color temperature sensing 1003-pixel RGB array Matrix metering is arguably more accurate than any old tech handheld meter under more varied color temperature conditions. At it's best (and I do believe my F5 has one of, if not THE best), it's also entirely capable of instant metering between frames at 8 FPS and was famous for almost never blowing out chrome film highlights. It's only when the scene has a subject brightness range that exceeds many chrome films that it starts to make educated guesses about what to clip. I'm not remembering whether the Nikons used DX coding to know whether the film type is chrome or negative, but they might have.

Leigh
26-Jun-2011, 21:22
But a whole stop off seems pretty odd.
It's not necessarily a whole stop off.

It could be 1/4 stop or less, depending on the resolution and display options of the various meters.

For example, if a reading is slightly over 8.5 it may display as 9, while slightly below 8.5 may display as 8.

My Sekonic meters display 1/10 f-stop increments, but other meters have other defaults.

- Leigh

Lightbender
26-Jun-2011, 21:26
Do a search on 18% grey.. You will see that there is some debate that 18% grey is not intended to be "middle grey".

Armin Seeholzer
27-Jun-2011, 02:47
Don't forgett the Sensors have a not a curve like film, they have a strait line not a curve! And the high lights are very fast burned out! So most of them do a bit underexpose for to save the high lights!

Cheers Armin

z_photo
27-Jun-2011, 04:09
as i recall there was a discussion of this on luminous landscape (or one of those sites) that seemed to be spot on. (sorry about that)

msk2193
27-Jun-2011, 05:00
Most pro level dSLRs will allow you to set a + or - fractions of stops as the "natural" setting. I find that all Nikons for instance measure about 1/2 step low, and have set the camera bodies to automatically add 1/2 stop to the metered reading to get proper results.