SD1 Merrill Metering Headroom Test (technical post).

Still raining here, so I tested how close my Merrill in-camera metering is to "correct" in accordance with ISO/CIPA. Since the SD1 EXIF tag for ISO type is the ISO sensor saturation method the metering should provide a little under-exposure (-1/2 EV) at the sensor.

The good news is that it does. Shooting each side of the Kodak R27 gray/white card under the same lighting with 0 EC causes the green channel to be 13% of saturation which is about right since the green channel does approximate luminosity.

I shot at ISO 100 and used spot metering centrally on the card. Illumination was by soft white LEDs (approx incandescent - don't tell me about blue spikes).

Raw histograms for each side (18%/90% reflectance):

18% at left

550/4080 (sat) = 13.4%

Opened them in SPP 5.5.3 and converted to sRGB TIFF:

Ignore the cast and vignetting - I am measuring the center area

Now, for saturation-based ISO, the conversion, without EC, should give 100/255 in sRGB after selection of the best WB and color-balancing with the picker.

the GIMP says it does not:

see original for the numbers at bottom right

The SPP conversion has given a level of 151/255 !!

Before I get told that "base ISO" is 200 for a Merrill, allow me to remind us that the Merrill is ISO-less and a shot at ISO 200, also with no EC, would have given the same level.

I think this one reason why folks praise the highlight "recovery" of the Merrills. Neutral objects captured correctly in a scene would look too light (keeping it simple) and require SPP's EC slider to be backed off a bit.