selmslie wrote:
Absolutely incorrect!
Film camera meters may not have been very accurate but they did not need to be. Film was very forgiving.
Hand-held meters were more accurate, especially if they could be easily adjusted.
But digital camera meters are extremely accurate and even more precise, otherwise we could not use them to make digital images. Their precision and accuracy can be easily tested by shooting raw images and analyzing the results with RawDigger.
I have been testing four cameras using a standard calibrated source and three of them (Z7, A7 III and X100T) come up with exactly the same spot meter reading.
The only one that is different is the A7 II which has had the Bayer array removed to make it record only monochrome images. To accommodate that change I need to leave the EC set permanently to EC+1 and that lets it agree with the other three.
What the cameras does with the raw information in the creation of a JPEG varies slightly but that has nothing to do with the meter.
RawDigger has made it possible for me to study how the cameras collect information.
Absolutely incorrect! br br Film camera meters ma... (
show quote)
Amazing.
First:
"Calibrating an instrument means a statistical procedure in which it is compared to an instrument of higher accuracy and which is traceable to a standard."
This holds, because it is the *very definition of calibration" It is true for light meters, volt meters, even tuning forks and rulers.
Second:
"The camera is not such thing, it has many more sources of inaccuracy like shutter speed, iso and aperture, plus in camera processing."
Still holds. While it is true that modern cameras are more accurate than film, still can not be used as a standard to compare any instrument (light meter, or whatever) because:
a) you don't know if your camera had any calibration issues after it was released from the factory
b) it lacks a laboratoy certificate of calibration
Believe me, I design electronic instruments and had to study this trough and in fact, in a occasion - because I lacked a proper meter (assistant forgot it) - I tried to make a make a luminance metering of a road with my A7C and it had such a dispersion that it was not statistically meaningful.
The next night, I could perform the measurement with a proper meter with a valid certificate.
Besides, a illuminance or luminance meter should be calibrated once a year (typically after DIN 5032 Part 7) , after that the certificate is no longer valid.
Besides 2, we are talking engineering here. Any statement without a number is meaningless... your curves lack the one and two standard deviation markers, se we really don't know what they mean. "
DMWAE rule.