Quote Originally Posted by David Allen View Post
"be sure to use the same camera, shutter and lens throughout" - the reason for this is that the testing procedure automatically compensates for flare.
We've discussed this point on a number of occasions. In fact, this was the topic of Michael's question. Long story short. It doesn't compensate for flare. Flare is caused by higher luminances affecting lower illuminances within an optical system. Traditional ZS testing uses a middle gray target of a single tone. Even if flare existed, the influence of flare diminishes as you move up the curve. The metering of the target and stopping down may produce the shadow exposure but it's still the middle tone luminance where even in normal flare conditions, there is little influence. And this is completely separate issue from the question of introducing such a variable into a test.

This is a good example of my earlier topic about the helpfulness of understanding theory. Helpful not so much as part of achieve precision testing. That's not necessary for most photographers. It's helpful in knowing the limitations of the tests.

"any variations introduced by a change of equipment (such as the possible difference in aperture calibration or flare introduced by changing lenses) should be quite apparent if they are significant." - i.e if you have calibrated with one camera, one lens, one meter, one developer, one thermometer, etc introducing a variable such as a different lens will immediately show if flare, for example, is a significant factor. If it is, you then know that you need to re-calibrate for that particular lens by redoing the testing sequence.
This can only be true if the photographer shoots with the same lens at the same f/Stop and shutter speed. And this is a good example of my other point about good testing and bad testing. My contention is that speed testing isn't necessary for most and the only testing really needed is for contrast (and that only applies to maybe 10% of photographers). In order to achieve knowable, quantifiable results, the variables and testing conditions must be known and controlled. In other words, scientific testing. Many of the disagreements about testing come from people arguing from different perspectives. It isn't a question of which is more accurate. But there's a big difference between testing for film speed (scientific) and testing for EI. They are different things and have different purposes.

Sorry, ran out of time.