FWIW, only. My result of dialing in +2/6 EV for the multi-pattern metering seemed to throw the entire metering system out of whack for, high-contrast, varied-subject images. The result was a mix of about 80% near perfect exposures that came within 0.1 EV of ideal, to around 20% that were nearly a full EV over exposed. Very weird. To confirm, I turned it off and things are back to Sony normal, meaning most files were pretty consistently ⅓ EV under ideal. Me, I'd rather live with a consistent ⅓ under than wildly mixed. To be clear, this was NOT an intensive or scientific test, just a quick point and shoot session under varied high-contrast light -- but it was enough to tell me to leave well enough alone *for me*.
j4nu wrote:
Not sure what you intended to convey here, but my point was that zebras are good enough to protect highlights from clipping (if zebra level lower limit is properly set), which in turn maximizes captured DR.
I understand your point better now, and I agree that properly calibrated zebras can be good enough in practice to protect important highlights from raw clipping. My objection was only to the wording that zebras show the “actual RAW DR latitude,” since they normally derive their warning from a processed signal rather than directly displaying raw-channel values.
So I think we largely agree on the practical method. My qualification is simply that calibrated zebras are a very good proxy for raw clipping—not a literal display of the raw data itself.
