[StephenHampshire]

I find the preview histogram to be least reliable in scenes with high contrast and a relatively small area containing highlights. Of course in such scenes blown highlights aren't necessarily a problem even on digital, but when I do want to expose for the highlights in such a scene I will use some combination of spot metering and/or deliberate "underexposure". Such as when shooting sunlit white flowers against a background of shaded foliage:

Agree 100%, white flowers in sun are tricky, I have found that using the "link ae to af" option can be useful here too. The oppositew is true for my steam train pictres as I normally focs on some part of the loco which is often black

[maltfalc]

First, boosting ISO via analog amplification is only done at low ISO levels where noise just isn't a major concern anyhow. Who really cares about the difference in noise between ISO 100 and ISO 140? And once you get to a point where you do care about noise - say, ISO 1600 (specifics vary by model) - all ISO boosting in camera is digital anyhow. So there is *no difference* at high ISO settings.

Second, while there might be some theoeretical advantage to analog amplification over digital, there is pretty much zero chance that you would ever be able to *see* the difference except perhaps in the most extreme cases - eg, an ISO 100 image pushed four stops (digital amplification) versus one shot at ISO 1600 (probably mostly analog application on most camera models).

The point being, unless we are talking doing about something really stupid like shooting four stops underexposed at low ISO then pushing in software, the types of differences we are actually talking about here - typically ETTR might result in what, 0.3 - 1.0 EV higher exposure? - *just will not matter in the real world*. Even pixel peeping, it will be incredibly unlikely anyone would be able to tell the difference between a half stop push in camera or in PP. Which is to say, the actual benefit is next to nil. And then you have to balance that against the very real risk of blowing highlights.

Again, here I am talking ETTR via pushing ISO. That is what I am saying is pointless, and indeed counterproductive given the risk. If you instead perform it by slowing the shutter or opening the aperture, only then can you actually gain the theoretical advantage of this approach, as yes, it is indeed possible to see a half stop difference in noise, especially at higher ISO levels. But again, at those higher ISO levels, it's pretty unlikely you have the luxury of slowing the shutter or opening the aperture - you wouldn't be at such high ISO levels if you weren't already hitting maximum aperture and minimum shutter speed for your shooting situation.

So, I'd reduce the whole concept of ETTR to this: don't be stupid and shoot more than a stop underexposed or more than a stop higher ISO than necessary if you can help it. All else is just not worth worrying about, and does nothing but confuse people and risk ruining pictures.

rather than go on a long-winded dismissive rant, please just admit that your original claim was factually incorrect and learn to take critisism. also, noise IS a major concern even at low iso if you're making large prints, cropping or trying to bring detail out of the shadows.

[Marc Sabatella]

rather than go on a long-winded dismissive rant, please just admit that your original claim was factually incorrect and learn to take critisism. also, noise IS a major concern even at low iso if you're making large prints, cropping or trying to bring detail out of the shadows.

I did not post a dismissive rant. I posted a clearly explained technical explanation of why your claims were factually incorrect in some respects and irrelevant in others. At high ISO, there is no difference whatsoever. At low ISO, I still challenge any person on earth to reliably tell the difference between a half stop push via analog versus digital amplification, at any print size you choose. I claim it can't be done. Do you have evidence to the contrary? If so, simply post it.

[robgo2]

The article saysnimbus more than that - it also points out that even if you do succeed in avoiding blown highlights, the actual benefits of ETTR are not nearly as great as some would have you believe. So, with a potential downside that is greater than many might realize, and a potential upside that is smaller than most might expect, it really does give one pause as to whether it is really to one's advantage to deliberately aim for a higher exposure than one actually wants (which isf course the central tenet of ETTR).

Exactimento! To put it in economic terms, the cost/benefit ratio of ETTR is too high for it to serve as the basis for determining exposure. It is much more prudent to leave a good bit of room to the right of the histogram.

Rob