rmalarz Loc: Tempe, Arizona

I have demonstrated it with images. Look at any images I've posted over the last year.
--Bob

selmslie Loc: Fernandina Beach, FL, USA

Did you miss where I said, "When it comes to moving subjects you might need to increase the shutter speed and possibly raise the ISO or open the aperture. But people are not then going to be looking for noise in those images. They will be more interested in whether you captured the action."

rmalarz Loc: Tempe, Arizona

Well, you were the one who brought the numbers game into play.

BTW, here's an excellently presented post on just this very topic. http://www.uglyhedgehog.com/t-433990-1.html

As for images, again look at anything I've posted over the last year or so.

--Bob

selmslie Loc: Fernandina Beach, FL, USA

You have never demonstrated the difference between using ETTR and using a lower ISO without ETTR.

I have posted real evidence that there is no difference. You can't point out any fallacy in my examples.

For example, suppose you post an image taken using your ISO 400 and +1-1/3 ETTR. You then ajust the Exosure slider in your editor by -1.33 Ev. Then post the same scene at ISO 160 using the same exposure (aperture and shutter speed), no ETTR.

Same exposure, same noise level, same ending tonal values. Where is there a difference?

You have never done this so we are back to the list of three excuses.

Scotty

selmslie Loc: Fernandina Beach, FL, USA

I understand what it is and how to do it. That has been covered ad nauseum.

I have been asking you, "Why I should bother?"

You can't show us any benefits that I can't get to more easily by lowering the ISO.

Let's not rehash this entire question from the beginning. Shows us some real images.

robertjerl Loc: Corona, California

Didn't miss it at all. And I didn't miss all the times you pounded "it doesn't work" or "prove it". Well you prove it is always true your way!

Just give it a rest, or start saying "In my opinion... or I prefer...or The way I do it is..."

selmslie Loc: Fernandina Beach, FL, USA

I never said ETTR does not work.

What I said was that any reduction in noise comes from the increase in exposure, not from the use of ETTR.

You can also get an increase in exposure by lowering the ISO and not using ETTR.

selmslie Loc: Fernandina Beach, FL, USA

ISO invariance may be at the bottom of this whole disagreement.

My cameras (D610, A7 II) appear to be invariant. However, so is Bob's D800.

Bob's lack of curiosity about how this affects the current discussion is troubling.

catchlight.. Loc: Wisconsin USA- Halden Norway

Scotty is correct again...
"Most 14-bit cameras on the market do not merit 14-bit recording. The noise is more than four levels in 14-bit units on the Nikon D3/D300, Canon 1D3/1Ds3 and 40D. The additional two bits are randomly fluctuating, since the levels are randomly fluctuating by +/- four levels or more. Twelve bits are perfectly adequate to record the image data without any loss of image quality, for any of these cameras (though the D3 comes quite close to warranting a 13th bit). A somewhat different technology is employed in Fuji cameras, whereby there are two sets of pixels of differing sensitivity. Each type of pixel has less than 12 bits of dynamic range, but the total range spanned from the top end of the less sensitive pixel to the bottom end of the more sensitive pixel is more than 13 stops, and so 14-bit recording is warranted.

...and

"Consider exposed to the right at ISO 3200. Then consider taking the same image, with the same shutter speed and aperture, at ISO 1600. The latter image will be one stop underexposed according to the ETTR ideology. In particular, the idea that the benefit of ETTR comes from the "number of available levels" suggests that the image quality would be one stop worse for the ISO 1600 image than it is for the ISO 3200 image, since by being one stop down from the right edge of the histogram, fully half the available levels are not being used. However, as we have seen, noise is much more than two levels in all exposure zones at these ISO's, so the extra levels used in the ISO 3200 image simply go into digitizing the noise, and are thus of no benefit in improving image quality. In fact, the quality of the two images will be very nearly the same (rather than one stop different):

The photon shot noise will be the same in photon counts for the same exposure (aperture and shutter speed). The photon noise is already present in the light before it hits the sensor, and couldn't care less what ISO you happened to choose. In raw levels, the photon noise will be half as much for the ISO 1600 shot because the gain (photons/raw level) is twice as much; a given number of photon counts of noise translates into half the noise in ADU at half the ISO.
The read noise for ISO 1600 (13.4 raw levels) is roughly half that of ISO 3200 (26.2 raw levels). Thus both read noise and photon shot noise will be half as much in ADU for the ISO 1600 image as they are for the ISO 3200 image.
Each feature of the raw histogram of the ISO 1600 image will occur at half the raw value it occurs at for the ISO 3200 image.
Thus the raw histogram, and the noise associated to each portion of it, will be half as much in raw levels for the "underexposed" ISO 1600 shot as it is for the ETTR ISO 3200 shot with the same aperture/shutter speed. Simply doubling all the pixel raw values of the ISO 1600 shot will yield a nearly identical image to the ISO 3200 shot; the fact that only the even raw levels are populated after this doubling of raw values is of no consequence, since as we have seen the dithering provided by the inherent noise of the image is more than sufficient to wash out the effect of the finer quantization of levels in the ISO 3200 shot."

To read more: http://theory.uchicago.edu/~ejm/pix/20d/tests/noise/noise-p3.html

TriX Loc: Raleigh, NC

An excellent reference and well worth reading, especially the section on "s/n and exposure decisions" which is extremely pertinent to this discussion and nicely speaks to the differences in optimum exposure techniques depending on the subject/situation and the brand of camera employed.

catchlight.. Loc: Wisconsin USA- Halden Norway

Thank you!... very good read that covers a lot of ground.

selmslie Loc: Fernandina Beach, FL, USA

Thanks for the link. It's clear that there is a lot more science behind this question than meets the eye (pun intended).

That's why we need visual evidence to establish whether ETTR actually provides real benefits we can see. Arguments and visual aids are simply not enough.

catchlight.. Loc: Wisconsin USA- Halden Norway

Looking forward to hear more after you read.

selmslie Loc: Fernandina Beach, FL, USA

It was certainly a challenging article. The main differences between this article and Michael Reichmann's are that it was published about seven years after his (2010 vs. 2003), it looks at the difference between 12- and 14-bit raw files and it considers Nikon sensors. It is also vastly more scholarly and detailed based on the author's obvious technical background. Although Reichmann was an avid photographer and writer I have seen no evidence of his technical background outside of the field of photography.

Several points are significant:

"Raw data is never posterized. ... the bit depth of the raw data is never the culprit. ... while the raw data is not posterized, post-processing can make it so" This is not going to happen in the brighter zones, only in the darker areas where the raw file represents tonality with the lower bit values.

"Curiously, most 14-bit cameras on the market (as of this writing) do not merit 14-bit recording." There has been a lot of progress since 2010. It is different today and some cameras are even using 16-bit encoding.

"Naively it would seem obvious that the highest quality image data would arise from concentrating the image histogram in the higher exposure zones, where the abundance of levels allows finer tonal transitions." Ouch! ETTR is a naive concept?

"Rather, the point is that by exposing to the right, one achieves a higher signal to noise ratio in the raw data. The number of available raw levels has little to do with the proper reason to expose right ..." If you can achieve a higher S/N ratio without ETTR, you can also reduce the visibility of noise.

"The proper reason to expose to the right comes from figure 12 on page 2, showing the rise in signal-to-noise ratio with increasing exposure. By increasing the number of photons captured, the S/N ratio improves ..." Ditto.

"... the preceding discussion might leave the impression that, for a fixed choice of the shutter speed and aperture, it doesn't matter whether one has underexposed at lower ISO or exposed to the right at higher ISO. In fact it typically does matter ..." My point in a nutshell.

"However, the benefit from the use of higher ISO comes in the shadows, not in the highlights where "there are more levels" " Now we get down to the crux of the matter. It is only in the shadows where there might be any visible benefit. If cameras have improved so much since 2010 that you cannot see a difference in the shadows, then ETTR is archaic, no longer necessary. It is also unnecessary if you don't care what the shadows look like - if they are just background to the overall composition.

There is a brief description of the benefits of "lossy" compression and how it throws off some of this mathematical analysis but the take away is that it is a misnomer. It may be an improvement over lossless compression.

a6k Loc: Detroit & Sanibel

I found this article helpful even though much of it is over my head. Perhaps some of you will also want to read it.
http://www.clarkvision.com/articles/does.pixel.size.matter/