From: Bob Larter on
Porte Rouge wrote:
>> You are aware that ETTR is a two part process, part one
>> being boost the exposure to maximum possible, and part
>> two is to later reduce it to whatever it is that makes
>> your heart beat fast enough.
>>
>> You *have* to do both.
>>
>
> Yes, I understand. My question is, after ETTR what exactly do you
> yourself do in post processing. Do you adjust exposure, black point,
> or something else?

I adjust black point, which retains the highlights.


--
W
. | ,. w , "Some people are alive only because
\|/ \|/ it is illegal to kill them." Perna condita delenda est
---^----^---------------------------------------------------------------
From: David J Taylor on

"Bob Larter" <bobbylarter(a)gmail.com> wrote in message
news:4ad6a1cf$1(a)dnews.tpgi.com.au...
[]
> As should be obvious from a moments thought, you're going to get a lot
> more dynamic range/tonal resolution from a 12 bit RAW file than from an
> 8 bit JPEG.

Yes, while the gamma-corrected JPEG may have an even greater dynamic range
between darkest grey and lightest white, it does not have the same tonal
resolution as the raw file. The raw file will likely have more headroom
as well.

Cheers,
David

From: Floyd L. Davidson on
Bob Larter <bobbylarter(a)gmail.com> wrote:
>As should be obvious from a moments thought, you're going to get a lot
>more dynamic range/tonal resolution from a 12 bit RAW file than from an
>8 bit JPEG.

Actually... not "a lot". About 1 fstop.

The maximum possible dynamic range of a 12 bit linear
encoded raw data file is fairly easy to determine for
just about any definition of dynamic range. But for an
8 bit gamma corrected JPEG encoding it is not so easy.
By one definition the JPEG actually has more dynamic
range (not for practical definitions though, because
most of the extended range is quantization distortion).

For practical purposes, a "useful dynamic range" is
limited to where each fstop is represented by at least 8
tone values in the encoding because with fewer values
quantization noise is visible as posterization.

Linear encoding, compared to an 8 bit gamma corrected
coding from a JPEG, looks like this:

Exposure Linear Encoding Bits 2.2 Gamma Corrected
Zone 14 12 10 8 8 bits

1 8192 2048 512 128 69
2 4096 1024 256 64 50
3 2048 512 128 32 37
4 1024 256 64 16 27
5 512 128 32 8 20
6 256 64 16 4 14
7 128 32 8 2 10
8 64 16 4 1 8
9 32 8 2 6
10 16 4 1 4
11 8 2 3
12 4 1
13 2
14 1

The useful dynamic range for linear encoding is
therefore 11 fstops for 14 bit depth, 9 for 12 bit
depth, and 7 for 10 bit depth.

The useful dynamic range for a JPEG, using 8 bit 2.2
Gamma correction, is 8 fstops. (It actually does
continue to wiggle bits down to about 18 or 19 zones,
but there is no useful information.)

--
Floyd L. Davidson <http://www.apaflo.com/floyd_davidson>
Ukpeagvik (Barrow, Alaska) floyd(a)apaflo.com
From: Porte Rouge on
On Oct 14, 11:56 pm, fl...(a)apaflo.com (Floyd L. Davidson) wrote:
> Porte Rouge <porterouge...(a)gmail.com> wrote:
> >> So you take a wild guess at 1 or 2 stops more!  Even
> >> with film and no histograms most accomplished
> >> photographers attempted to get within 1/2 a stop.
> >> Today, with digital and histograms, it is entirely
> >> possible to be within 1/3rd of a stop.
>
> >Floyd, baby, chill out. "Wild guess" is over the top. I'm a slow
>
> It's a wild guess.
>
> --
> Floyd L. Davidson <http://www.apaflo.com/floyd_davidson>
> Ukpeagvik (Barrow, Alaska)              fl...(a)apaflo.com

Oh, well, that's too bad. I guess that's as far as we go, Floyd.
Thanks for answering my questions.

Porte
From: Wilba on
Worth a look for a discussion of 13 images and their histograms -
http://www.juzaphoto.com/eng/articles/exposure_and_histogram.htm