Hi David,
the information I have is that manufacturers set the exposure balance
to slightly underexpose, with most doing this to -0.3 EV. However,
photographers shooting in RAW, who don't care about the SOOC image, will
push the histogram as far to the right as possible to avoid clipping and to
reduce noise. I tend to be a bit conservative on this as when the histogram
nears clipping the saturation is reduced. Aurelien explains this in some of
his videos.
When I teach my photography students about histograms I tell them that if
they shoot in JPG (SOOC), then the histogram only confirms the bleeding
obvious that the picture is under or over exposed. However, I then show an
image and histogram of a picture taken in the middle of an Indian desert of
a fort with lots of haze. The picture looks washed out and over exposed. I
ask the students to put their hands up if they feel the image is
underexposed. Most students do not put their hands up, I can not remember
any student putting their hand up. But, then I raise my hand and say yes it
is. I then explain that the histogram is placed very central with plenty of
room to move to the right. There is no clipping and the histogram is very
narrow due to the atmospheric haze. I open the image in editing software
and use options such as levels to stretch the histogram wider and then
adjust the gamma and show how well the image can be processed to lift the
contrast and fix the exposure.
I might agree with you that I give camera manufacturers too much credit
that they want to protect the highlights. But, one of the challenges we
face is having people realise that the SOOC image is not necessarily how
the image should look or how it can best look. It is just the
manufacturer's algorithm producing a version of the scene.
Cheers
On Tue, 18 Jan 2022 at 13:29, David Vincent-Jones <david...@gmail.com>
wrote:
> I am not sure that camera manufacturers have too much interest in
> protecting the highlights. However you set your meter .. for spot, local or
> whole image .. the meter (if you are using it solely) is simply going to
> provide information assuming that the metered area is pivoted around 'mid
> gray' (18% black). Whether the histogram (data range) is broad or narrow,
> and the position of the extremes, is totally another problem.
> On 2022-01-17 21:18, Terry Pinfold wrote:
>
> Hi All,
> camera manufacturers actually set the camera to underexpose a little
> bit because this protects the highlights and produces a pleasing result.
> However, when shooting in RAW you want to push the exposure towards the
> right of the histogram without clipping. The picture may look over exposed
> bt if the highlights are not clipped then DT can correct the exposure and
> the noise is reduced compared to the SOOC JPG. It is a little more effort
> but gives the best picture.
>
> On Mon, 17 Jan 2022 at 21:14, Frank <synapse...@gmail.com>
> <synapse...@gmail.com> wrote:
>
>
> On Thu, 13 Jan 2022 at 01:42, bwh vanBeest <b...@xs4all.nl> <b...@xs4all.nl>
> wrote:
>
>
> Thanks for all the reactions to my posting. I initially thought something
> was wrong with my camera-DT-combination, but now I learned otherwise. I
> have now learned what to do. Still, somehow it is unsatisfactory that a
> picture of which you know is exposed correctly is initially displayed in DT
> as underexposed.
>
> Regards
> Bertwim
>
> What Darktable is actually showing you is that your images are underexposed,
> and you can expose farther to the right and still have highlights that are
> not blown. You will come to appreciate this if you have the patience. You
> cannot trust your in camera histogram for exposure, this is a fact.
>
>
> I have edited the exposure preset to apply +2 stops on import, this is about
> right for my Olympus cameras
>
>
> ____________________________________________________________________________
> darktable user mailing list to unsubscribe send a mail
> todarktable-user+unsubscr...@lists.darktable.org
>
> --
>
> ____________________________________________________________________________
> darktable user mailing list
> to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
>
>
> ____________________________________________________________________________
> darktable user mailing list to unsubscribe send a mail to
> darktable-user+unsubscr...@lists.darktable.org
>
--
____________________________________________________________________________
darktable user mailing list
to unsubscribe send a mail to darktable-user+unsubscr...@lists.darktable.org
