Olympus-OM
[Top] [All Lists]

Re: [OM] TOP - DxO Gives OM-D High Marks

Subject: Re: [OM] TOP - DxO Gives OM-D High Marks
From: Ken Norton <ken@xxxxxxxxxxx>
Date: Thu, 4 Oct 2012 09:54:37 -0500
> I get that about film and filters.  But you said in a previous post in
> this thread that "raw is not raw."  Are you making the case that we have
> to shoot daylight WB and use filters with digital capture?
> And if so, how does CaptureOne then come into the picture.

What I am implying is that the camera itself is providing color
correction at the chip level to "normalize" the sensor to a
standardized white balance. They also do a bit of tweaking to align
the overall sensitivity to their normalized ISO settings. I won't get
into how they determine standardized ISOs, but they do micro-adjust
the gain on the sensor to bring it in line with not only WB
standardizations but speed standardizations too. As has been proven
through actual real-life usage, not everybody gets this right. My
Minolta A1 is a good example of this, where the ISO setting of 100
gives the exposure equivalent to ISO 160.

Onto the subject of using lens filters to color-correct a digital
camera file as compared to doing the correction in post, well, this
has been a subject of a lot of testing about ten years ago. What
happened is that convenience and optical purity won out. Let's take
the example of an indoor shoot under 3600K lights. You have a choice
of placing a blue filter on the lens or do the correction later. When
you use the blue filter, the noise in the blue channel stays lower,
but the noise in the green and red channels go up. If you do it in
post, the blue channel noise increases. However, one nasty side effect
of NOT using color correction filters in this case is that we run a
risk of blowing out the red channel (or even the green channels,
depending on camera) which causes clipping and other artifacts. If you
are shooting the entire project under 3600K lighting, and there is a
lot of dynamic range to the scene, you actually do get better results
if you use a CC filter. If you are under stage lighting, for example,
that is 3600K, the highlight fringing/transitions of any lights in the
scene are nasty in the digital files--especially if the lighting is
red-gelled. A blue CC filter allows you to balance the lighting so
that all three colors in the sensor will clip at or near the same
point. The overall noise-level will increase since you have to
increase exposure somehow, but the clip point and overall dynamic
range is improved. (As a side note, stage lighting, thanks to LEDs has
been shifting towards daylight white-balance and is no longer as warm
as it used to be).

So, let's run this to a landscape shot where we KNOW we're going to
warm up the image. Under cloudy conditions or right after sunset, the
overall lighting will skew to the blues. We warm the image up with
color correction via the white-balance adjustment. If we are
photographing a scene (the waterfalls on our trip) where we have the
bluish cast that needs correction and the scene itself contains
highlights and shadows that need to be preserved, we can end up
clipping one end or the other. A warming filter placed on the lens
will allow us to effectively increase the dynamic range of the scene
as captured on the sensor. Specifically to this, the shadows won't
pick up a blue noise cast and dithering artifacts as you lean into the
last two bits.

A third scenario, which is more unique to some cameras than others is
that when the coloring of the scene is far too intense in one of the
primary colors, bad things can happen. The DMC-L1 is a classic example
of this. When the coloring of the scene contains too intense of
oranges, it will just blow them right out. The E-1 isn't too far
behind. Let's say that you are photographing a red rose. To keep the
petals from blowing out, you have to underexpose the scene by
sometimes multiple stops. Then you bring the exposure back up during
post. If you don't do this, you end up with a bright red blob with no
details and Moose rightfully complaining about it. If, however, you
place a filter on the lens to de-red the scene a bit, you will bring
the exposure range of the reds down and increase the greens or
whatever else.

The idea which I'm presenting here is that we think of the sensor not
as just one sensor, but three sensors. You have a green sensor, a blue
sensor and a red sensor. Instead of exposing the picture based on the
combination of all three colors (or in monochrome), you adjust your
exposure so that none of the three color sensors is clipped on either
end. Since you can't really expose them separately, what you would do
is analyze the scene and determine which color sensor will clip and by
how much and offset that with a filter of the opposite color.

Overall, this is more academic than practical. Few of us travel around
with a set of CC filters. And in all honesty, it is a rare image which
would really benefit from this level of attention.

-- 
Ken Norton
ken@xxxxxxxxxxx
http://www.zone-10.com
-- 
_________________________________________________________________
Options: http://lists.thomasclausen.net/mailman/listinfo/olympus
Archives: http://lists.thomasclausen.net/mailman/private/olympus/
Themed Olympus Photo Exhibition: http://www.tope.nl/

<Prev in Thread] Current Thread [Next in Thread>
Sponsored by Tako
Impressum | Datenschutz