> It's becoming obvious that the trick to using glass filters is to
> record the image as RAW colour, convert it to JPEG colour, and then B&W. Or
> something like that. I find that doing this processing with the camera is a
> bit cumbersome as I have a lim ited opportunity to judge the saturation, etc.
> before doing the conversions.
It gets a little more complicated. Yes, LR is a good option, but the
problem comes in that there is an inherent flaw in the way nearly all
raw converters combine the three or four pixels together to get a
merged RGB value for the pixel. When one set of sensels has a full
luminance range of values, but the other sensels don't there is an
averaging that kicks in which screws everything up.
Frankly, you're usually better off just shooting it unfiltered and
"fix er in post". The exception to this would be if you are using
RawDigger to extract just the one channel. That worked pretty good for
me. But resolution and noise goes very wonky.
Let's say you place a red filter on the lens and take the photograph.
Just for illustration (cause there is some bleed-over, but not enough
to make this wrong), lets say that RawDigger reveals a nice histogram
profile on the red sensels, with nothing of note on the blue and green
sensels. If you retrieve JUST the red sensels at this point and run a
monochrome TIFF conversion, you would end up with a file based on just
the red sensels and the 50% loss in resolution.
However, other than RawDigger and a couple other one-offs, there is no
way to retrieve just the red sensels. Soooooooo, let's look at what
Lightroom does.
1. If Olympus, the ACR converter engine uses a four sensel (pixel)
merge. This is essentially an averaging process that determines
luminance value as well as color value for any given output pixel.
(with some other cameras, the merge uses only three sensels in the
mix).
2. Since the Blue and Green sensels in the merge have a value of "0",
these are averaged in with the Red sensel value. Depending on the
algorithm and a few other settings that introduce some fuzzy logic
into the mix, the output luminance value for the pixel MIGHT be what
the input value from the Red sensel is, but is more likely an average
value of the four used in the mix.
3. The resulting output image is then low in contrast because of the averaging.
I know Chuck is going to cry fowl here and say that the value of the
red sensel is applied to the red channel and the output of the green
sensels are applied to the green channel and the value of the blue
sensel is applied to the blue channel of the given output pixel. The
answer to that is absolutely not. The RGB values applied to any given
output pixel are POST mix. The mix happens first then the values are
applied. That's why you see reds turn to yellow, etc. The ONLY, and I
mean ONLY way to extract just the value of the red sensels is with a
program like RawDigger which lets you do just that. Lightroom/ACR
absolutely does not. I can guarantee you that ACR is mixing first for
luminance and applying color values second, but not without goobering
things up because it actually generates false colors as part of the
luminance merge.
I know I explained this very poorly. The upshot is that the ugliness
is the result of the raw conversion stage.
--
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/
|