I do HDR shots and processing. The tone mapping processes themselves
tend to produce halos and shadows around parts of images in my experience.
On 2/26/21 11:27 PM, [email protected] wrote:
" I know at least /.../ darktable can output in linear color space
without applying a gamma curve "
Okay, so part of the solution is to go into Darktable and disable the
base curve. This worked, though it's too bad it adds another step to the
workflow. Choosing darktable option from hugin import doesn't give any
options to specify, like styles that could be used to define linear
settings with... does it work well for you? I'd try it but for now it
just crashes.
Good luck with Darktable. I've never been able to figure out its user
interface. I use RawTherapee to process my RAW files.
I'm pretty sure all programs that process RAW files have a way to set
and save preferences, so I'd think you could set a preference to not
apply a color space?
"The response curve is applied to all color channels. /.../ Maybe there
is a problem with the white balance instead."
Camera manufacturers add subtle changes to the RGB output channels
in-camera. It's mostly done to keep a cohesive look to their cameras,
like a modern photorapher may use a certain LUT they like so all of
their work looks the same. Normally raw files are immune to this because
there is no in-camera processing going on, but point here is to check
for inconsistencies. It could be anything - sensor age or wear.
Camera manufacturers could also be compensating for issues in the
sensors themselves.
I've also seen sensor tests (I think at DXOMark?) indicating that
sensors' color response (in bit depth) changes over the ISO range.
Generally, the higher the ISO, the lower the bit depth.
I always shoot at 100ISO, so I don't know what impact shooting at higher
ISOs might have.
"Hmm, if you opt for linear color, does color space matter?"
Technically those are two different things. A colour space can be
whichever coordinate system you can describe within the limits of human
visual ability (400-700nm), while linearity is having a constant step
"distance" within that coordinate system. I imagine.
I tend to think of linearity as an absolute number recording the
sensor's response to whatever light frequencies it's sensitive to. For
instance, some digital camera sensors are sensitive to the infrared
ranges used by some autonomous/assisted car systems, and their sensors
have been damaged by such cars' laser-generated beams.
Color space, yeah, that's different.
"What impact would [15bit colour depth] that have on "linear" color space?"
More colours between full black (0,0,0) and full white (1,1,1). If the
system is new or very uncommon it could cause demosaicing problems, but
that is a different topic.
Color depth is just the intensity range as detected by the particular
sensor. Some sensors aren't an even color depth across the three colors.
Some cameras truncate their sensor readings before recording them as a
RAW file. My old camera's sensor recorded 16-bit color but cut off the
high bit in the RAW file. I think some other cameras use 15-bit sensors
but truncate the colors down to 12-bit when recording. I think it's a
function of how much noise removal the camera manufacturer wants in
their camera?
On Friday, February 26, 2021 at 6:44:07 AM UTC+1 GnomeNomad wrote:
On 2/25/21 7:49 AM, T. Modes wrote:
>
>
> [email protected] schrieb am Dienstag, 23. Februar 2021 um
21:53:38 UTC+1:
>
> I want to check if the photo stacks I am working with are linear
or not.
> Yes; raw files are supposed to be linear, but Hugin does not work
> with CR2, only tiff. If I convert the cr2 to tiff it will probably
> add a gamma curve,
>
> I know at least dcraw, RawTherapee and darktable can output in
linear
> color space without applying a gamma curve. What Adobe does I
don't know.
Hmm, if you opt for linear color, does color space matter?
Also, I learned something when I changed cameras. I changed from a
camera whose sensor shot 15-bit per channel (each color) to one that I
understand shoots different bit depths per channel. What impact would
that have on "linear" color space?
> In the attached picture green shift is seen in the cloud. To my
> knowledge this means that the gamma curve has been calculated badly.
>
> Not sure about this. The response curve is applied to all color
> channels. When there is a color shift it is probably not related
to the
> response curve. Maybe there is a problem with the white balance
instead.
What happens if the sensor's response curve varies for each color
channel?
--
David W. Jones
[email protected]
wandering the landscape of god
http://dancingtreefrog.com
My password is the last 8 digits of π.
--
A list of frequently asked questions is available at:
http://wiki.panotools.org/Hugin_FAQ
---
You received this message because you are subscribed to the Google Groups "hugin and other free panoramic software" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/hugin-ptx/81879039-b22f-47b5-11d6-c0b951b6afff%40gmail.com.
