> So you say we can have a mask between two values right? When you open an
> image in the darkroom module it has a default WB value and EV (i'm guessing
> 0EV). If you enable a mask and change the EV or WB value we could apply the
> mask between the new value and the base value. Am i wrong on this?
>
>
>
Ok, seeing Hanato's reply below I was wrong about exposure. So I won't
speak about exposure and only talk about WB here.
First a quick reminder/crash course about what WB is... that will make
discussing the specifics easier later.
A digital pixel is color-blind. It only measures a number of photon hitting
it
To get color images, we add a filter (red, green or blue) on top of the
pixels so each pixel can only see photons of a specific color. During the
RAW developement process, we blend each pixel with its (differently
colored) neighbours to reconstruct the image. That's demosaicing.
However pixels do not respond in the same way to every color. In particular
they are not very good at seein the green color.
To compensate for that, our sensors have more geen pixels than red or blue
pixels. A typical sensor (all the ones supported by DT) have the following
color pattern
RGRGRGRG
GBGBGBGB
RGRGRGRG
GBGBGBGB
As you can see, there are twice as many green pixels than red or blue.
If we treat the image in a simple way, we would have a very strong color
cast. To compensate for that, we multiply each type of pixel with a float
very early in the pipe. That's white balance
The typical R,G,B coeffs are close to (1, 0.5, 1) but the exact values
depend a lot on the brand/model of sensor and the brand/model of the filter
on top.
If you want to have "no" WB, there is a preset for that in DT, you can try
it for fun, but it's unusable. The green cast is just too strong.
Now, back to your question. Internally, BlendIf and Masks are "merging" i.e
we take the image before, we take the image after and we merge the results.
That can't work for WB because
* we are so early in the pipe (before demosaic) that the idea of an image
"before" doesn't make sense, so we have nothing to merge with
* Applying WB multiple time is a bad idea. Whe you multiple/divide multiple
times in a row you loose a lot of image quality compared to multiplying
once with the agregated value
So, we could have (UI-wise) multiple instance of WB, the first one with
blendif/maks disable and internally merging these instances to have
multiple variable WB coeffs resulting from the agregation of the different
filters in different areas, but from a point code of view, this is very
complicated because we are not reusing blendif/mask code at all. We are
reimplementing it in a WB specific way. So much more work
hanatos, houz, whoever knows better : feel free to correct me if I said
anything stupid
------------------------------------------------------------------------------
Get your SQL database under version control now!
Version control is standard for application code, but databases havent
caught up. So what steps can you take to put your SQL databases under
version control? Why should you start doing it? Read more to find out.
http://pubads.g.doubleclick.net/gampad/clk?id=49501711&iu=/4140/ostg.clktrk
_______________________________________________
Darktable-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/darktable-users
