RAW vs White balance

Started Nov 20, 2006 | Discussions thread
Jonas B Forum Pro • Posts: 14,596
Re: Unequivocal evidence - here too now

GordonBGood wrote:

Jonas B wrote:

What does it all matter with regards to real photos? That is the
important question to me. Sometimes I get very good colors,
sometimes not. As I have my own profile for ACR it works fine. I
have the white balance set to sunshine all the time except for some
times when I have choosen to make a manual balance. I never thought
of the possibly need for another profile at those occassions. Then
I sometimes got colors a bit off but I think I just blamed the
light for this.

I can't see that most raw convertor's that give an accurate as shot
white balance wouldn't take the varying gains in the RGB channels
into account, as otherwise one would have to do a manual WB for
every image. Usually raw files record the actual relative gains of
the individual RGB channels in order to get a neutral grey or
white. The the case where the pre-amplifiers can be set with
various gains, the recorded required extra gain would be (1,1,1).

So what are we thinking now? With the whibal set to manual the amplification needed should be close to (1,1,1)? When it is set to Tungsten the correct values might be (1,1.2,0.9)? I'm just making up the figures of course.

If the Pentax software has been aware of this all the time maybe that explains the good colors people have reported getting from it (never even installed it myself)? And in what way do the other raw converters handle this? I remember that Adobe changed the DS profile in ACR when updating to v3 (I think it was then at least). But noone ever commented on different profiles depending on WB settings.

A big Hmmm here...!! What on earth says that I get the correct profile when manually dragging the raw converter WB slider?

I think that the main thing learned is that there is an advantage
to guessing at the proper WB preset or use Custom WB where the
white point is very warm or very cool in order to help avoid
clipping the red or blue channels, respectively.

Perhaps the times when you get less accurate colours with your
standard workflow is the times when one of the raw colour channels
is clipping due to extreme WB situations? If you previously just
left WB to Sunny, you must have done a manual WB for evey image, or
let you convertor do an automatic (redetermined) WB? If the second
case, clipped raw channels might affect the determination.

I always have done it manually for every picture. Sometimes I use the Photoshop plug-in InCamera from Pictocolor, but not very often. It's only ocassionally I bring a Gretag ColorChecker chart with me. A grey card is always in the bag though, next to the notebook.

As pictures files in raw format have to be converted to something
else to let us look at them the raw converter is very important.
This make for another tedious set of tests: We don't know if ACR,
Capture One and Silkypix takes this into account. Heck, we even
don't know if the Pentax software handles the files correctly.

As I said, it would be hard to believe that Pentax software
wouldn't handle the WB correctly, and newer versions of Silkypix
likely do given the co-operation between them and Pentax, but other
convertors would have to be tested to see how the convertor WB
compares to that of the embedded JPEG.

Again, thanks for all your work in showing that this is real in the
'D' series and likely in the K100D.

We have to thank vortout and Curtis here. Or curse them...

For now I just neglect all this, hoping somebody will show any possible effect to real world pictures. Maybe the difference is small enough to just make me go from Sunshine to Tungsten and make a new ACR calibration.

regards,

Jonas

Post (hide subjects) Posted by
w
w
w
ddd
Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow