Author Topic: AWB without shift  (Read 3513 times)

domasa

  • Member
  • ***
  • Posts: 112
AWB without shift
« on: February 27, 2016, 04:27:55 PM »
When I shoot a documentary video I sometimes mistakenly turn on "Adjusting white balance...".
This set wrong "White balance" and "WB Shift". When I switch to AWB by Canon menu (for fast fix) the "WB shift" is still wrong.

Is it possible:

1. turn on "Adjusting white balance..." by longer way for prevention
(e. g. by holding RATE button - similary as open ML menu by holding arrow button; or by double-click),

2. reset "WB shift" after set AWB?

domasa

  • Member
  • ***
  • Posts: 112
Re: AWB without shift
« Reply #1 on: March 03, 2018, 11:38:54 PM »
I modified question :)

Levas

  • Contributor
  • Hero Member
  • *****
  • Posts: 1741
  • 6d - Nightly build user
Re: AWB without shift
« Reply #2 on: March 04, 2018, 12:08:02 PM »
Had the same problem, it's because auto-white-balance is standard set as a shortcut key.
You can turn the whole auto white balance shortcut key option off.
By going into this menu: (Solved the problem for me)

domasa

  • Member
  • ***
  • Posts: 112
Re: AWB without shift
« Reply #3 on: March 04, 2018, 02:08:04 PM »
Had the same problem, it's because auto-white-balance is standard set as a shortcut key.
In your case AWB still has shortcut key: "Foc Pnts" in LiveView (6D).

In my case AWB has shortuct key: "RATE" (5DIII). It is good. But mostly I press this button by mistake.

a1ex

  • Administrator
  • Hero Member
  • *****
  • Posts: 12564
Re: AWB without shift
« Reply #4 on: March 04, 2018, 05:39:54 PM »
This should be re-implemented from raw-based data anyway. The way it works right now is by looking at the YUV buffer and performing an iterative procedure using Kelvin/GM adjustments until it gets this right.

There are routines to do the math (kelvin.c extracted from UFRaw); the custom WB multipliers from Canon may be a better choice for the auto adjustment backend and won't require WBShift adjustments, but their exact meaning is not very well understood (they don't seem to be applied directly to the raw Bayer channels).