EM-5: Confused by Auto White-Balance results

Started Feb 12, 2013 | Discussions thread
Timur Born
Senior MemberPosts: 3,894
Like?
Re: EM-5: Confused by Auto White-Balance results
In reply to micksh6, Feb 12, 2013

micksh6 wrote:

There can't be pure daylight indoors, unless all objects inside completely lack colors. Part of the light is always reflected.

Of course, what I meant to say is that there was only one light-source, which was non artificial and thus full spectrum.

Another thing that may explain differences in the first shots - color response curves may vary for different ISO sensitivity. The second shot was made at ISO12800. At that ISO dynamic range and color depth are too low. Simply speaking, there aren't many colors left to choose from.

Yes, the extra high ISO was something that may have an effect. On the other hand I did WB tests with the X10 at different ISO settings once and the main "overall color appearance" related difference seemed to be from noise in shadows (black appearing more and more blue). Once I find time I will have a look on that with my gray/white card.

BTW, how much of a difference should a gray card make versus a white card? I understand that the 18% gray card is meant for helping with exposure, while the white card is meant to help with white-balance. 18% gray is nothing but darkened white anyway and thus should ideally result in the same outcome. But skipping over the different WB methods listed in the PDF linked by DM it may make a difference with some cameras and maybe so with some raw converters!?

Something to test in free time I guess.

Now, you were saying that your halogen lights are rated at 2900-3000K. It's important to realize that this may not be the case after the lights were mounted in your particular room. As I tried to explain in my previous post the light spectrum will change after it's reflected from walls, ceiling, clothes and other objects that aren't neutral. The light mix will change after you change camera position resulting different WB value. This should explain inconsistency in Auto WB results.

Yes, but the whole circumstances with all the reflections and imperfections don't change too much from one shot to the other, so I would expect WB measurements to come up with results that are closer to each other instead of being worlds apart. The first example is somewhat close, but the second example really is worlds apart. It's the same skin, the same purple bathtub and the same walls reflecting the light. What changed was the additional of TTL fill-flash bounced towards the same ceiling that the halogen light came from (even at -1 EV flash power).

So I am really still baffled at the difference of 2400K vs. 5100K even while knowing that mixed light is difficult. And my X10 examples show that "usable" Auto WB is still possible even in the most wild mixed light situations. I don't consider the E-M5's bathtub results usable, or at least I don't remember my baby suffering from second degree burns on her skin from the bath.

Obviously I never trusted Auto WB and my own skills too much, else I wouldn't have to use RAW files.

Some months ago I got me some cloth foldout white/graycard for outdoor shooting in the garden where I expected lots of green and changes of very bright sunlight and dark shadows to cause problems. Plus it's easy to put them against some plant without asking kids and other people to hold still (or at least its more fun to throw it to my boy like a Frisbee and ask him to hold it for a sec).

I used that for my comparison of how different raw converters handle "as shot", auto and WB picker, which can be seen in the table I posted earlier.



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