White balance M9

[xtoph]

Advertisement (gone after registration)

What I am responding to in these post is the assertion that WB has no effect when Shooting to RAW (DNG). If the assertion is correct there should be no difference no matter how WB is set. However thats not what I have found.

 

what you seem to be comparing are two raw files converted in the raw converter with different white balance settings. if you match the settings, there will be no difference between the two pictures.

 

when you set a mwb in camera, it simply appends that data to the unchanged dng file in place of the value the awb would have picked. then, whether you view the file on the camera lcd or through a raw converter, it will read the mwb setting and apply it.

 

you cannot view a raw file without applying some sort of wb. the question is, does the white balance setting actually <i>change</i> the raw data, or simply append the wb info to the correct metadata tag, as it should. i have yet to see any indication that the raw file is changed.

[swamiji]

what you seem to be comparing are two raw files converted in the raw converter with different white balance settings. if you match the settings, there will be no difference between the two pictures.

 

when you set a mwb in camera, it simply appends that data to the unchanged dng file in place of the value the awb would have picked. then, whether you view the file on the camera lcd or through a raw converter, it will read the mwb setting and apply it.

 

you cannot view a raw file without applying some sort of wb. the question is, does the white balance setting actually <i>change</i> the raw data, or simply append the wb info to the correct metadata tag, as it should. i have yet to see any indication that the raw file is changed.

 

Ok, How do I prove this one way, or another?

[xtoph]

Ok, How do I prove this one way, or another?

 

simple--in aperture, just copy the white balance (and any other tweaks, so the files are comparable) from the correct file and stamp it onto the awb file. they should look exactly the same.

 

i have actually done this myself just to be absolutely certain that this is how it works, and it does do the trick in my experiments.

[swamiji]

simple--in aperture, just copy the white balance (and any other tweaks, so the files are comparable) from the correct file and stamp it onto the awb file. they should look exactly the same.

 

i have actually done this myself just to be absolutely certain that this is how it works, and it does do the trick in my experiments.

 

They don't, the MWB has a temp 3062 and tint of 87, AWB has a temp of 3630 and a tint of 10. Changing the AWB to the same temp/tint as the MWB gives a blueish tint.

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[xtoph]

They don't, the MWB has a temp 3062 and tint of 87, AWB has a temp of 3630 and a tint of 10. Changing the AWB to the same temp/tint as the MWB gives a blueish tint.

 

obviously there's something you're doing that i don't follow. i have never seen anything remotely resembling what you are showing here. it is possible that what you're experiencing is related to the high shutter speed in combination with 'energy saving' bulbs, but it is suspicious that the mwb shot happened to coincide perfectly with the shot to which you applied it and not to the one before. it could be just dumb luck. at any rate, with any discontinuous light source, it is much safer from a color (and even exposure) standpoint to use a shutter speed close to the hertz cycle of your power source (with some systems, double that). otherwise, even when using a mwb, the color will vary wildly.

 

do you have a way to share the raw files?

[LondonM]

Energy saving bulbs are not a problem.

 

I shot 5 uncompressed DNGs with Tungsten, Flourescent 1, Daylight, cloudy and shady camera presets, imported them into Aperture 3.0. Of course they all had different colour casts (last 5 pictures). I then WB using eyedropper on the tissue sticking out of the kellenex box, which restored colour, but showed definite overexposure on the last 2 images. I then did auto-exposure correct and they got much better. There is still a hint of clipping on the last (shady) one right on the dome of the light (first 5 pictures).

 

Moral of the story: For Colour, the preset WB on the camera makes little difference. However when you import and WB correct, you can lose some dynamic range as the bright whites in the image saturate. In the last image, the RGB of the white tissue went from 238,165,55 to 199,199,199. Autoexposure cleaned the saturation above the bulb up up.

 

BTW, Can anyone tell me why the RGB vlaues are only to 255,255,255 ??? Isn't the sensor deeper than this ?

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[xtoph]

Advertisement (gone after registration)

Energy saving bulbs are not a problem.

[...] I then WB using eyedropper on the tissue sticking out of the kellenex box, which restored colour, but showed definite overexposure on the last 2 images. [...] For Colour, the preset WB on the camera makes little difference. However when you import and WB correct, you can lose some dynamic range as the bright whites in the image saturate.

 

can you clarify whether you were shooting in auto exposure, or at constant manual exposure settings for all 5 photos? also, did you standardize the wb on all photos in aperture, or did your custom eyedropper corrections give different values for each photo?

 

when i do this test using constant exposure, i can copy and paste the same wb to all shots and get exactly the same colors, no differences in clipping or otherwise.

[LondonM]

Shooting in autoexposure. Exposure lock (button half-press) on the tissue each time, as best I could. I already trashed those pics, so I can't go back to the metadata and check what exposures it picked. I can easily redo it though (tonight, after the sun goes down).

 

I bet you are right on the constant exposure thing.

 

I used custom corrections, i.e. eyedropper on the tissue in each of the 5 different frames with different camera WB.

 

I'm not sure what you mean by standardize WB on all photos. Actually, I think I do, but let me repeat the test before I put my foot in my mouth. I think you are saying the DNG is the same....no matter what. The colour cast you get in Aperture is because it reads the WB you shot with and applies that to the displayed image. If you define some other WB, then that is what would be used for display of the DNG. I think you are right on that and I will verify for my own satisfaction and report back.

[xtoph]

Shooting in autoexposure. Exposure lock (button half-press) on the tissue each time, as best I could. I already trashed those pics, so I can't go back to the metadata and check what exposures it picked. I can easily redo it though (tonight, after the sun goes down).

 

I bet you are right on the constant exposure thing.

 

I used custom corrections, i.e. eyedropper on the tissue in each of the 5 different frames with different camera WB.

 

I'm not sure what you mean by standardize WB on all photos. Actually, I think I do, but let me repeat the test before I put my foot in my mouth. I think you are saying the DNG is the same....no matter what. The colour cast you get in Aperture is because it reads the WB you shot with and applies that to the displayed image. If you define some other WB, then that is what would be used for display of the DNG. I think you are right on that and I will verify for my own satisfaction and report back.

 

by all means, try it again. by standardize the white balance i simple mean that for the same scene shot two different ways (eg, mwb and autowb), if i apply exactly the same numerical values to the white balance in aperture, the colors look exactly the same (provided i shot in manual, constant exposure).