Flash Light Temperature

[ksinkel]

I enlarged a calibration target I photographed and I can clearly see variations in color temperature between different small regions within a gray patch. So if your target is like mine, when you click with the probe, you are selecting areas of somewhat different temperature each time you click and so getting different results.

Kiril

[tomczak]

Thank you - the large amount/level of noise reduction seem to make the WB much more consistent. All in all the actual light temperature of the flash (a build-in, compact flash) as I figured doing the test above seems to be around 6500K, this is quite different from both: the PWP RAW flash preset and in-camera flash preset.

p.s. Will chroma noise reduction slider in RAW be an option in the PWP6?

[tomczak]

I noticed that as PWP RAW recalculates the preview, the Noise Reduction (if it's used) is applied twice: once for the WB and once afterward. Is that how it works? If it is, would it be possible and would it make sense to always apply the NR for the white balance (regardless if NR is actually used or not for the image)? A related question: since I believe that NR preview is fast because it processes just the subset of pixels required for the preview, could similar technique work for the white balance blurring (i.e. only blurring a kernel? of pixels needed for white balance?

[tomczak]

When PWP RAW uses WB=Camera WB, where does it read the value(s) from? An EXIF/MakerNotes? I can't match the values of Colour Temp and Tint with anything that I can see in the EXIF.

[ksinkel]

White balance is calculated from data extracted by dcraw and is not simple. It is calculated from camera multiplier and image white balance data which are expressed as three or four element vectors, depending upon whether the camera has three or four different sensor pigments. This sets both the temperature and tint controls to their initial values when you choose camera WB.

Kiril

[tomczak]

There are channel multiplier values listed in EXIF/MakerNotes. Is that where dcraw reads them from to, then, derive Colour Temp and Tint values?

[ksinkel]

Maker notes are different for each camera manufacturer and even camera model. So dcraw may use different methods for different cameras.

Kiril

[tomczak]

I removed EXIF:MakerNotes from a .CR2 file to see what will happen. It looks like PWP RAW defaults to 'no WB' when it can't read the WB data, but still reports 'Camera WB'.