[arkav]

Since getting my K30, I've mostly been using the camera JPGs because I haven't been entirely happy with the results I was getting from any of the RAW converters I had available, but
lately I've been working a bit harder to get proficient with them. All along, I've been mostly shooting RAW+, so I'd have the RAW files available when I settled on a preferred RAW converter.

Yesterday, I was out shooting under bright winter sunlight, with WB set to "Daylight" on my K30. I worked the RAW files using Silkypix 3.0 that came bundled with the camera.

Most of the shots come up with the same WB temperature and tint ( called "colour deflection" in Silkypix ) which is what I would expect, but a few shots have different values for temperature and tint.

This makes no sense to me if the camera is setting WB to fixed values for "Daylight".

To double check this, I picked 2 photos. I verified on the camera that both photos were shot with WB set to daylight. Then I opened both photos in 3 different converters.
And yes, I verified that each converter was showing the "camera setting" for WB, not a custom/altered value. As stated above, Silkypix shows the two photos with different WB values.

I tried opening the same files in ACR ( via PSE 9.0 ). Again, I see a discrepancy in temperature/tint between the two photos, but the values displayed are not the same as those shown by Silkypix. PSE 9.0
predates the K30, but the files are in DNG format, and PSE says it's using the embedded camera profile. Maybe Tint doesn't really mean the same thing as "colour deflection", but temperature should
be the same. They're not wildly different, just slightly.

In Rawtherapee, I also see a difference in the WB settings between the two photos. And again, the colour temperature values don't match those shown by the other 2 converters. In this case, Rawtherapee
presents tint differently ( not a +/- integer, but a value relative to 1.0 ), so I don't expect the tint value to match what the other two converters present.

Here are the temperature values shown for the two photos by each converter:

Silkypix: 5244, 4905
PSE 9.0 ( ACR ):5250, 4900
Rawtherapee: 5337, 5024

It looks as though ACR is probably just rounding the temperature to the closest multiple of 50. Not sure what's going on with Rawtherapee - maybe it's presenting temperature in fahrenheit...

Ignoring the differences in values between the converters, they all show different temperatures for the two photos that were shot with the same WB setting on the camera. If both photos were shot with the K30 set to "Daylight", shouldn't they both have the same WB values in the DNG file?

Have I completely misunderstood how WB works?

[max_pyne]

I'm not an expert in shooting raw but from what i had been reading is that the RAW in Pentax case PEF or DNG is not including a WB setting it should be the information straight from the sensor.
And this is the whole clue of shooting in raw you don't have to bother about it and can adjust it in your raw processor in post-processing without loosing any information contained in the RAW image container...

Edit:
The image you usally see is embedded int he RAW file and is only for displaying purposes if you want to compare pictures you can for instance convert them in to a JPEG or TIFF using the same settings in your RAW converters. Different RAW converter can produce slightly different JPEG output from my experience.

[Ole]

There are three cases were fixed WB isn't fixed:
1 - Auto and Scene modes use AWB no matter what
2 - When using fill flash. Custom function 10 by default allows the camera to adjust WB
3 - WB tungsten light setting (not relevant in your case), custom function 11 allows the camera to adjust WB

Other than that I would expect a fixed value for temperature and tint except that the K-30 manual states that WB Daylight means approx. 5200K! Whatever that's supposed to imply.

[BrianR]

Cameras will do some fine tuning to the WB even in the preset modes. There's a setting on the k-5/3 cameras called "WB Adjustable Range" that lets you disable this fine tuning for the consistent results you're expecting. I don't see any such option in the k-50's manual.

[arkav]

There are three cases were fixed WB isn't fixed:
1 - Auto and Scene modes use AWB no matter what
2 - When using fill flash. Custom function 10 by default allows the camera to adjust WB
3 - WB tungsten light setting (not relevant in your case), custom function 11 allows the camera to adjust WB

Other than that I would expect a fixed value for temperature and tint except that the K-30 manual states that WB Daylight means approx. 5200K! Whatever that's supposed to imply.

Right, I meant to mention that I was shooting in Av mode, no flash. Normally, I would have shot manual, but I wanted to check auto metering with this lens that I had just had serviced due to
wonky exposures. I almost never use auto or scene modes.

If the camera is adjusting the fixed WB setting, the fact that most of the shots have WB set to ~4900 instead of the advertised 5200K probably has something to do with the light conditions - very bright afternoon sun on snowscapes - trees coated in ice and snow. Lots of bright white and blue. WB seemed to toggle between ~4900K and 5250 depending on the direction I was facing ( IIRC, I'll double check if it was consistent ).

Actually, I did shoot a series of shots for panorama stitching in manual mode. I know that I forced them all to have the same WB settings when I converted them, but I don't remember if there
were discrepancies in the camera settings. If the camera is tweaking the WB between shots, that could make stitching a series of camera JPGs awkward if you're thinking that the WB is constant because
you've set it to 'daylight'.

Cameras will do some fine tuning to the WB even in the preset modes. There's a setting on the k-5/3 cameras called "WB Adjustable Range" that lets you disable this fine tuning for the consistent results you're expecting. I don't see any such option in the k-50's manual.

I don't recall ever seeing a WB adjustable range setting in the custom menus, so that is probably not a feature of the K30.

Thanks to both of you for this information. As long as I know that the camera is doing this, and it isn't something weird in the RAW converter(s), that's fine. In some circumstances, I can see it making sense
for the camera to second guess my settings.

It's probably time for me to get a gray card anyways so I can set manual WB. Hopefully, the camera doesn't adjust it when you're on manual WB!

Cheers!

[Bruce Clark]

AS BrianR points out the "WB Adjustable" feature can change the so called fixed WB setting even in RAW. From my limited experience I found this prevalent when trying to get a photo of essentially a monotone scene. A course I was doing required a photo of a gray card for test purposes. Impossible with this setting on. The camera seemed not to believe what it saw and attempted to change reality and introduce false colours noise etc. In varied scenes however I have not noticed a great deal of difference. Nevertheless, I leave mine off (K7). If this setting is not available in your camera, I am afraid this is all a bit academic but it may at least provide some explanation for what you are seeing.

[arkav]

Nevertheless, I leave mine off (K7). If this setting is not available in your camera, I am afraid this is all a bit academic but it may at least provide some explanation for what you are seeing.

That's torn it. From now on, only flagship models for me.

[IchabodCrane]

Since getting my K30, I've mostly been using the camera JPGs because I haven't been entirely happy with the results I was getting from any of the RAW converters I had available, but
lately I've been working a bit harder to get proficient with them. All along, I've been mostly shooting RAW+, so I'd have the RAW files available when I settled on a preferred RAW converter.

Yesterday, I was out shooting under bright winter sunlight, with WB set to "Daylight" on my K30. I worked the RAW files using Silkypix 3.0 that came bundled with the camera.

Most of the shots come up with the same WB temperature and tint ( called "colour deflection" in Silkypix ) which is what I would expect, but a few shots have different values for temperature and tint.

This makes no sense to me if the camera is setting WB to fixed values for "Daylight".

To double check this, I picked 2 photos. I verified on the camera that both photos were shot with WB set to daylight. Then I opened both photos in 3 different converters.
And yes, I verified that each converter was showing the "camera setting" for WB, not a custom/altered value. As stated above, Silkypix shows the two photos with different WB values.

I tried opening the same files in ACR ( via PSE 9.0 ). Again, I see a discrepancy in temperature/tint between the two photos, but the values displayed are not the same as those shown by Silkypix. PSE 9.0
predates the K30, but the files are in DNG format, and PSE says it's using the embedded camera profile. Maybe Tint doesn't really mean the same thing as "colour deflection", but temperature should
be the same. They're not wildly different, just slightly.

In Rawtherapee, I also see a difference in the WB settings between the two photos. And again, the colour temperature values don't match those shown by the other 2 converters. In this case, Rawtherapee
presents tint differently ( not a +/- integer, but a value relative to 1.0 ), so I don't expect the tint value to match what the other two converters present.

Here are the temperature values shown for the two photos by each converter:

Silkypix: 5244, 4905
PSE 9.0 ( ACR ):5250, 4900
Rawtherapee: 5337, 5024

It looks as though ACR is probably just rounding the temperature to the closest multiple of 50. Not sure what's going on with Rawtherapee - maybe it's presenting temperature in fahrenheit...

Ignoring the differences in values between the converters, they all show different temperatures for the two photos that were shot with the same WB setting on the camera. If both photos were shot with the K30 set to "Daylight", shouldn't they both have the same WB values in the DNG file?

Have I completely misunderstood how WB works?

For comparison purposes, you might want to try the experiment again (as close as possible lighting as your first experiment) with AWB. Would be interesting to see how the converters report WB temperature for these.

[arkav]

For comparison purposes, you might want to try the experiment again (as close as possible lighting as your first experiment) with AWB. Would be interesting to see how the converters report WB temperature for these.

Ah well, too bad I didn't think of that myself or read your post earlier. I was out skiing in the same area today, and it was similarly sunny, and I had the camera with me. But all the shots I took were with WB set to daylight. I will check to see if the camera decided "daylight" needed to be set to a different temperature today.

[Ploki]

Some RAW converters interpret embedded WB information better than others.

If you take the same photo, open it in Lightroom 5 and Aperture 3 and just save it, it will look slightly different

I shoot RAW exclusively, and if I'm not too lazy I use custom WB, else I just don't bother and fixed it in Aperture.

[arkav]

Some RAW converters interpret embedded WB information better than others.

I get that. But I would have thought that a numeric value for temperature would be the same across different converters. Unless the camera isn't actually embedding a numeric colour temperature value, and that value is something that the converter computes from some other value(s) embedded in the RAW file.

I recently installed DxO. I'm just looking at some photos shot last summer. I've opened one particular photo in Silkypix, RawTherapee, and DxO. They all show a different value for WB temperature recorded by the Camera.

So we've got two confusing phenomena:

1) The camera will sometimes vary the WB regardless of the fact that it is set on a fixed WB setting
2) Different converters display a different value for colour temperture for the same RAW file

If you aren't aware of either of these effects, it can seem like the camera is very inconsistent in recording WB values. The fact that I can override the camera WB setting in the RAW converter is beside the point. That's obvious, since I'm playing around with several different RAW converters at the same time.