MAWB on the K3

Can anyone point me at a description/explanation of how the 'Multi Auto WB' feature on the K3 is implemented.

Thanks
dave k

From what understand the sensor on the K3 can read different light sources within the scene and can then work out the correct WB for each different light temp to provide a more accurate WB reading. However, the explaining the science behind how the K3 can do this, I have no idea.

I was wondering whether the information about these differing segments is somehow mashed into the RAW data, or whether it is passed in Exif Makernotes to the RAW converter.

If this second scenario is correct, it presumably means that only the 'Digital camera Utility' will understand how to use it, and software such as ACR and LR will ignore it, with unpredictable results.

I suppose so experimentation is in order..

dave k

Good question! JPEG wise the data is embedded, however shooting RAW is different. When I have loaded images taken in RWA with a K3 and loaded them into Lightroom 5 they right WB wise in comparison to the camera screen image, although the camera display a JPEG image. Lightroom 5 does not yet support the K3 but it will only be a matter of time before an update is released that does. I have not used Pentax camera utility software but will take a look.

All I know is the exposure of the K3 is the best from Pentax so far.

There's a bit about it on the Ricoh website:

http://www.ricoh.com/about/company/technology/tech/026.html

Still doesn't indicate how it might apply to RAW images though.
The updated Silkypix Utility 5 that came with the camera should surely support I would hope.

John K

Interesting read, thanks for the link.

This is something that Ricoh has brought to the table, it looks like a very useful addition, now we can have different WB in different parts of the image.

This should be true for out-of-camera JPGs ( I haven't tried it), but I can't reach definite conclusions about RAW images.

I have played around with this today, comparing AWB & MAWB. I shot some images in a room lit with natural light from a window and a fluorescent light that gave some (not very dark) shadows. I then converted them in both ACR and the supplied DCU5 using no PP.

I couldn't see any real difference in colour between the ACR and DCU5 versions and in both cases AWB gave me significantly better fidelity than MAWB.

I retired confused.....

dave k

Is it really different WB in different parts of the shot, or is it just an averaged setting for the whole image based on multiple readings?

Also, I'ts always been difficult to be certain exactly what camera recorded metadata will be read and acted upon by different RAW converters. I agree with JAK above that if PDCU has been updated for the K3, it would be a bit strange not to include the new multi WB setting - based on my experienced with various RAW software, at this stage at least, I would be very surprised if it was supported by any other converters.

But the DNG format does store everything from the camera, so theoretically Adobe products can read it - the question is where each piece of metadata is stored and maps onto the Adobe parameter fields specifically.

Hence my confusion...

PDCU (it is now RCDU in V5) identifies the fact that MAWB was used, but I couldn't see any colour difference between this MAWB-aware converter and ACR v8.20 which is presumably not MAWB-aware.

Colours were more realistic in both converters when normal AWB was used.

ACR will by default apply the 'as shot' preset as recorded in the RAW - whatever the camera was set to when the shot was recorded will become the 'as shot' preset.

I'm assuming here Dave you did not see a 'MAWB' preset in ACR's dropdown, but only in PDCU V5 ? So you can't select it as an option in ACR? If so, then you won't be able to choose MAWB in ACR.

What was the image shot with? If it was MAWB then ACR would just load this WB temperature at default, so thats why you wouldn't see any difference between the two programs. Maybe the MAWB setting was just not suitable for that particular image.

My interest in this arises because I always shoot RAW images, and convert in ACR. Not knowing any of the details of how MAWB is implemented gives rise to a concern that using it could have an adverse effect on my conversions.

After reading a Ricoh Paper here , it seems that MAWB is concerned with WB in light and dark areas, so I shot images with two light sources and light/shadow areas, expecting to see that RDCU5 would give different results for the AWB and MAWB images.

I used two images, one with AWB and one with MAWB. For both images ACR just offered 'as-shot' Temps and Tints and the usual pre-sets. RDCU5 indicated correctly that one used AWB and the other MAWB but didn't seem to improve colours. It could well be that my conditions were not suitable for comparing AWB/MAWB. I will try a more rigorous test later this week and also check out-of-camera JPGs.

In the meantime, until I understand more about how MAWB is implemented and could affect my workflow, I'll avoid using MAWB.

davek

You can apply a different white balance to different parts of the image easily in Lightroom


No need to a buy a K3 for that


.

Yes, but it's still not clear (well not to me based on what has been said here) whether that is what the K3 actually does, (ie some kind of selective toning based on different zones of an image and their differing colour casts) or whether it is, more simply, creating a better averaged setting based on taking multiple readings around a scene.

If its the latter then I wouldn't be surprised if many images look very much the same on AWB or MAWB - those that might benefit from MAWB would likely be scenes that included very distinct mixed lighting sources (say a tungsten lit interior with a sunny day going on outside through the window).

Dave, you mentioned ACR 'as shot' setting above - my understanding of this is that it will display the temperature of the image as recorded by the camera settings - the slider defaults are just new 'zero points' to make adjustments from the 'as shot' starting place.

The Ricoh article is very clear. You could have different light sources on subject and background and the WB would be corrected individually for the different areas.

They show an example using flash on the foreground subject.