Why does WB matter when raw shooting ?

[MarcoC]

Many have noted the perhaps less than desirable shift in auto WB peroformace that came with a700 firmware v4, when shooting indoors and using flash. Garry Friedman recently reported and indeed demonstrated it in the latest version of his a700 e-book - indoor shots with flash using v4 tend towards a warmer more yellow WB. Fine, that can be changed in raw processing.

BUT - I can't understand why some posters over on dpreview and Dyxum insist that getting the WB correct at the time of shooting is important, even when shooting raw. Sure, it's good practice to do so, but does the dialed in WB actually affect the way the raw image is created and stored or is WB just a tag that is added to the raw file and then read by raw processors? In which case it shouldn't matter whether you got the WB wrong when you shot, if you are shooting raw.

Ao can any tech heads who really understand the way sensors record data enlighten me - does the dialed in WB acxtually affect the way the sensor records a raw image and if so, is it therefore worthwhile busting a gut to get WB right at the time you shoot, even in raw ?

[Yagil Henkin]

but it may save you time post-processing and espcially when batch-processing.

[Dr. Harout]

And welcome to the forum MarcoC.

[pakodominguez]

Any sensor is optimized for a certain standard (most of the time daylight @ 200 ISO) and everithing else is programing. That means if you use a 80A filter you will get a better file than if you set the white balance to Tungsten. If you shot under tungsten light with your white balance set to 5500K (daylight) you can recover that "mistake" while procesing the RAW, but you will get a better quality if you set it right from the begining. Some extreme corrections just won't work properly.

It isn't because you can fix things during the post production that you are allow been sloppy, unless the quality provided reach your standards.

Regards

[jcoffin]

Many have noted the perhaps less than desirable shift in auto WB peroformace that came with a700 firmware v4, when shooting indoors and using flash. Garry Friedman recently reported and indeed demonstrated it in the latest version of his a700 e-book - indoor shots with flash using v4 tend towards a warmer more yellow WB. Fine, that can be changed in raw processing.

BUT - I can't understand why some posters over on dpreview and Dyxum insist that getting the WB correct at the time of shooting is important, even when shooting raw. Sure, it's good practice to do so, but does the dialed in WB actually affect the way the raw image is created and stored or is WB just a tag that is added to the raw file and then read by raw processors? In which case it shouldn't matter whether you got the WB wrong when you shot, if you are shooting raw.

Ao can any tech heads who really understand the way sensors record data enlighten me - does the dialed in WB acxtually affect the way the sensor records a raw image and if so, is it therefore worthwhile busting a gut to get WB right at the time you shoot, even in raw ?

My guess is no, but I doubt anybody outside of Sony can say with absolute certainty. The situation is that the sensor collects some data. That's then converted to digital. When you adjust the ISO, you're changing the gain of an amplifier between the sensor and the A->D converter (often contained in the same package along with the A->D itself -- clearly so in the case of CMOS sensors with integrated A-D). For ISO adjustment, you're adjusting all of them together by the same amount (well, effectively the same amount -- there's probably some compensation for differences in sensitivity of the different channels). By adjusting the amplifiers separately from each other (the relatively gain of the red and blue channels in particular) you could do at least some of the white balancing before the data got to A->D converters.

Realistically, I'd guess the chances that anybody is doing such a thing at the present time are just about nil. It's just drastically easier to capture the data "flat" and compensate after the fact instead.

The other side of it is that if/when you can, it's likely to be a lot easier to do a good job of color balancing when you take the shot than after the fact. If you're in a really critical situation, there are grey cards specifically for the purpose of white balancing. For most normal situations, however, a piece of plain white copier paper will do the job quite nicely -- and while the current Sony's don't make custom white balancing quite as accessible as the KM7D did, they still make it easier than any more than the most trivial adjustment in post processing.

[peterottaway]

I think I prefer " live long and prosper " to the collective.

[MarcoC]

Thanks to everyone (esp. Sonolta) for their thoughtful and informative replies. It seems that nobody knows for sure if WB affects the way the a700 sensor records the image, other than through affecting exposure, although Sonolta's experience (which I respect - you shoot far more than I do), suggests that it is clearly advantageous under certain conditions to take the time to get WB right when shooting. So thank you to everyone - I've learned a lot.

David - what are your thoughts ?

[David Kilpatrick]

I'm still in Mallorca, shot the last sunset from Port D'Andratx (cRAW but set to Daylight and AWB for various shots because it helps preset the raw converter, and give me a decent preview on the camera). As far as I have been able to tell, WB settings do not affect gain or D to A, but may change exposure (need to check). I can tell you for sure than using an 80B blue filter and daylight WB gives 1 to 1.5 stops more highlight clipping adjustment than using AWB, with all KM/Sony cameras to date. I use 80B/tungsten and it works.

David

[braeside]

You are missing the snow in the borders tonight David.

[pakodominguez]

I can tell you for sure than using an 80B blue filter and daylight WB gives 1 to 1.5 stops more highlight clipping adjustment than using AWB, with all KM/Sony cameras to date. I use 80B/tungsten and it works.

With Tungsten and Fluorescent lights, right?

[alphaPDX]

I was caught off guard when my A200 slowed down while shooting RAW recently. Turned out I had left WB-Bracket on from shooting jpgs earlier in the day, and it applied same to three RAWs at a time. That really times out the A200 in a hurry! I was surprised by that setting not being overwritten when shooting RAW, but will now consider it a 'feature' & read my back panel a bit more carefully in the future..

[David Kilpatrick]

I can tell you for sure than using an 80B blue filter and daylight WB gives 1 to 1.5 stops more highlight clipping adjustment than using AWB, with all KM/Sony cameras to date. I use 80B/tungsten and it works.

With Tungsten and Fluorescent lights, right?

I have not tried a fluorescent correction filter, since there are so many types and you need one exactly matched to the lights in use. I only use the 80B when shooting by tungsten light (normally tungsten halogen or Halostar modelling lamps) and only in the studio; for any quick shots 'in the field' I just use preset or auto white balance. For fluorescent light, or energy-saver bulbs, I use AWB (the preset is very rarely correct, because it applies only to one type of fluorescent light). If you want precision correction for various types of light, Singh-Ray used to make such filters; you could correct for anything from mercury vapour to named types of building floodlighting.

David