Originally posted by photoptimist Not everything is an approximation. It's "color temperature" that is a very poor approximation of white balance for most of the light sources found in the modern world. I agree with Steve's hypothesis is that Sony's reported "Color Temperature" number would not be useful for the purpose of picking gels to match it.
That said, I agree with you that it would be nice if Pentax' MWB implementation provided some indication of where a given measured MWB is on the blue-amber and magenta-cyan axes. If it did that, you could do your own measurements of each gel, write them down, and then know which one to use when you measured the scene.
Ok... so... lets be clear now then. If both of you are saying that the fact Sony spits out a Kelvin value when you grey card a scene in MWB, and that the value given is 'not useful for the purpose of picking gels to match it' then you're actually picking a bone with Robert Hall, the guy behind making that video about how and when to gel a camera. If you watch the video I think you will find he is spot on the money for what he is saying, I just find it a little hard that he would include this part (about MWB and using the Kelvin given to grab an approrpriate gel) if he didn't actually think it worked and was of value. Like I say, we don't need to be exact, just improve the situation the photographer is in. We don't want to be firing 5500k light from our strobe and get that light on the subject if all the ambient light around them is 3200k, it looks weird. What we want is someway for our camera to tell us the WB of the environment we're in, and sadly even when we grey card the scene Pentax won't tell us that (but other brand/s do... could well be Canon and Nikon also do this..). Up til now I have just eyeballed it and manually adjusted the temp till I get something on the Live View screen I like. But when we factor in any Custom Image settings it can skew things... would be really nice if the Grey Card/MWB thing was actually giving us a value.
Furthermore he reviews gels a lot, saying which ones are good value, which are accurate and how some change in values of a tenth etc etc. He goes to say how Rogue for example will tell you the strength of the gel (1/4 CTO)
and supply you with a chart for kelvin temps for the gels. So a good gel (pack) should provide you with (already) printed values on the gels themselves, or at least what the heck they are...
Originally posted by stevebrot So...what happens if you plug the displayed K° back into the camera? Do you get the same result?
I am serious.
This is not true. A manual gray card WB setting results in four color correction factors, (1R, 2G, 1B) that are applied as part of RAW processing. The numbers are what is required to render equal values for the set of (RGGB) pixels in the Bayer matrix (result = gray) within the sampled area and nothing more. Manual entry of K° value results in the same four numbers, but may not result in the same four numbers as a gray card measurement even if illumination is from a standard light source. It may not be obvious, but the relationship of white balance to a K° isotherm spectrum on the camera side of things is purely coincidental and tied to historic color film response specifications.
Steve
I can't test some of that out because I don't have a Sony. What I can do (and will do) is tripod my camera up at a scene, use the MWB mode and hold a grey card in front of the camera to generate a WB. Take a shot. Then quickly (before much light changes too much etc) eject the SD Card, import to LR and see what LR the shot is reporting back in Kelvins was taken at. Then I shall take a second shot quickly but this time dial in the White Balance to be as close to what the file in LR is saying, and then compare if they are similar or quite off... I guess if I was tethered to LR this could be easier.
Oh... and I also meant that like every RAW file has a kelvin temp taken at the time of the shot, whether its AWB or Multi-AWB or whatever, I'm simply saying its
recorded in the file and you get that info in LR when imported.
I just think that the point about the grey card
is to do with flashes,
nothing else (ok perhaps relevant to Jpgers as well... but still). If shooting RAW and not using flashes you can do whatever the heck you like with the WB at the time (because you can choose a better one in post), but you
CANNOT fix in post different colours in the scene caused by flash giving off one temp and ambient light another, that's not workable. That's why not getting a kelvin temp after using a Grey Card is absolutely baffling to me...
its like the whole point... once again though Pentax and Flashes seem to really be an afterthought with them...
Originally posted by stevebrot OK...here is a workaround...
1. Save off the manual white balance to file
2. Open the file in Lightroom and modify the WB using the develop module with a small change, say +1 K°.
3. Read the resulting K° from LR display and subtract the change.
The result is based on a back-calculation of the correction factors in the EXIF and should be fairly close. I have not tested this (need to go out and trim some trees), but it should work, perhaps even in LR mobile.
Steve
It's not a workaround when on a shoot though is it...
I have a shoot next weekend at a posh hotel in Sydney, a function room for a party. I will be gelling my flashes to match the room ambience, but how the heck do I find out the room temperature if even when using a grey card it won't tell me the kelvin value to let me choose an appropriate gel to match. What I don't want is a cool flash tone on the faces of the portraits and yellow/orange ambient lights in the background, that looks bad. This is the entire point in grey carding (imo) and it's not been executed properly by Pentax!