arthurking83
29-12-2010, 11:25pm
It's time once again, for me to annoy as many folks as I possibly can, thus living up to my avatar and become the PITA through distorted logic, confusing concepts and illogical thought processes! :D
Whitebalance, not everyone knows what it is, and how it really works, even though many of us know how important it is in digital photography.
I use(almost exclusively) Auto WB setting in my camera, except in specific situations(such as a studio shoot a while back in Marlo).
but sometimes you just can't get the exact perfect Wb setting due to specific conditions.. but usually post processing helps there.
And that's where I fall down flat on my face with fits of confusion and impatience!
My choice of software is actually very simple easy to use and effective, but I hate the time taken to post process images.. and I am referring to the many minutes it takes to, say.. create a gradient selection and a simple desat, or colour balance setting and apply that to an image. all 1 minute and 30 seconds of it. I think if there's a way to do it, without the need to waste those 90seconds, I want to do it that way. Batch processing! :th3: ... the seemingly more efficient way to do it, but with one caveat.. all images have to be almost exactly the same, ie. requiring the exact same precess step(or layer if you prefer).. which makes for extremely boring photography(I reckon).
So you need a particular type of scene, and sometimes you can get the situation where you see an of balance WB colour cast across the entire frame for some reason or another.
And you now think think to yourself... what? :confused:.. how can WB be different across a frame. The sensor always acts as a whole, and WB is going to be correct across the entire single exposure.
So yes, you maybe correct in thinking I'm now being a right PITA again in trying to cause more confusion ... but I have some experience in this area(in fact quite a lot).
GNDs.. more accurately stacked GNDs. heavy use of stacked GNDs almost always produce an imbalance in the WB setting. Stack 9 stops of graduated filtering on your scene, and watch how the darkened area turns a nice shade of deep magenta as your clear based section of the scene turns all blue. no mater what WB setting you choose in camera, it'll almost certainly be tainted with a tint of varying degrees. I do this a lot. shooting into a bright setting sun, with the foreground in shadow and the upper sky bathed in glorious golden light. Sky will always turn deep red, the more I warm up the shadow areas for correct WB there. So WB adjustment is not an answer in PP.. it's more of a problem. Fo rthis reason I decided that a magenta sky is not really a problem(purple is another variation of magenta!) and I've had lots of comments on 'my purple skies' over the years. i kind of like them now(both magenta and purple) as they display a different opinion of how the scene looked a the time(according to my view that is). While to the naked eye, it looked different, through the viewfinder it looked as I display it(usually).. I'm not normally a crossprocessor type of photographer. Cross processing is where you apply different WB settings that distort the reality. WB can also be replaced with colour switching.. it basically amounts to the same thing.
I think in the film world cross processing was where they used one film processing type(say C-41) onto a non similar film type(say E-6).
I remember this from a non direct handling of the film types being the delivery boy for many photographers back in the old days.
Anyhow.. back on track and to the point. The point being using filters.
I recently uploaded an image of my window blind that had me wondering why it looked the way it did. About two weeks before this tho, I had the idea to use a specific filter to try to balance the WB of the scene when using grads in a heavy handed manner. it turns out that the window blind scene was a perfect way to test my idea .. almost without getting out of bed!(bedroom window, silly! :D)
Here's the window scene:
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3599&gal=gallery&type=full
6924
I have to say too, that I did in fact purchase a cheapo graduated blue filter on a whim. :o
(I've been told by a good friend that I can be a bit whimsical.. but I no longer see this friend due to other differences of opinion, but she was a nice person tho).
Ok, so I now have a grad blue filter.. and in the digital age a redundant piece of gear as adding a glue graduation in PP is trivial, simple, easy and much more effective as you have a lot more control over it. No need to stuff around with guess work and such hit and miss measures. But then again, about two/three years back I also got a cheapo graduated tobacco filter too(just for some fun experimentation). I hardly ever use it, almost never for that 'sunset effect' that it was originally intended to be used for.
So armed with a grad orangey filter and a grad blue filter, you'd think that this image makes more sense.. but the reality is that it doesn't!
it'd be easy to assume that with the two filters loaded upside down to each other this scene would be easy to make. that is orangey filter up top, and bluey filter reversed to get the bottom of the blind looking blue. .. right!?
Wrong! totally. this image is untouched both in terms of filters and post processing. PP steps was setting Picture Control(in camera processing for non Nikon peeps) from Standard to Neutral. then saved in jpg format.
and yet it's blue down low and orangey up top.
my issue is not colour balance, and I'll explain what probably causes this window to show such a varitone image early in the morning.
First up was my need to balance the light properly, and eliminate the blown highlight without losing shadow detail.. so i turn to my always faithful GND. First the 2stop, but then the three stopper for better exposure.
Doing this subsequently distorted the WB reality.. as always.
Using the dark grey area up top to lower the brightness level only exaggerated the warm to cool colour balance and the upper section of the blind became more saturated with red(warmer and more browney-orange). in this case not a problem as it was the closest rendition of reality I could get.
What causes this dual tone is that the window faces away from the morning sun, and the light that's reflected back onto the widow is off my next door neighbour's house which is rendered and painted orange! Easy explanation .. except for the blue section. That's even easier to explain actually. The lower section of the window is shielded by the paling fence between us, and hence doesn't get any of the bright orange reflected light from their wall. Shadows areas always like to turn blue when camera can't decide what the right WB setting should be when such a disparity exists .. just as I get when using GNDs in a massive effort to balance dynamic range ;)
In a shadowy scene, set camera WB to daylight(5000-5500K and your shadowy scene will turn blue). in shadowy conditions camera WB should be set to cloudy to shade.. in Nikon terms this means 6000K to 8000K. It all depends on the scene, and the operators preference.
Hence why my skies sometimes look purpley-red.. I usually set Wb for the shadows, and then prefer to use artistic license for the colour of the sky. Let the camera decide what was appropriate and disregard the use of the graduated process step to balance the colours for photographic correctness.
(summary of this: I'm lazy .. lazy in terms of post processing).
So I added a graduated blue filter to the scene, to try to resemble what should be a correct colour balance.
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3600&gal=gallery&type=full
6925
blue is now almost gone from the bottom of the window, even though they are the same scene, taken at the same time of the day(seconds apart). Later in the day the window looks more like this as the sun subsequently casts direct light onto the window. and not reflected orangey light from next door's orangey house. So adding the blue filter has almost removed the blue cast.
Does that make sense at all?
Those that understand why may be feeling smug, and the confusing answer to this seemingly lunatic logic is easy as can be! AWB
Auto WB. The camera is always going to try to produce for you the best exposure it can. Better cameras(newer) produce better AWB set images due to more intricate systems that the camera manufacturers use. better WB sensors, colour exposure meters.. etc, etc. AWB is much better on a 2007 D300, than on a D70 from 2005(or 6 or whenever) simple technology advancements.
So the blue filter has done something that the camera doesn't know about. produced a blue cast to the upper part of the scene, and hence the AWB sensor then adjusted the setting to produce slightly warmer setting. Warmer WB setting means less blue as much as is possible.
The central section is still orangey because I probably didn't ustilize the full graduation surface of the blue filter. the blue grad was used doubled up with the 3 stop GND, as the blue filter isn't very capable of reducing exposure. that's not their purpose. They only exist to either blue up the blue skies in traditional film use(common use). Lee also produce them as well as Cokin, and their marketing always refers to using them as a blue enhancing method, but never as a colour balance filter.
The reason why there is no blue cast, as you'd be expecting to see is because the camera is seeing a lot of blue now and trying to counter the issue with a more neutral setting .. hence the blue down low is also gone.
I could have played with WB settings to produce an exact replica of what the scene really looked like with the grad plus blue filter in place. but that wasn't the purpose of the test. The test was a lazy way of checking out if the blue over heavy GND filter really would help to remove the magenta cast that multiple GNDs produce.. and I think it will(help) but not totally eliminate the problem.
Not that I'm going to make heavy use of the filter. Sometimes I like the magenta cast skies now and then in come images it just gives a slightly different lok.
also for those interested, WB set by camera:
for the unfiltered image: is approximately 4200K and no tint adjustment, which turns out to be close to a setting of High Colour Rendering Flouro! :D(I initially checked out daylight, 5200K)
and for the filtered image: is about Daylight 5200K plus a -1(out of 12 steps) tint adjustment( ie. into the red channel)
Just to give you an idea of how easily fooled Wb is by the camera.. and therefore how to manipulate it to your advantage.
Note I also tried the grad tobacco on the window as well, using it on the lower blue section and it balanced perfectly.. a nice smooth copper-brown transition across the scene from top to bottom.
I've tried using the grad tobacco in the heavy handed GND situations, and it doesn't work(well). gives an interesting colour perspective, but not good balance, IMO. I'm hoping that the blue is going to do better tho.
will test at the next opportunity.. but it may just end up being a curio to play with more than anything else.
Now that I've bored you into a coma, silly you should have taken the time to do this all for yourself!
Whitebalance, not everyone knows what it is, and how it really works, even though many of us know how important it is in digital photography.
I use(almost exclusively) Auto WB setting in my camera, except in specific situations(such as a studio shoot a while back in Marlo).
but sometimes you just can't get the exact perfect Wb setting due to specific conditions.. but usually post processing helps there.
And that's where I fall down flat on my face with fits of confusion and impatience!
My choice of software is actually very simple easy to use and effective, but I hate the time taken to post process images.. and I am referring to the many minutes it takes to, say.. create a gradient selection and a simple desat, or colour balance setting and apply that to an image. all 1 minute and 30 seconds of it. I think if there's a way to do it, without the need to waste those 90seconds, I want to do it that way. Batch processing! :th3: ... the seemingly more efficient way to do it, but with one caveat.. all images have to be almost exactly the same, ie. requiring the exact same precess step(or layer if you prefer).. which makes for extremely boring photography(I reckon).
So you need a particular type of scene, and sometimes you can get the situation where you see an of balance WB colour cast across the entire frame for some reason or another.
And you now think think to yourself... what? :confused:.. how can WB be different across a frame. The sensor always acts as a whole, and WB is going to be correct across the entire single exposure.
So yes, you maybe correct in thinking I'm now being a right PITA again in trying to cause more confusion ... but I have some experience in this area(in fact quite a lot).
GNDs.. more accurately stacked GNDs. heavy use of stacked GNDs almost always produce an imbalance in the WB setting. Stack 9 stops of graduated filtering on your scene, and watch how the darkened area turns a nice shade of deep magenta as your clear based section of the scene turns all blue. no mater what WB setting you choose in camera, it'll almost certainly be tainted with a tint of varying degrees. I do this a lot. shooting into a bright setting sun, with the foreground in shadow and the upper sky bathed in glorious golden light. Sky will always turn deep red, the more I warm up the shadow areas for correct WB there. So WB adjustment is not an answer in PP.. it's more of a problem. Fo rthis reason I decided that a magenta sky is not really a problem(purple is another variation of magenta!) and I've had lots of comments on 'my purple skies' over the years. i kind of like them now(both magenta and purple) as they display a different opinion of how the scene looked a the time(according to my view that is). While to the naked eye, it looked different, through the viewfinder it looked as I display it(usually).. I'm not normally a crossprocessor type of photographer. Cross processing is where you apply different WB settings that distort the reality. WB can also be replaced with colour switching.. it basically amounts to the same thing.
I think in the film world cross processing was where they used one film processing type(say C-41) onto a non similar film type(say E-6).
I remember this from a non direct handling of the film types being the delivery boy for many photographers back in the old days.
Anyhow.. back on track and to the point. The point being using filters.
I recently uploaded an image of my window blind that had me wondering why it looked the way it did. About two weeks before this tho, I had the idea to use a specific filter to try to balance the WB of the scene when using grads in a heavy handed manner. it turns out that the window blind scene was a perfect way to test my idea .. almost without getting out of bed!(bedroom window, silly! :D)
Here's the window scene:
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3599&gal=gallery&type=full
6924
I have to say too, that I did in fact purchase a cheapo graduated blue filter on a whim. :o
(I've been told by a good friend that I can be a bit whimsical.. but I no longer see this friend due to other differences of opinion, but she was a nice person tho).
Ok, so I now have a grad blue filter.. and in the digital age a redundant piece of gear as adding a glue graduation in PP is trivial, simple, easy and much more effective as you have a lot more control over it. No need to stuff around with guess work and such hit and miss measures. But then again, about two/three years back I also got a cheapo graduated tobacco filter too(just for some fun experimentation). I hardly ever use it, almost never for that 'sunset effect' that it was originally intended to be used for.
So armed with a grad orangey filter and a grad blue filter, you'd think that this image makes more sense.. but the reality is that it doesn't!
it'd be easy to assume that with the two filters loaded upside down to each other this scene would be easy to make. that is orangey filter up top, and bluey filter reversed to get the bottom of the blind looking blue. .. right!?
Wrong! totally. this image is untouched both in terms of filters and post processing. PP steps was setting Picture Control(in camera processing for non Nikon peeps) from Standard to Neutral. then saved in jpg format.
and yet it's blue down low and orangey up top.
my issue is not colour balance, and I'll explain what probably causes this window to show such a varitone image early in the morning.
First up was my need to balance the light properly, and eliminate the blown highlight without losing shadow detail.. so i turn to my always faithful GND. First the 2stop, but then the three stopper for better exposure.
Doing this subsequently distorted the WB reality.. as always.
Using the dark grey area up top to lower the brightness level only exaggerated the warm to cool colour balance and the upper section of the blind became more saturated with red(warmer and more browney-orange). in this case not a problem as it was the closest rendition of reality I could get.
What causes this dual tone is that the window faces away from the morning sun, and the light that's reflected back onto the widow is off my next door neighbour's house which is rendered and painted orange! Easy explanation .. except for the blue section. That's even easier to explain actually. The lower section of the window is shielded by the paling fence between us, and hence doesn't get any of the bright orange reflected light from their wall. Shadows areas always like to turn blue when camera can't decide what the right WB setting should be when such a disparity exists .. just as I get when using GNDs in a massive effort to balance dynamic range ;)
In a shadowy scene, set camera WB to daylight(5000-5500K and your shadowy scene will turn blue). in shadowy conditions camera WB should be set to cloudy to shade.. in Nikon terms this means 6000K to 8000K. It all depends on the scene, and the operators preference.
Hence why my skies sometimes look purpley-red.. I usually set Wb for the shadows, and then prefer to use artistic license for the colour of the sky. Let the camera decide what was appropriate and disregard the use of the graduated process step to balance the colours for photographic correctness.
(summary of this: I'm lazy .. lazy in terms of post processing).
So I added a graduated blue filter to the scene, to try to resemble what should be a correct colour balance.
http://www.ausphotography.net.au/forum/dbtgallery.php?do=gallery_image&id=3600&gal=gallery&type=full
6925
blue is now almost gone from the bottom of the window, even though they are the same scene, taken at the same time of the day(seconds apart). Later in the day the window looks more like this as the sun subsequently casts direct light onto the window. and not reflected orangey light from next door's orangey house. So adding the blue filter has almost removed the blue cast.
Does that make sense at all?
Those that understand why may be feeling smug, and the confusing answer to this seemingly lunatic logic is easy as can be! AWB
Auto WB. The camera is always going to try to produce for you the best exposure it can. Better cameras(newer) produce better AWB set images due to more intricate systems that the camera manufacturers use. better WB sensors, colour exposure meters.. etc, etc. AWB is much better on a 2007 D300, than on a D70 from 2005(or 6 or whenever) simple technology advancements.
So the blue filter has done something that the camera doesn't know about. produced a blue cast to the upper part of the scene, and hence the AWB sensor then adjusted the setting to produce slightly warmer setting. Warmer WB setting means less blue as much as is possible.
The central section is still orangey because I probably didn't ustilize the full graduation surface of the blue filter. the blue grad was used doubled up with the 3 stop GND, as the blue filter isn't very capable of reducing exposure. that's not their purpose. They only exist to either blue up the blue skies in traditional film use(common use). Lee also produce them as well as Cokin, and their marketing always refers to using them as a blue enhancing method, but never as a colour balance filter.
The reason why there is no blue cast, as you'd be expecting to see is because the camera is seeing a lot of blue now and trying to counter the issue with a more neutral setting .. hence the blue down low is also gone.
I could have played with WB settings to produce an exact replica of what the scene really looked like with the grad plus blue filter in place. but that wasn't the purpose of the test. The test was a lazy way of checking out if the blue over heavy GND filter really would help to remove the magenta cast that multiple GNDs produce.. and I think it will(help) but not totally eliminate the problem.
Not that I'm going to make heavy use of the filter. Sometimes I like the magenta cast skies now and then in come images it just gives a slightly different lok.
also for those interested, WB set by camera:
for the unfiltered image: is approximately 4200K and no tint adjustment, which turns out to be close to a setting of High Colour Rendering Flouro! :D(I initially checked out daylight, 5200K)
and for the filtered image: is about Daylight 5200K plus a -1(out of 12 steps) tint adjustment( ie. into the red channel)
Just to give you an idea of how easily fooled Wb is by the camera.. and therefore how to manipulate it to your advantage.
Note I also tried the grad tobacco on the window as well, using it on the lower blue section and it balanced perfectly.. a nice smooth copper-brown transition across the scene from top to bottom.
I've tried using the grad tobacco in the heavy handed GND situations, and it doesn't work(well). gives an interesting colour perspective, but not good balance, IMO. I'm hoping that the blue is going to do better tho.
will test at the next opportunity.. but it may just end up being a curio to play with more than anything else.
Now that I've bored you into a coma, silly you should have taken the time to do this all for yourself!