Matching White Balance Of Two Photos.

[ByRiker'sBeard6]

I took two photos of myself. First I was standing still with a gray card, in second one I was throwing a ball in the air to get a blur. I used ND filters for latter.

F/7.1 for both
1/500 for first, 1/8 for second.

I find longer exposures tend to be reder and short exposures bluer. I used white siding of nearby house as white reference in second image. Still don't quite have equal exposure but mostly I'm concerned with my blue shirt turning purple in longer exposure. Why might that be?

Ultimately I'd like to be able to superimpose a still person in front of a moving background, but need to normalize colors in both.

 

[petrochemist122 147]

I took two photos of myself. First I was standing still with a gray card, in second one I was throwing a ball in the air to get a blur. I used ND filters for latter.

F/7.1 for both
1/500 for first, 1/8 for second.

I find longer exposures tend to be reder and short exposures bluer. I used white siding of nearby house as white reference in second image. Still don't quite have equal exposure but mostly I'm concerned with my blue shirt turning purple in longer exposure. Why might that be?
yyou
Ultimately I'd like to be able to superimpose a still person in front of a moving background, but need to normalize colors in both.

Many image editing tools have a white balance picker, that allows you to select a point in the image to make neutral.

If you used the same siding as WB on each shot (the second through the ND filter) the colours should match. Do be aware that ND filters are never totally neutral, the better ones might be close, my cheapest are WAY off (but they don't actually claim to be ND filters)

 

[AlanKlein86 3]

ND filter can add blue. The better ones less so.

 

[Rickbb69 8]

Instead of using a filter try lowering the ISO and upping the F-stop for the blurred motion shot. Maybe use a telephoto lens on a tripod to create the blur if you can't get the shutter speed low enough.

You could also set for auto white balance and have the same white object in both shots in case you need to edit it in post.

As everyone said filters introduce artifacts, combining shots of with and without will be tricky to get perfect color match. You may have no choice depending on what you're shooting, in that case a good post editing software is where you will have to work.

 

[ByRiker'sBeard6]

Many image editing tools have a white balance picker, that allows you to select a point in the image to make neutral.

If you used the same siding as WB on each shot (the second through the ND filter) the colours should match. Do be aware that ND filters are never totally neutral, the better ones might be close, my cheapest are WAY off (but they don't actually claim to be ND filters)

White balance selectors work as well on white card as gray card?

 

[ByRiker'sBeard6]

Instead of using a filter try lowering the ISO and upping the F-stop for the blurred motion shot. Maybe use a telephoto lens on a tripod to create the blur if you can't get the shutter speed low enough.

You could also set for auto white balance and have the same white object in both shots in case you need to edit it in post.

As everyone said filters introduce artifacts, combining shots of with and without will be tricky to get perfect color match. You may have no choice depending on what you're shooting, in that case a good post editing software is where you will have to work.

Has anyone created a true ND filter? Why the coloring?

 

[petrochemist122 147]

Has anyone created a true ND filter? Why the coloring?

No, because it's impossible to get it perfect.
Many filters are made by mixing dyes and however much effort is applied some residual hue will remain, in addition this may vary slightly from camera to camera.
I gather some are made using fine metallic particles, which means they also block near infra red (which goes through most dyes) These will probably be more uniform but will have slight tints for impurities etc.

 

[petrochemist122 147]

White balance selectors work as well on white card as gray card?

They should do as long as none of the channels are maxed out. I've used white paint on the side of a race track to correct white balance. For infrared work it is common to use foliage to set white balance, though concrete, skin, PTFE, clouds and no doubt many other materials are also used.
All a software base white balance does is look at the relative strengths of the three channels & apply corrections to make all three the same. For artistic effects white balance can even be set on distinctly coloured parts of an image.

The benefit of a grey card is that it's designed to match the 18% grey that camera systems try to aim for in exposures.

 

[ByRiker'sBeard6]

No, because it's impossible to get it perfect.
Many filters are made by mixing dyes and however much effort is applied some residual hue will remain, in addition this may vary slightly from camera to camera.
I gather some are made using fine metallic particles, which means they also block near infra red (which goes through most dyes) These will probably be more uniform but will have slight tints for impurities etc.

Are other filters pretty close to perfect? Like red, green, or blue? Are attempts at black the only issue?

 

[petrochemist122 147]

Are other filters pretty close to perfect? Like red, green, or blue? Are attempts at black the only issue?

Polarising filters will also often introduce a colour cast and they are certainly not 100% transmitting of one polarisation & 100% blocking of the perpendicular polarisation. Some are 'high extinction' designed to be better at blocking the unwanted polarisation, others to be 'high transmission' letting more of the unwanted polarisation in order to maximize the total light.

When I've looked at filters that are supposedly colour neutral on a visual spectrometer (I made use of the one at work) they always had some degree of tilt to the visible spectrum with additional finer details such as bumps & pits. So some wavelengths of red were transmitted more than other nearby reds (as with what we would see as blues
& greens).

The purpose of coloured filters allows much more leeway, sometimes very much more.
A strong red filter will usually be a long pass design that blocks nearly all light below a set wavelength goes through a transmission area then transmits at more than 90% of longer wavelengths. The wavelength of the transmission zone will very with different types of red filter. Fainter reds will also allow more of the shorter wavelengths, but I've not really used those.
Blue filters that are designed for three colour work are similar with the short end of visual light, but will typically leak significant amounts of near infra red, which some digital cameras can see (though it's generally only significant with converted cameras). There are plenty of other blue tinged filters used for cooling the image that are far more subtle in their effect - these are typically only used for film.
Green filters will cut light at both ends of the visual spectrum but are in my experience a lot more variable. I have three different #58 green filters which are fairly similar through the visual range but differ markedly in the near IR - one transmits loads, one very little & the third in between - only one works for my application Other types of green filter (different wratten numbers) can differ very significantly. One tiny filter I have looks almost silvery as it reflects most light & only transmits a few nanometers of green light (I think it was 5nm either side of the peak transmission) I guess the lens I took it off was previously used for working with green lasers of it's peak wavelength.

Humans perception of colour is actually rather complicated & a LONG way from what the light actually is.

While we see three colours light in the visual band of the electromagnetic spectrum covers ~400 to ~700nm each photon will has a specific wavelength in this range - specific to much less than 0.01nm (I've never worked with instruments that can measure differences less than 0.01nm) which leave well over 30,000 actual possible wavelengths. White light from continuum sources will typically include all of these wavelengths.

The cells in our eyes that respond to colours over lap considerably in their responses, with the relative response being used by our brain to interpret colour. There are 'metamers' mixtures of two pure wavelengths that we see as a completely different colour. Other animals see colours differently to us, some only have two colour receptors & some at least 13 different types. Some peoples vision differs too (colour blind etc.)

Digital cameras are usually designed to mimic our eyes as nearly as possible, but the filters used in one camera will differ than the ones used in another brand and none are perfect. Lenses are often not complete colour neutral either some having quite distinct tints. Fortunately for pictorial use perfection is not needed & often not even desirable.

 

[AlanKlein86 3]

Polarising filters will also often introduce a colour cast and they are certainly not 100% transmitting of one polarisation & 100% blocking of the perpendicular polarisation. Some are 'high extinction' designed to be better at blocking the unwanted polarisation, others to be 'high transmission' letting more of the unwanted polarisation in order to maximize the total light.

When I've looked at filters that are supposedly colour neutral on a visual spectrometer (I made use of the one at work) they always had some degree of tilt to the visible spectrum with additional finer details such as bumps & pits. So some wavelengths of red were transmitted more than other nearby reds (as with what we would see as blues
& greens).

The purpose of coloured filters allows much more leeway, sometimes very much more.
A strong red filter will usually be a long pass design that blocks nearly all light below a set wavelength goes through a transmission area then transmits at more than 90% of longer wavelengths. The wavelength of the transmission zone will very with different types of red filter. Fainter reds will also allow more of the shorter wavelengths, but I've not really used those.
Blue filters that are designed for three colour work are similar with the short end of visual light, but will typically leak significant amounts of near infra red, which some digital cameras can see (though it's generally only significant with converted cameras). There are plenty of other blue tinged filters used for cooling the image that are far more subtle in their effect - these are typically only used for film.
Green filters will cut light at both ends of the visual spectrum but are in my experience a lot more variable. I have three different #58 green filters which are fairly similar through the visual range but differ markedly in the near IR - one transmits loads, one very little & the third in between - only one works for my application Other types of green filter (different wratten numbers) can differ very significantly. One tiny filter I have looks almost silvery as it reflects most light & only transmits a few nanometers of green light (I think it was 5nm either side of the peak transmission) I guess the lens I took it off was previously used for working with green lasers of it's peak wavelength.

Humans perception of colour is actually rather complicated & a LONG way from what the light actually is.

While we see three colours light in the visual band of the electromagnetic spectrum covers ~400 to ~700nm each photon will has a specific wavelength in this range - specific to much less than 0.01nm (I've never worked with instruments that can measure differences less than 0.01nm) which leave well over 30,000 actual possible wavelengths. White light from continuum sources will typically include all of these wavelengths.

The cells in our eyes that respond to colours over lap considerably in their responses, with the relative response being used by our brain to interpret colour. There are 'metamers' mixtures of two pure wavelengths that we see as a completely different colour. Other animals see colours differently to us, some only have two colour receptors & some at least 13 different types. Some peoples vision differs too (colour blind etc.)

Digital cameras are usually designed to mimic our eyes as nearly as possible, but the filters used in one camera will differ than the ones used in another brand and none are perfect. Lenses are often not complete colour neutral either some having quite distinct tints. Fortunately for pictorial use perfection is not needed & often not even desirable.

What do you mean by this: "Some are 'high extinction' designed to be better at blocking the unwanted polarisation, others to be 'high transmission' letting more of the unwanted polarisation in order to maximize the total light."

I never knew there were two types. With my standard camera polarizers, I tend to only go to about 50%. I find if I go to full polarization, it kills the life out of the scene by deleting too much light especially on the foliage and grass, even while saturating colors at 100%. Which type would be better for me, and where do they sell them?

 

[petrochemist122 147]

What do you mean by this: "Some are 'high extinction' designed to be better at blocking the unwanted polarisation, others to be 'high transmission' letting more of the unwanted polarisation in order to maximize the total light."

I never knew there were two types. With my standard camera polarizers, I tend to only go to about 50%. I find if I go to full polarization, it kills the life out of the scene by deleting too much light especially on the foliage and grass, even while saturating colors at 100%. Which type would be better for me, and where do they sell them?

I think most common polarisers are somewhere in between these two types. If you never use full polarisation then the high transmission type should be good for you. I've seen those at quite a few places including Amazon, e-bay & some camera shops IIRC. The high extinction type might only be via technical places like Edmund Optics.

For some of the special effects I sometimes shoot higher extinction would probably be better though I use a normal polariser carefully crossed to a laptop screen (polarised light source) for this:
stressed stencil by Mike Kanssen, on Flickr

I also use them to boost reflections, almost as much as to kill them yet hardly ever for more normal scenes where intermittent positions would be preferable.

 

[ByRiker'sBeard6]

I think most common polarisers are somewhere in between these two types. If you never use full polarisation then the high transmission type should be good for you. I've seen those at quite a few places including Amazon, e-bay & some camera shops IIRC. The high extinction type might only be via technical places like Edmund Optics.

For some of the special effects I sometimes shoot higher extinction would probably be better though I use a normal polariser carefully crossed to a laptop screen (polarised light source) for this:
stressed stencil by Mike Kanssen, on Flickr

I also use them to boost reflections, almost as much as to kill them yet hardly ever for more normal scenes where intermittent positions would be preferable.

I have been wanting to photograph Christmas lights at night but find that a long enough exposure to get them bright enough to pop also blows them out to white. I've been wondering if I could simply use primary color filters to block anything too far outside the range of a given light, and from 3 photos combine color channels manually.

 

[Rickbb69 8]

I have been wanting to photograph Christmas lights at night but find that a long enough exposure to get them bright enough to pop also blows them out to white. I've been wondering if I could simply use primary color filters to block anything too far outside the range of a given light, and from 3 photos combine color channels manually.

You would probably do better with image stacking in post. 1 exposure for the lights, one for everything else then combine/stack them.