Teacher suggests warming filter

Scatterbrained wrote in post #16672590
Instead of using auto WB use the eyedropper tool on the shirt. I'd imagine they'll be quite a bit closer.

I used the Auto thinking that way there wouldn't be any 'user variation', and just let the software decide.
__________

Isn't the whole point of this thread whether or not AWB will negate the effect of a warming filter?

I would think that adding a yellow 81 filter, or using a lens that has a yellow cast would be pretty much the same thing, right? I'm still going to see if I can find a Yellow 81 laying around, and I'll do a comp tomorrow. <shrug>

I can now state to my own satisfaction anyway that AWB does not correct the effects of colour altering filters completely if at all. Here are two images, shot within seconds, the only salient difference being the presence of a deep blue 80A filter on one, while AWB was set for both and shutter speed and aperture held constant, while allowing ISO to vary to determine exposure. The second one does look relatively underexposed to me, but that may indicate some colour sensitivity in the exposure assessment. In any event, the blue cast due to the presence of the filter is obvious, despite any attempts the AWB algorithm may have made to correct it. I would expect the same to hold true for any colour-altering filter.

You can add warming or cooling filters in Photoshop. Here's a screen shot showing a warming filter being added.

xarqi wrote in post #16672663
I can now state to my own satisfaction anyway that AWB does not correct the effects of colour altering filters completely if at all. Here are two images, shot within seconds, the only salient difference being the presence of a deep blue 80A filter on one, while AWB was set for both and shutter speed and aperture held constant, while allowing ISO to vary to determine exposure. The second one does look relatively underexposed to me, but that may indicate some colour sensitivity in the exposure assessment. In any event, the blue cast due to the presence of the filter is obvious, despite any attempts the AWB algorithm may have made to correct it. I would expect the same to hold true for any colour-altering filter.

Thanks for that xarqi, it confirms what I thought.

We all know all sorts of filters and effects can be achieved through post processing, but it comes down to whether you prefer to do it in camera or after the fact. Personally, I'm of the camp of get the exposure, framing, composition in camera the way you want it. Then play with it in post for other aspects.

Not every portrait is going to benefit from a warming filter, and if it doesn't give you what you want SOOC, it'll be that much more to correct in post. Besides, doing it in post allows for much more creativity and subtler changes, depending on personal tastes.

xarqi - Have you tried to correct the second shot in post? How well does your software do, removing the filter's effect? Could you post that too?

KirkS518 wrote in post #16672724
xarqi - Have you tried to correct the second shot in post? How well does your software do, removing the filter's effect? Could you post that too?

I'll get back to you on that - a bit of a logistical hassle just at the moment, sorry.

KirkS518 wrote in post #16672522
I understand that, but from what I'm gathering from the posts, is that AWB will correct any color cast from a filter (warming/cooling or whatever).

Not all colour casts can be corrected by adjusting colour temperature, or even by also adjusting the "tint" slider which is usually found in the same panel of raw converters. The AWB algorithm may also have sanity checks on how far it will move those sliders (as xarqi appears to have shown).

However, it should be possible to correct, to a good approximation, a good 81A filter by just changing the white balance slider. The correction won't be perfect, because dyes were used to make the filter; see, e.g., the transmission curve here.

So I got curious about the effects of a colored filter and the ability of AWB and/or post processing to eliminate its effects. I don't a warming filter, but I do have some gel sheets large enough to cover the lens.

Setup:

5DIII with a 50mm f/1.4 on tripod
580EX II mounted on the hotshoe, set to 1/4 power, bounced off a white ceiling.
Exposure set manually to 1/160 sec, f/5, ISO 400 per Sekonic light meter. I metered in front of the middle car.

I shot the scene without any gel, with a 1/4 CTO, and a full CTO. One shot of each was with the camera manually set to 5500K and one with it on AWB.

Looking at the LR previews, it was obvious that AWB didn't correctly compensate for the presence of the gels. In fact, the camera went the wrong direction, increasing the temperature instead of lowering it.

After correcting each shot with the eyedropper, and increasing exposure on the gel shots to match the non-gel shot, there's still a difference in color rendering. The green seems to suffer most while the red looks more orange in the shots through the gels.

3213- No Gel

Here' the reworked image taken through the 80A filter. I used the LR temperature eyedropper tool on a fairly neutral spot (I didn't think to include my Spyder Cube in the frame, alas). It maxed out at 50000 K - not quite enough. I bumped the exposure by about a stop. It's a pretty decent correction, but not perfect. Somebody with a Macbeth chart and some filters could doubtless do it definitively.

Oh - I also did bump the vibrance and clarity a bit - force of habit I'm afraid.

John from PA wrote in post #16672383
People should be more open minded as opposed to saying every person's opinion, including this teacher, is wrong. Each of have different likes/dislikes in how an image should look. It is entirely possible that some people might prefer a warmer look, who are we to say that is incorrect? I'm paraphrasing a bit, but Moose Peterson, an accomplished landscape and wildlife photographer, once stated if you can do something "up front", as opposed to post processing, then do it up front. This is one of the reasons he developed the Hoya Moose Peterson circular polarizer, which I personally prefer for landscape work.

I agree. These issues are not entirely black and white, right or wrong.

melcat wrote in post #16672771
Not all colour casts can be corrected by adjusting colour temperature, or even by also adjusting the "tint" slider which is usually found in the same panel of raw converters. The AWB algorithm may also have sanity checks on how far it will move those sliders (as xarqi appears to have shown).

However, it should be possible to correct, to a good approximation, a good 81A filter by just changing the white balance slider. The correction won't be perfect, because dyes were used to make the filter; see, e.g., the transmission curve here.

This was my initial thinking, although I'd still prefer to do it in post.

Also maybe the students are shooting .jpeg and although I would have thought AWB would compensate somewhat I guess it doesn't.

So maybe the teacher did know what she was saying? Although it's not really the way most people would go about it with digital.

I am not sure why we are limiting this discussion to auto white balance. If you want the image warmed up a bit, switch over to the better white balance tool - degrees kelvin, represented by the "K" in the camera's white balance options. If you want something warmed up a bit, just crank up the Kelvin temperature by a few hundred degrees. No need to use a filter. And no need to worry about what auto white balance may or may not do. Auto white balance is fickle at best. If you're trying to get the picture the way you want it in the camera, don't you want to take control of your white balance, rather than letting the camera pick what it thinks you want?

Auto WB = often inaccurate
warming filters = unnecessary

SkipD wrote in post #16671813
You are correct.

The teacher is ill-informed for the digital world.

This. And you should get rid of your other useless "protection" filter.

Tom Reichner wrote in post #16674990
I am not sure why we are limiting this discussion to auto white balance.

I guess any limitation is only because it's virtually a given that in any mode except AWB, a colour-altering filter will have an effect on the image, whereas with AWB it was uncertain, with theory suggesting it would not, but experiment suggesting it does.

Mike_d,

You don't get an average WB with an on camera flash. The camera knows the color temperature of the flash and sets the WB for approximately 5000. The setting varies slightly based on flash duration. The camera doesn't actually measure the reflected flash nor is it aware if the flash is gelled.

Bearmann wrote in post #16675090
Mike_d,

You don't get an average WB with an on camera flash. The camera knows the color temperature of the flash and sets the WB for approximately 5000. The setting varies slightly based on flash duration. The camera doesn't actually measure the reflected flash nor is it aware if the flash is gelled.

I believe the camera only does that if the WB mode is set to flash in the camera. Also, I had the flash power locked down so if it were setting the WB based on flash power, it wouldn't have changed as it did. I did not gel the flash itself, but held the sheet in front of the lens.