jaxonspictures PRO 4:30pm, 7 May 2009
This was part of my reply to another post / topic in another forum, but I thought it might get lost so I am posting it with a new topic. Just wondering if my understanding of the basic physics of white balance is correct.

Somebody correct me if I'm wrong, but aren't you effectively increasing the ISO, and therefore noise, for a particular color channel by using a white balance that adds more of that color to make up for the lack of it in the ambient light? I mean, if you are shooting under incandescent lighting, the white balance increases the blue gain to make up for the lack of that color from the lighting, right? If that is the case, you are turning up the camera's sensitivity to that color and therefore adding more noise in that channel.

I notice this when I take pictures using a bic lighter for my key light. The flame lacks blue, so when I either manually white balance or correct for it in RAW, I end up with alot of blue noise.

I guess what I'm asking is: Can you really say that you used, say, 100 ISO when your white balance had to correct for lack of blue in the lighting and was effectively working at 800 ISO on that particular channel?
agnisflugen 9 years ago
i've struggled with ISO and white balance issues too, particualary when i try to shoot in black and white, so i really look forward to a response to this question....
Damien Franco PRO 9 years ago

Okay, that's just me answering the question of

"Can you really say that you used, say, 100 ISO when your white balance had to correct for lack of blue in the lighting and was effectively working at 800 ISO on that particular channel?"

You can, in my opinion, say that you did use ISO 100 even if the camera is correcting the WB.

The real question may be whether or not the ISO is the same.

I think there is a lot of grey area here. In the old days of film, especially 35mm, most photographers rarely had their exosures spot on. Plus the lab did the color correcting, not photographers (mostly), but the ISO (ASA) of the film, which is effectively the light sensitivety and correlates to the amount of grain (or noise in digital), would not actually change.

I guess what I'm saying is that there really shouldn't be a difference in how we say that an image came to it's fruition. If we were still shooting film we wouldn't say

"I shot this image with film speed 200 but underexposed by two stops. Plus I was using daylight balanced film (5500K) when I should have been using tungsten B (3200K). Then the lab probably had to color correct for the bright red curtains that were throwing red reflected light all over the place."

Sometimes I think that having too much metadata (or information in general) can actually tie you down.

As long as the images look good in the end, and you didn't push the pixels past their boundries, the image was shot at whatever ISO the camera was set to.

Just my thoughts.
jaxonspictures PRO 9 years ago
Damien, thank you for your thoughtful response!

I agree that if the image looks good, it is good. I apologize for a rambling question and apreciate your simple....yes.

I have found through my research that ISO is a number used for calculating proper aperture and shutter speed in order to produce a fully exposed photo. What happens after the sensor records it's info has nothing to do with the ISO.

What I was trying to get at was that there is no free lunch when it comes to white balance. If there was a severe lack of a certain wavelength of light when the photo was taken, you will pay with noise in that wavelength when you correct the white balance as the camera (or the image editing software) is increasing the sensitivity or gain at the wavelength that is lacking. If you are already pushing the camera's limits by taking a very low light photo in something other than perfectly white light, you will raise the noise to a visible level in whatever portion of the spectrum that was lacking and therefore corrected for with white balance adjustment.

What I would be interested in knowing is what color temperature would give the sensor the best signal to noise ratio in all three channels. Is it 5500K? Or is it something else? Or does it vary between camera and sensor designs?

I'm rambling again. Sorry!

Thank's for the consideration of my musings!

Damien Franco PRO 9 years ago
My quick guess on research I've done in the past is that the "best" signal to noise ratio will vary considerably by manufacturer and model.

Cameras can't do everything we wish they could all the time. That's not to say that I think you're asking for some sort of miracle by any means.

It seems you're just looking for what most people are looking for.

How do I get the most optimal image from my camera?

Find out where your camera handles it's best ISO.

Most photographers think that it's at the lowest ISO but that's not actually the case. Some cameras actually handle better at 200 ISO than at 100 ISO or even 50 ISO.

After that, it's just a matter of trial and error. It takes many many images and hardcore pixel pushing and examination to truly understand just how far you can comfortably push your camera.

Truthfully, the color tempature is probably the easiest question to answer, even if the subject itself is very complex.

If you prefer warm tones then you should initially capture your images at a warmer Kelvin. Otherwise you'll be compensating and manipulating the images in post processing to get the look you wanted to begin with.

Truth be told, I think that the fewer steps that pixels have to take to go from capture to presentation the less artifacting and anomoly you'll get. This will typically also translate to less noise as well.

If noise becomes an issue because you're forced to shoot under sub-optimum lighting and you are pushing your camera to it's limits then a noise removal software may be what you need. This is not a bad thing. Many top professionals live and die by software programs like Noise Ninja.

We have many tools at our disposal as photographers and its up to us individually to use what we can to get the results we're looking for.
jaxonspictures PRO 9 years ago
Do you have personal experience with Noise Ninja? I have been considering buying it but am not sure where in the "flow" to use it.
Damien Franco PRO 9 years ago
I've tested it but don't actually use any noise reduction software. It's just in the workflow for the type of photography that I do.

However, I have seen several great reviews about their latest iterations from very trusted sources.

If I remember correctly, it's used at the very end, just before print.

I do think that depends on how heavy your workflow is. I've heard of people doing crazy things with Noise Ninja, like separating chanels and running the software, then recombining the channels, then running Noise Ninja again. I'm not sure how that works out, and for what purposes that very extreme amount of pixel pushing is used, so as always, procede with caution.

Here's an early review by a very trusted source on digital photography. www.luminous-landscape.com/reviews/software/noise-ninja.s...
Groups Beta