How do it know?

[Doug_Kerr]

Hi, Drew,

The Color Parrot has a catchy name, and sounds like a nice concept.

But I just can't see how it can work. What we need to do in our white balance reference shot is catch the chromaticity of the ambient light on the subject. That's why we sometimes use a gray card at the subject location, and other times use an ExpoDisk and, at the subject location, aim the camera so that the face of the disk is parallel to the principal surface of the subject (I know it is rarely said that way, but that is what theoretically produces the most appropriate result).

How can a camera equipped with a Color Parrot, or any other instrument, by regarding the scene, discern the chromaticity of the ambient light? To take an extreme case, suppose the entire scene is a red barn. Clearly the light accepted by any apparatus that regards the scene will have a chromaticity affected by the reflective chromaticity of the barn. (That's why it "looks red" to us.)

In any case, the light perceived by the "instrument" has a chromaticity that is influenced by the overall reflective chromaticity of the scene.

The automatic white balance facility of our cameras attempts to overcome this by making intelligent assumptions as to the range of reflective chromaticities in the scene. Often it guesses well, many times not.

But clearly the Color Parrot cannot engage in any such decision making.

So, what am I missing? How does the Color Parrot overcome the problem that the light it receives from the scene is, for each area within the "view" of the device, has a spectrum that is the product, over wavelength, of the spectrum of the ambient light (which we would like to evaluate) and the reflective spectrum of that scene item?

What we need to do in our white balance reference shot is catch the chromaticity of the ambient light on the subject. That's why we sometimes use a gray card at the subject location, and other times use an ExpoDisk and, at the subject location, aim the camera so that the face of the disk is parallel to the principal surface of the subject (I know it is rarely said that way, but that is what theoretically produces the most appropriate result).

Thanks for the compliments. Great to see you back after so long.

I'm a bit confused by your post here. It seems you are saying that the ExpoDisk type concept is a functional way of achieving a white balance reference shot. Is that accurate?

[Kiran Thind]

Doug,

Well said! Your overview of automatic white balance establishment in digital imaging systems is the best I have seen.

Drew,

To establish a white balance reference, we need to know the spectrum of light illuminating each area in the captured scene. The only way you can do this from the camera position is by reflecting the illuminant with a spectrally neutral object (grey card) placed at the position of the subject. The more direct method is to use an ExpoDisc type product at the subject location as Doug detailed above. The only problem with the ExpoDisc is that it is not spectrally neutral (they attempt to compensate for the short wavelength absorbing (purple, blue) white substrate by adding a blue filter).

Kiran

[DougAxford]

My best guess on this Doug is that this device great for many applications, but not for some.

I'm thinking of particular uses when I can't easily do a white/grey card and I'm not sure enough to guess on the K. It should work similarly to a color temperature meter or a AWB in a camera except (for me anyway) I do not want the the camera to change WB with every shot.

Doug

To establish a white balance reference, we need to know the spectrum of light illuminating each area in the captured scene. The only way you can do this from the camera position is by reflecting the illuminant with a spectrally neutral object (grey card) placed at the position of the subject.

Um, I appreciate you guys summing up white balance for us from a more technical standpoint. However, we've already been over this ad nauseum elsewhere. If you are willing and able from a practical standpoint, I would suggest the use of a gray card as most likely giving the absolute most accurate reading.

The situations and target market for this product are clearly outlined in the article and here in the forum.

Most live events do not, in any way, lend themselves to the use of a gray card.

With this product you will save time and you will get much closer than you will using awb the vast majority of the time. Those are just the facts.

[robertburkhalter]

Drew,
I applaud your last statement. It in no way suggests that your product (pointed at the subject) produces a superior white balance to the Expodisc (pointed at the light source.)
Good, clean statement.

However, when you say "most live events do not, in any way, lend themselves to the use of a gray card", I start to think, "so what, I've got the Expodisc that, when pointed at the light source, spins rings around the parrot". by the way, I like the name parrot. It's cute (this is a compliment).

[Doug_Kerr]

Originally Posted by drew

I'm a bit confused by your post here. It seems you are saying that the ExpoDisk type concept is a functional way of achieving a white balance reference shot. Is that accurate?

The ExpoDisk concept is great if we place the "instrument" (that is, the camera with the ExpoDisk on the lens) at the subject position with the disk parallel to the subject surface, so that it receives light just as the subject will.

No "instrument" will theoretically make the determination we seek if put at the camera location for the shot and aim it into the scene.

Many instruments may give a good result that way, but this is a triumph of good luck over physics.

Best regards,

Doug