A suggestion of a topic: what is "white"

Before explaining white balance, is it worthwhile to explain what is white (and neutral, in general)?

Depending on how far down the adaptation rabbit hole you want to go, that could get pretty complex. Staying with just enough to talk about white balance (which can also get complicated, since there are so many algorithms for changing shite balance, but I'll ignore that here), a sufficient condition (but not a necessary one) for unmixed lighting situations is equal reflectivity at all visible wavelengths. If cameras were Luther-Ives devices, you could say a reflectance spectrum that, when illuminated with the single illuminant for the scene, has the same chromaticity as the illuminant itself. But cameras aren't, so that gets a bit messier.

There's a reason why good gray cards and scanner calibration chips aren't cheap.

Does a beginner need to know any of this? I doubt it. If it looks gray to the beginner, it's probably good enough to white balance off of. The beginner isn't going to be looking for wonderful accuracy.

I was thinking understanding "white" helps to understand "why white balance".

That would be whine/neutral reflectivity. What is white (source of) light?

Human adaptation is not total, but the convention is to consider that the illuminant defines the white point.

A halogen light in a studio appears to be white. When the same halogen light is lit outdoors in broad daylight, it appears to be yellow ;)

I'm not sure this exactly fits under this topic, but it's always irked me that whoever first designed WYSIWYG word processors chose "maximum white" as the default background to look like "paper". As a photo and video hobbyist, the maximum output of the display should IMHO be reserved for image highlights, and a screenful of that is way too bright. I try to set the UI on all the programs I use to a dark background (keeping my fingers crossed, DPRevived!) but I'm still not able to set my monitor as bright as I'd like to for photo and video viewing.

When people compare computer monitors and TV screens, I think the white level is the major difference between them. There's no technical reason why that needs to be the case, the fault lies squarely on the shoulders of that anonymous software developer....

Right, hence my question. Of course, one can always say - if it looks white, it is white since color is a perception...

From time to time when I say that in a lecture the audience doesn't accept it and asks for a demonstration.

And Xerox, with the Star, in 1981, predating the 1983 Lisa introduction. The Star and Lisa were both unsuccessful, however, and for much the same reason.

As the there is no color in pure 255/255/255 white, I’d say yes. For white balance you need viable color in all three channels to do any “balancing”.

There is indeed color in 255/255/255 on a monitor. The color is the white point of the monitor. Perhaps you meant in a raw file, but the precision of virtually all raw files is, thankfully, not 8 bits.

Only if that can be done in one paragraph of conversational English.

Personally, that's my goal when responding to a beginning photographer's question. Explain each concept in a single paragraph of conversational English or, better yet, a single sentence. I'm marginally successful at that. I do well at keeping a post conversational but can often get a bit wordy.

As a photographer gains experience and develops an interest in acquiring a deeper, more complete understanding of a concept, that's when more complex, layered, and precise technical discussions become essential. But as a general rule for the questions that come up in the Beginners forum, I recommend keeping posts conversational, broadly accurate and focused on the specific question.

If the post requires more than 4-5 sips of coffee to read, it's probably too long. I speak from personal experience on this point :)

N/N/N is white only for a colour-neutral space. Today's cameras record RGGB raw data in a non-neutral colour spaces (YCC pseudo-raw is a different beast, it is pre-balanced).

Good point, I was just thinking from a WB shooting/WB standpoint. I stand corrected.

Here's a start that I wrote for the Lensrentals article:

Human vision evolved when it was a matter of survival to be able to recognize a predator or a potential meal regardless of the lighting. One of the effects of this is that people unconsciously tend to see objects whose reflectance spectra are the same as the same color, even if they are illuminated differently. This effect is called color constancy. It’s not perfect, and it’s influenced by many things, but there’s one aspect to it that we can take advantage of to make our images look more like the real world.

When a scene is illuminated by single-spectrum light – think sunlight, electronic flash, candlelight, etc. – people adapt to the illumination, and, within limits, the colors of objects appear the same regardless of the illuminant. In general, we view images in a different state of adaptation than we’d experience in the original scene. Thus, in order for the colors to appear realistic, we need to change them to account for the difference in adaptation. It’s sort of a paradox: it is necessary to make the colors wrong in an absolute sense so that they appear right to the viewer. The simplest form of this operation is called white balancing. There are many ways to perform corrections for viewer adaptation, such as XYZ Scaling, induced opponent response, Lab shifting, and Von Kries. The algorithm that I prefer is called Bradford. I vaguely remember from somewhere — Eric Chan? — that Lightroom uses Bradford.

The last two sentences are technical and could be eliminated.

Thanks, some of that went so far over my head, my hair didn’t move. I’ll bow out of this one and just read. :)