I have a noob level question about RGB images- Pixels in RGB images are in the range of (0-255), so technically each channel ranges from 0(black) to 1(white), then why we say that each image has a Red, Green, Blue channel component, although they are only Gray channel. In short, how the image is formed?
-
1$\begingroup$ The red, green, and blue channels aren't grey. They are, respectively, black to red, black to green, and black to blue. Hold a magnifying glass up to your (LCD) monitor and you'll see the individual channels. $\endgroup$– Simon FCommented Aug 29, 2019 at 15:31
-
$\begingroup$ okay, I have another doubt, maximum of each is represented as 255 then what mystery goes behind each channel that every 255 corresponds to a different color. $\endgroup$– MarkCommented Aug 29, 2019 at 15:36
-
$\begingroup$ I don't understand your question. $\endgroup$– Simon FCommented Aug 29, 2019 at 15:38
-
1$\begingroup$ Do you mean, given an RGB triple, {128, 35, 199}, which is a medium bright purple, why does the 128 affect red? Because thats how the hardware is wired up. $\endgroup$– Simon FCommented Aug 29, 2019 at 15:54
-
1$\begingroup$ Yes @SimonF, I meant that only. $\endgroup$– MarkCommented Aug 29, 2019 at 16:12
2 Answers
Each channel ranges not from black to white, but from off to full intensity. What colour is meant by full intensity is down to convention and the presentation technology.
For example, a given image file format might say there are three bytes, of which the first is the red channel, the second is the green channel, and the third is the blue channel. There are many different colours, so to be a real standard, it has to say exactly what that "red" means, usually by reference to a particular colour standard such as sRGB or CIELAB. These international standards provide an absolute way to describe colours.
After some processing, your red, green, and blue values find their way to your LCD monitor. There, the electronics uses them to control the intensity of three sub-pixels: one of which is a red filter, one a green filter, and one a blue filter. The 0-255 value in the red channel is turned into some voltage which controls the transparency of the liquid crystal behind the red filter, turning it from completely opaque at 0 to lets all light through at 255. Since this is attached to a red filter, it contributes black or red to the image.
The same idea is everywhere in computers. The only things inside a computer are numbers. They only mean what we say they mean because of conventions (all software interprets them the same way) and/or because they control some values inside some output hardware. e.g. one number means the red channel in a pixel on your screen, another number means a letter of the alphabet in an email, another number is a pressure sample in an audio file which will be sent to your sound card and control the voltage to a speaker, and so on.
CRT monitors are a thing of the past, but I think they are the best example to describe to you "how the image is formed". Each pixel on the monitor is actually a group of 3 phosphors. A green phosphor, a red phosphor and a blue phosphor. These phosphors will glow at a certain intensity when they are charged. Inside the monitor there is an electron beam that charges these phosphors. The intensity of which, is determined by the number 0-255 of that channel. So a completely black pixel would not be hit at all, whereas a bright red pixel would only strike the red phosphor, and at a high intensity (around 200-255). This is, of course, a simplification of the process, but it answers "how" the red channel relates to a red pixel.
-
$\begingroup$ "bounced off a mirror"? IIRC there is no "mirror" in an CRT. To control the direction of the beams there are charged plates at the back of the 'tube'. A 'shadow mask' at the screen end stops electrons intended to illuminate, say, a red phosphor from iluminating the green or blue one. see. e.g. image.slidesharecdn.com/… Please correct your post. $\endgroup$– Simon FCommented Sep 5, 2019 at 9:44
-
$\begingroup$ Please look up reflective vs refractive CRT and specifically the Schmidt optical system. Also Texas Instruments had a micro-mirror CRT that was released using mirrors rather than electrostatic deflection plates. Either way, I don't feel this aspect of the discussion helps the original poster. $\endgroup$ Commented Sep 5, 2019 at 22:46
-
$\begingroup$ "Texas Instruments had a micro-mirror CRT that was released using mirrors " Do you mean the DLP technology ti.com/dlp-chip/getting-started.html ? IIRC that wasn't a CRT - it was all optical. $\endgroup$– Simon FCommented Sep 6, 2019 at 13:35