CCD: The heart of a digital camera (how a charge-coupled device works) 

Flying butt monkeys create rainbows in the sky as they dance in rhythm for the pleasure of their god, the Flying Spaghetti Monster. This sounds implausible, but you've seen the results yourself.

"if it didn't happen, you wouldn't see me in colors right now" lol

They had that for a while, but the imps evaporated, and kept spilling their paint if the camera was jostled.

So that's where all the unemployed Keebler elves go. I was wondering why my DSLR smelled like a shortbread cookie.

IMPS? RIP AND TEAR WITH YOUR CAMERA

It's just magic, plain and simple. Just like have airplanes fly.

@@kahlzun You clearly know more about the nature and working of imps than you should do...

😅😅

Co-inventor sorry. George Elwood Smith (born May 10, 1930) is an American scientist, applied physicist, and co-inventor of the charge-coupled device (CCD). He was awarded a one-quarter share in the 2009 Nobel Prize in Physics for "the invention of an imaging semiconductor circuit-the CCD sensor, which has become an electronic eye in almost all areas of photography".

Do you realizr it's sped up Edit: I was pretty much a dumbass kid back when I made this comment

*realize

@@MigTheFalz Do you realize it's a joke? Well, of course you don't, dummy!

It's whatever, it is of course sped up with the magic of editing

Sorry about that though, let's just say that it's some kind of a joke.

(1/2) I am currently reading his book in which he illustrates it and details it a little bit more so here it goes... From a simplified perspective, the key to moving the electrons is the charge of the MOS (metal oxide semiconductor). What's really amazing about a CCD is that it alternates a charge of about .5 and .10 volts between each pixel and MOS, as it does this it changes the actual location of where the electrons hang out within each row of the silicon slab... A higher voltage means that the electrons are closer to the surface, while a lower voltage (.10 V in this example) gives room for electrons to fall to the base of the silicon to let protons take their place at the surface (since we know light is both a partial and a wave, we know that both of these items would exist within the Silicon after it has been exposed to light to take an image). The key is the voltage fluctuation, as the voltage to the adjacent row falls to match the one next to it, the electrons now have an option

(2/2) of where they want to hangout BETWEEN pixels as apposed to just where in the silicon they can hangout within a SINGLE pixel. But this doesn't last long, as the CCD now makes the charge of the old location a higher voltage and forces the electrons to stay in its new found location on the CCD (which is just now one pixel or row over to the right, if we imagine the flow in the video). That's it! Now just repeat and repeat this flow of voltage fluctuations enough, and the electrons end up at the end of the CCD and read by the camera as a measure of light exposure, and thus an image is created. Just imagine you put a marble on top of a piano key and you wanted to move the marble down the piano without ever touching it, to do so you would press down the adjacent key and lift the first key the marble was on to force it over one, and then just continue that process. That's what a CCD does except the marble is an electron and the piano key is the CCD's individual rows. Hope this helped!

+Dana Ludwig Yes he seemed to skim over a couple of things in that one. How does it shift the charges down and how does it know what colour the adjacent pixel is?

Adjacent pixels colour is known because colour filter array geometry is known. It's simply hardcoded into image processor.

Short version: Those aluminium electrodes over the top can create an electrical potential barrier, a wall which separates pixels, but this 'wall' can be raised or lowered depending on whether you apply a voltage or not. Each pixel spans three times the distance between electrodes. Electrodes are connected so that every third one is connected, so we have three sets of electrodes, say A, B, C. Now use capital letters to represent the 'wall' state, lower case to represent no wall. We start in state Abc (i.e. AbcAbcAbcAbc...) Charge can accumulate between the walls, in the 'bc' space. We hold it like this during exposure. Now to read out we cycle the voltages on the electrodes. We shift to state ABcABcABC... which shifts the charge which was in 'bc' spaces all into 'c'. Then we go to aBcaBcaBc (charge shifts to 'ca' spaces) then aBCaBCaBC (charge in 'a' spaces) then abCabCabC, then AbCAbCAbC then back to AbcAbcAbc and the charges have all shifted rightwards one pixel. At the end of each row you catch these little buckets of charge and send them off to analogue-digital converter(s) to turn them into numbers. If you overexpose, the potential barriers caused by the electrodes are insufficient to contain the charge, and it leaks into adjacent pixels. The barriers in the perpendicular direction are caused by doping in the silicon and are much more resistant to leakage. This means overly bright lights cause streaks in one direction (e.g. vertical) but not the other (horizontal.)

I've seen the video of the trillion frames per second camera. Technically speaking, as far as I understand and remember, that footage of the light beam is a "composite" of multiple shots. Obviously the result is still incredible, but it's not really film once and play back, from what I remember

Feeling confused is a good thing! Feeling like you know everything is what you should worry about

Yeah, at first glance I thought "oh cool! Mark Hamill is going to teach me about cameras!". Great video though, even with the lack of joker's voice :)

Mark Hamill was on his way to Tosche station to pick up some power converters.

That happens to me.

hahhaahaha

Welcome l hope help me orphansneedWe putdifficulthe havechildrenWe needyesterdayno There isfor us housing livingtentmy seekersThe helphelpsickneed medicineuntilinsurancehealthyCut offAbout MeI hope helpI have Proof my wordspay when God

I think it may have been simplified in the video for the sake of brevity.

A CCD reads out easy line at once, while a CMOS reads each line one by one in sequence. There are other differences but I believe this to be the most notable one.

vx1000 best 3ccd camera

Welcome l hope help me orphansneedWe putdifficulthe havechildrenWe needyesterdayno There isfor us housing livingtentmy seekersThe helphelpsickneed medicineuntilinsurancehealthyCut offAbout MeI hope helpI have Proof my wordspay when God

There's a bit more to it: luminance is not completely preserved because each color filter (rgb) blocks the light/luminance that comes with part of the spectrum that's blocked. This is the reason why the grid has twice as many green pixels than reds and blues (2/4 green, 1/4 red and 1/4 blue): the human eye is more sensitive to green, so in order to preserve luminance the most, green must be captured in more detail. Neat++

When I was an astronomy student c1990, we had engineers in our department building a CCD camera. Some of the CCD detectors cost over $100k each. (If you want to see the results of this camera, do a web search on Sloane Digital Sky Survey. The UV sensitive CCDs were the very expensive ones.)

CCDs revolutionized astronomy. You could easily do digital processing and measurement and automate such measurements, but more importantly CCDs are about 40 times more sensitive than ordinary photographic film, and about 5 times as sensitive as super-sensitive-bizarrely-processed astronomical film. Suddenly you could do with a 1m telescope what had previously been done with a 2m or 4m telescope.

​@@michaelwoodhams7866 Welcome l hope help me orphansneedWe putdifficulthe havechildrenWe needyesterdayno There isfor us housing livingtentmy seekersThe helphelpsickneed medicineuntilinsurancehealthyCut offAbout MeI hope helpI have Proof my wordspay when God

Just making a guess: Probably because the images are rectangular. No rearrangement process is required this way.

Display pixels are completely different from camera sensor pixels. It's easy to estimate that the pixels are rectangular because the images that come out of the camera are rectangular, they're actually squared. Al of the photosites on a camera sensor have the same dimensions across.

This is not true. High-end digital TV and cinema cameras all use the exact same bayer color filter array as the cheapest flip phone cameras.

It's pretty cool that the CMOS is actually going backwards to the y/x flow of electrons that he said didn't work... If I'm not mistaken, it's because this method now uses an Active Pixel Sensors assembly to fluctuate circuit flow to overcome capacitive coupling... Which is really cool and can give us even more pixels when an image is taken!