Recently i’ve been looking at buying an LCD monitor.
I interested to know what causes dead pixels and if there is a way to prevent them.
B_Black
Interesting that you are asking this. It’s a misnomer. LCD’s don’t have “pixels”, they have liquid crystals. You tap an LCD hard enough, and you can shatter the clear cover. Or, the facing of the crystal matrix. if you have an older LCD that lacks the facing, you can burst individual liquid crystals.
As far as blown pixels? The camera I use all the time at the studio has a blown pixel. I always thought it was a physical issue with the CCD’s, but I come to find out at the N.A.B. show in Vegas last month that in fact you can- with some broadcast cameras- wipe the CMOS clean and indoing so, reset the entire chip array to their factory settings. In doing this, you can make “blown pixels” go away.
If you go to Macie Video’s website, you can contact Roger Macie and ask him all about it. Fascinating stuff !
Cartooniverse
Eh, in what way is it a misnomer? A pixel is a picture element, and the term equally applies to computer monitors, stored images, and CCDs. A dead pixel on an LCD is one that is permanently black. At least that’s what wikipedia and I think.
Neither of us can tell you exactly what’s wrong with it though.
Of course they have pixels. The pixels are created with liquid crystals. Simplistically speaking, a set of thin horizontal conductors are “drawn” on a glass plate, and a set of similar vertical conductors are drawn on another (if you look at an LCD watch or calculator screen at the right angle, you can see these conductors). The liquid crystal material is sandwiched between them, and the pixels are created when a voltage is applied across one pair of the conductors. A polarizing filter makes this visible. It’s much more complex in modern color hi-res LCD screens, but this is the basic mechanism by which they work. Dead pixels are usually caused by a fault in the driver circuitry which results in either a permanently “lit” pixel, or one which is permanently “dark”. Most LCD monitor and laptop manufacturers have included in the warranty a set number of dead pixels up to which will not be considered “defective”.
I’m not sure what exactly causes dead pixels, but if you have a monitor with one, you can sometimes (gently!) massage the area and restore the pixel to functionality. (I’ve done it myself with some success).
Cartooniverse: are you sure they’re ‘resetting’ the CMOS? I haven’t heard of this before. My Olympus D40 has a feature to improve long-exposure shots… in dark conditions and with long exposures (mine can do at least 13 seconds), ‘hot’ pixels (example) show up as pink or green dots.
By turning on the CCD for a while with the shutter closed, these pixels can be identified and later ignored, improving the results of night time shots.
Q.E.D., we are both right. Yes, LCD’s have a representational dot that can be considered a “pixel”, but in terms of the world of cameras and imaging, it is not a pixel in the traditional sense at all. However, if one could reset the drivers just as one can in the example I provided, one could clean up blown “pixels” in an LCD display. This would probably thrill owners of gargantual LCD 16x9 monitors who find that suddenly they see a few random missing dots, or dots of other colors.
Oh yes, Nanonda. That is exactly what they are doing. It turns out that on cameras, the so-called “blown pixel” is a small fault in the CMOS much more often than it is a damaged CCD. ( By comparison, you could take a bad “burn” in a tube camera but if you faced the camera to a 100% white panel for a few hours, frequently the “burn” would fade away. )
These folks apparently have great success in re-setting the CMOS’s but only for a specific set of broadcast cameras. Not all of them can be re-programmed. For many, a tiny bit of information is always ferschtummelled after a given failure of a small bit of data reading ability, and that shows as a blown pixel. The cameras at work at Sony, and so I will recommend that we send out the bodies over the summertime ( when we are dark ) to be re-set.
I think the ‘drivers’ that QED mentions are the actual transistor circuits that cause the LCD pixel/element to change state; if an individual transistor somewhere in the vast array simply fails (or never worked in the first place), then there will be a dead pixel that no amount of resetting will remedy.
The traditional sense is that of computer graphics and computer displays. Its use in the world of cameras and imaging came later, with the rise of the digital camera.
No offense, but you’re completely wrong. Television cameras pre-date the use of computer CRT displays by… decades? It is conservative to say that television cameras existed and were in regular daily use by 1948. First documented broadcast of a scheduled event? I’ve got November, 1930, when
It is even more conservative to say that computer CRT displays- as used in home personal computers- did not exist in any kind of numbers until the advent of the Scelbi, arguably the very first home computer kit sold. That was 1974. ( And a few months before the Altair Computer- so named by the inventor’s daughter who heard the name of a planet on the original Star Trek series.)
You do the math. It is 44 YEARS from 1930 to 1974. Despite what computer wonks wish to believe, the TRADITIONAL SENSE of discussing pixels that are blown out or not reading/scanning originated solely in the realm of broadcast television cameras, NOT computer displays using CRT’s. Pixels are and have been the representational unit of measure of display information since the 1920’s on cathode-ray tubes. ( I’m not sure that the rudimentary “television” spinning disk systems really had measurable pixels, so I don’t reach back farther than the early CRT systems. )
Plain and simple. On the other hand, if you can prove what you said, and prove my cites are completely wrong, then of course I will apologize. Otherwise, you’re wrong. This is G.Q. You either cite proof, or admit that perhaps what you believed to be true, is not. 
Cartooniverse
Well. You can either have a failure in what we might call the “driver circuitry”, which is the actual wiring and transistors that deliver the flow of information from the CMOS, or you can have yer basic programming glitch. ( That’s why I said we were both right. We were both referring to the same words, applied in slightly different ways. )
If you have a bit of damaged transistor or wiring, then yeah. That pixel, or pixel set is gonna be blown. OTOH, if you have a CMOS whose drivers in the software/controller sense of things is failing, you can re-install them.
A reasonable analogy is someone who uses a Windows machine. They notice that they cannot do all of the things in Outlook Express that they could do until recently. It is commonly held in many circles that it’s a fine idea to re-install a fresh OS into your computer every six months or so. Why? Because hard drives are physical things, reading arms sweep back and forth, iron oxide flakes now and then, lots of things happen and you get unreadable cylinders/sectors. Re-installing a fresh OS allows the computer to skip over the damaged cylinders/sectors and install the OS only in areas that are able to operate properly. In doing so, you defeat the “blown pixel” issue whether it was caused by physical or electronic causes.
To take it to a logical conclusion, as I wrote that last graph I realized that there is no such thing as an electronic cause. If the CMOS is losing bits of info, it is the physical CMOS that is doing so for whatever reason. You do a fresh install, and you are back at square one. We need to assume that the driver software is bug-free and operates as it should. Given that, when we turn on our computer/camera/LCD, the information flow from the software that’s stored on a hard drive of any kind is read and used in a consistent manner.
So, assuming a bug-free OS, we can say that a transistor is physically failing, OR that the driver software is failing but if we say that it’s the driver software, then what we are really saying is that the working install of the OS is not being read properly for any number of reasons, such as those given above.
You can have a piece of shit OS and a spectacular machine. You still get crap. OTOH of course you can have Mac OS-X Tiger and a 1983 Apple, and still have crap. All depends on how you look at it.
Sure, although I think that in terms of LCD monitors, dead pixels are typically caused by failure of a transistor in the control circuitry; it’s part of the reason that kept large LCD screens so expensive for so long and also part of the reason that manufacturers have insisted that a certain number of dead pixels should be acceptable; the number of transistors increases with the square of the screen size; more components=more potential points of failure.
No, I don’t think that’s the case at all. First of all, the word pixel didn’t appear until about 1969 (can anyone with an OED find an earlier occurrence?) and it was, as you say, originally used in connection with COLOR television images. Prior to that, black and white images were composed of lines–a term still used today to describe TV resolution. There weren’t any individual dots, like color CRTs would later have and thus no need for a term for them. Since CRT computer monitors and television sets are fundamentally identical in operating principles, it was only natural to apply the term pixel to those, too. The main difference here was that the dots, or pixels, which composed the images on the early monochrome monitors were more a function of the computer’s limitations, rather than a necessary design element, as was the case for color CRTs.
Nope. The OED cites the following sentence from Science magazine in 1969 as its earliest known usage:
The article in question is titled “Mariner 6 Television Pictures: First Report”, but unfortunately my institution doesn’t seem to have access to their online archives.
I didn’t say it was used in color television images. I said it was used in television images. And, more to the point, television cameras- that were not color until (nominally) 1960.
How can I have a blown pixel in a black and white studio viewfinder if it only applies to color monitors?
I will contact SMPTE tomorrow to see what they have to say about it. They are the governing body of Broadcast Television engineering. ( and of movies, since it stands for Society of Motion Picture & Television Engineers ).
I knew I was going out on a limb, mainly to provoke someone to look up some cites either way. 
My dictionary just put the etymology as “From TVs and computers.” The cites so far back that up. I was of course wrong to say it originated in the computer world.
However you’ve so far completely failed to come up with a cite for the usage of the word “pixel” at any time, and with this unproven sentence “Pixels are and have been the representational unit of measure of display information since the 1920’s on cathode-ray tubes.” shot your own statement about LCDs in the gut. Unless you mean to say LCDs don’t have display information. 
Pixels can only really exist within a digital system; by ‘digital’, I don’t mean the popular modern usage as in ‘digital TV’, but just digital as opposed to analogue - i.e. having a finite set of possible discrete states as opposed to having continuously variable ones.
Early TV was not digital at all, it was analogue; the image was captured in the camera not by a pixellated array of photosensors, but by analogue scanning processes; it was transmitted as a waveform and rendered on the TV set by analogue scanning processes - by modulation of an electron beam that was scanned repeatedly across the phosphor. There were no pixels.
Lazy. And, glad to have been of service !!
Well. Look. I didn’t get far with the SMPTE but I ain’t giving up yet. I am digging my heels in until I’m told otherwise by world-recognized experts and still insisting that pixels are an accurate phrase to be used to describe the dots appearing on a t.v. screen. ( NOT the spinning disk stuff of the early days, of course ). I refer to the cameras and sets used by oh, say, 1930-1940.
Even after the Baird-style spinning discs, early cathode ray television was still an analogue technology, implementing continuous scanning techniques; there were not ‘dots’ - only lines of modulated brightness; there would have been a minimum width for a region of brightness, but that’s still not a pixel, unless its position is also constrained within a grid, which was not the case.
For further reference, look up the term ‘image dissector’ - this is the name of the (analogue) image scanning tubes that were in use in TV cameras, and remained in use right up until the 1980s (although not exclusively by that time) - they do not use a photosensitive array or grid, just a surface that is scanned by a focused electron beam.
Pixels, I believe, as a concept, preceded their use in cameras; they were first implemented as a method of recording television footage.
( Cite )