Quote:
Originally Posted by Merlin66
here's the thread mentioned...
http://www.iceinspace.com.au/forum/s...ad.php?t=72640
Astroguy,
I'm interested in your experiments. I'm assuming that a conventional LCD screen is made up from individual pixels which emit either green, red or blue...these are then "combined" (?) through a bayer matrix solution to display any one of the 64 million colours...Is this correct?
If so, a couple of comments:
What's the size of the individual pixel? What's the average distance between centres?
In a DSLR the De-bayering process actually "manufactures" a virtual pixel at the intersections based on the mathematical model used ( Craig Stark has a very good write-up on his site) and it's this "virtual" coloured pixel you see in the viewfinder and RAW image of a DSLR.
I don't know how the individual pixels (if they are only RGB) in a LED can be made to emit a different colour...so it may be a combination of varying intensity for each pixel giving the "impression" of colour ie combined by our eye....
It is certainly possible to easily test the outcome. We have access to spectroscopes with bandwidth resolution of 0.01nm......
How can we help each other??? |
Did you get a chance to check out my LCD Digital Filter thread in the ATM/DIY section?
1) Basically on a good quality LCD screen there is very little or no inter-pixel space.
2) Pixels on LCD screens used for my project do not emit light, they channel it through RGB Liquid crystal. They allow light to be filtered so that individual RGB components can be filtered and focused in the eyepiece/camera/photo diode whatever light detector exists at the focal point.
to create a spectrascopic analyser for e.g.
The LCD screen has the backlight removed so it becomes transparent, by turning on all the RGB pixels through software you can generate an entire screen of a particular colour/hue. The screen is basically mounted in front of the scope aperature and becomes a colour filter. The light that can only travel through the programmed LCD colour filter reaches the photo diode at the eyepiece end when focused. The voltage is recorded and graphed for that colour/hue. By cycling throught all the possible colours/hues of the visible spectrum and recording their respective voltages and plotting them on a graph......well you can see where this is going right? by checking particular wavelength plot amplitudes, you can see emmision and absorbtion characteristics. Keep in mind though, a 24bit LCD screen can only create 16777216 different hues/colours. So the visual spectrum can only be diced up into as many slices. Take into account though that a LCD screen mainly works at 60 frames per second. That means you can only sample 60 individual spectral lines per second!
Using this system each spectral line capable of being recorded has a bandwidth of 0.00002086162567138671875 nm (nanometres). If you wanted to do the entire spectrum you could skip every 70,000 spectral lines, this would sample the visible spectrum at 240 points from say 350nm to 700nm and it would only take 4 seconds for a preview. Because remember, you can only scan any 60 spectral points per second. Preview would not be as accurate as re-sampling the same spectral points (colour/hue) many times and taking the average, basically many short exposures that are stacked can be done through the software.
I would rather use a photo diode to gather gather the sampled spectral wavelength voltages detected at any given spectral point in the visible spectrum. I think using colour CCD's would create to much work and complicate things, not to mention costs involved. You will however have to get the data of the spectral response characteristics of the diode in question and compensate for this in software.
You would need to have the star/subject area being analysed taking up the entire eyepiece view or at least focused on the photo diode's detector/sensor for this system to work accurately.
If you were able to see the eyepiece with the star taking up the entire eyepiece exit pupil area, you would see a kind of cycling of colour from deep violet to deep red and every other hue in between. The intensities of these hues/colours is the data you would be capturing on the photo diode sensor and recording or displaying/recording as spectrographs on the computer attached to the system.
It may sound a little complicated, and a little convoluted in the way I have explained it, but if you read it a couple of times you will see the light!
Anywhooooooo, I've probably bored you enough already
I'll send you a link to those cheap components ok. You won't be sorry you checked them out.
