Talk:Color filter array

3 new nameless CFAs from Kodak
At 2007-05-14 Kodak presented 3 4x4 CFAs: http://johncompton.1000nerds.kodak.com/default.asp?item=624876 —The preceding unsigned comment was added by Abu ari (talk • contribs).


 * Yes, CFAK is logical; maybe even pleasant. But it's not sourced; it is your original creation, and therefore not really within wikipedia guidelines.  It can have a life if published elsewhere first. Dicklyon 15:54, 7 August 2007 (UTC)

antialiasing
why arent those new lower-aliasing-prone filters that have alternating red/blue in a row not mentioned here? — Preceding unsigned comment added by 2001:6F8:1293:1:84EA:31D2:50BD:A7B9 (talk) 21:57, 6 January 2014 (UTC)

Merge from bayer filter
I suggest merging from bayer filter into this article. It has information about a number of other colour filter arrays that really belongs here. And I don't think it is sufficiently different from the other types of array to warrant its own article. So just merge it all over here and keep a redirect. What do other people think? --Imroy (talk) 18:43, 29 November 2007 (UTC)


 * I am absolutely against merging content about the Bayer filter simply because the Bayer filter proliferation is huge and obviously notable. Dealing with the RGBW filters is another question (IMO).  The RGBW stuff was put here because it could be seen as a variation of the Bayer filter with some unfiltered pixels.  If it stays or goes I'm much less decisive about. :)  Cburnett (talk) 22:41, 29 November 2007 (UTC)


 * I think the current arrangement of two articles is pretty satisfactory. Dicklyon (talk) 02:20, 30 November 2007 (UTC)

What are these filters made of?
The article could be improved by information on how color filter arrays are manufactured i.e. what material are they made of (resin? glass?), how is this material deposited on the sensor (vapor deposit? printed?). Unfortunately I don't know the answers and can't find them anywhere... Any expert willing to help? 146.203.21.99 (talk) 15:44, 3 October 2008 (UTC)

New sections
Some anon IPs (probably all one editor) have been adding some new stuff without proper citations, and a long list of "further reading". What we really need is to have the content made verifiable by citing sources. I will be happy to help convert the list of sources to proper footnotes if the editor can point out what parts of the material are verifiable in what source. It would be a giant mistake to let the article grow this way, when the editor has the sources handy and do the right thing. That's why I've reverted it until we can go about it correctly. Some of the "further reading" sources are sort of referred to in a sort of Harvard style, but that's not the style used in this article. The stuff about Foveon seems to unsourced and irrelevant here. Also, it is best to respect normal style, such as heading case, when adding new material, rather than leave a lot of work for others to clean up. Dicklyon (talk) 01:17, 9 November 2008 (UTC)

New sections
Technically Sandbox or Talk pages are the proper forum for expansion of this type. I have gone through significant labors to acquire the academic materials required to "get a handle" on this particular puzzle and I have made significant revisions. All (perhaps I mean "most"...) books you read on CCDs will "glance over" or ignore the issue of CFAs, and the semiconductor manufacturing books I have consulted rarely help. Most of the literature (that I have seen) on the subject of CFAs are mum on the subject and only seem to duiscuss the efficacy of their particular "formulation" (i.e. RGB, CMY, RGBW, OU812... blah blah blah) and/or interpolation algorithms.

If there are problems with my edits it cannot be (a) simultaneously too few citations and too many citations (b) that it has not been made verifiable when verification appears to be physically impossible, i.e. where trade secrets and/or unspoken industry standards disallow honest, open academic discussion. You will note that the citations made cover most/all of the major books on CCDs (i.e. silicon imagers), many IEEE articles on CFAs, and several references to SPIE sources. If the international electrical and electronics Engineering society and the Societe Photometrique Internationale Engineers don't print CFA construction particulars, it is reasonable to assume that it has never been printed. Prove me wrong. Show me the article that illustrates the proper way to manufacture CFAs. Now that would be constructive!

Harvard style? Tell that to Harvard! I think they should hear it. Anybody want to write me a letter of recommendation? I'd be a shoe-in for grad-school if only I had wikipedia's approval! If you want to add links to the "further reading" citations I have made, be my guest. That would make the aticle more user friendly! The IEEE articles are not obtainable on-line (without paying a fee). A good college library will have these articles and can be verified "the old fashioned way" (hint: it doesn't involve an internet connection. Some walk, some drive, but all access these things called "books").

Admittedly, I was quick with my analysis of several/many of the IEEE articles pointed out by Theuwissen's book. I lumped them all into the category of "unhelpful". Some articles indicated that "Gelatin" was the construction material, others indicated that the CFA is manufactured separately and attached to the image sensor die "late" in the manufacturing process. Some indicated both. Some indicated neither. I'm sort of fed up with the process, but I have acquired some other research material and I continue to investigate the interesting science of optical filters. My edits provide an interesting springboard for interested students. It does need links to other wikipedia articles like "dichroic filters" "narrow bandwidth filters", "thin film optical filters", "vacuum deposition", "anti-reflection films", etc. Its all one subject. We are, essentialy, talking about the particulars of microfabrication. If you want knowledge, I can't dumb it down or make it any more palatable than this. Things are uncertain, until I find the obscure reference that makes them certain. Until then I'll assume (a) plastic (with impregnated dyes or pigments) or (b) glass (with chemical dopants). (c) Thin-film interference filters and (d) silicon absorption-depth filters remain a possibility, though these two latter items are not necessarily relevant to CFAs (But they are undeniably relevant to the subject of Single-Sensor Color Imagers)!

My new research materials are the 2005 "The Photonics Handbook" put out by Laurin Publishing. There MIGHT be some interesting filter science stuff in here. I live in "intellectual siberia" and I have very little access to quality research (reading) and/or research (experimental) materials. So, from my point of view, what I have to add to this page is better than the nothing you have offered.

I didn't bring up "Foveon". Some other post-er did that. It is an intriguing sort of sensor, and does require some attention. Further, since James Janesick devotes much of his book to the efficient collection of light by a Silicon die, it might be useful to mention that Silicon is something of a natural "filter". I spared you an in depth discussion of Carver Mead, CalTech grad students, Franscisco Faggin, Intel microprocessors, etc.

How about this. If you want to discuss the interplay between pixel size, interpolation algorithms, MTFs, sensor "resolution", contrast, sensitivity, Signal to Noise Ratio, etc., move it to the page "Solid State Image Sensor". Make a big subcategory of solid state image sensor (i.e. "redicrection" links to "CCD Imager" page or "CMOS Imager" pages). Make big links on those pages to "Monochrome Image Sensors" and "Color Image Sensor". Further links to "color interpolation algorithm", or "signal integrity", or "CFAs", or "3-CCD cameras" or "Single-Chip Color Cameras".

In short, use the CFA page to mention only its construction, chemistry, and a word or two about the type and combinations of color. Use the "Color Interpolation Algorithm" page to gripe about the lack of knowledge regarding CFA construction or the sloppiness of the post-er on the "CFA" page. Yes, there is lots to clean up, but the start I provided is better than nothing! —Preceding unsigned comment added by 4.156.246.42 (talk) 22:14, 9 November 2008 (UTC)


 * As I mentioned, the problem is primarily the form or style of your contribution. See my lastest edit, where I converted an external link to a citation, to give you an idea how it's done in the usual citation style.  If you'll start to make an effort to integrate your material correctly, I'll chip in and help.  Please at least start with some refs, and fix the heading case, as a show of good faith. Dicklyon (talk) 22:52, 9 November 2008 (UTC)

OK, thanks for trying to fix it up. I did some work on it so you can see some of what I'm talking about. Please read the edit summary and study the edits, and do more of the ref merging via naming. Dicklyon (talk) 02:01, 10 November 2008 (UTC)

It's really too bad that you decided to do such a massive edit before learning how. With all my edits to try to clean it up, it's still got a long way to go. Please see Template:cite journal and use it where appropriate; and take quote mark off of book titles. And the sentence about "a curious note" is going to be removed shortly if you don't clarify where this info is from, to make it verifiable. What's that long list of citations about on it? Dicklyon (talk) 04:20, 11 November 2008 (UTC)

Sorry to be such a "blusterbutt"! I got the "ranting" out of my system. The article should be up to "usable" standards now. I am a guy who has academic information and, sometimes, I am just looking for a place to put it! After I put two more direct quotes on this page (one from Nakamura and one from Theuwissen)I will be finished (or, at least, I will be finished until I find some REAL answers. The answers I have given so far are the best approximation of truth that a marine biologist (Bachelor of Science), with 4 years of independent study on image sensors and imaging systems, can give.). —Preceding unsigned comment added by 216.204.233.178 (talk) 17:36, 14 November 2008 (UTC)


 * I don't understand why the wholesale change in reference style. Is there a problem with the present style that I reverted back to?  See WP:FOOT. Dicklyon (talk) 01:19, 15 November 2008 (UTC)

Nope, No Prob. Sorry, I am new at this. I think I was editing at the same time you were. The way you did it is fine. In fact, it is better. Thanks so much for your help. I need to find some more academic materal and (1) get smarter so corporations will pay me to make better products and (2) get smarter so the CFA article is smarter. Hopefully there is another expert out there--someone who is twice the expert I am on this particular subject---who will come along and say to themselves "wrong, wrong, wrong" and proceed to fix the CFA article. Until then, this will have to do! Its not false, its just not specific enough for my standards. I have to know every small detail!!!! I checked that reference I told you about (Photonics Handbook) and i pointed to two other referencesthat might help. One, by Edward d. Palik, "Optical Properties of Solids Handbook" and another "Technical Digest Series: Optical interference Coatings" (cannot find it in any library, but I have some close matches???!!?). If you are in NY, Boston, D.C., or Philly, you could go and research it. It would happen quicker. I have to wait a week or two to get them via interlibrary loan! Ideally, upsatate NY wikipedia-posters could go check some old Kodak and/or Corning corporate pubblications. This would giv the proper historical perspective for chemistry of solid/photographic color filters. Rochester, NY, I am sure, has excellent library materials for this subject and these items (i.e. Kodak/Corning Filters). —Preceding unsigned comment added by 4.157.110.172 (talk) 22:33, 16 November 2008 (UTC)


 * I think a lot of the recently added material is unsourced interpretations (WP:OR), and is essentially off-topic anyway, as the history of assorted color filtering approaches doesn't have a lot to do with CFAs. What do others think?  Dicklyon (talk) 16:28, 20 November 2008 (UTC)


 * Hearing no objection, I've removed much of the unsourced, speculative, and interpretive bits that were added, along with some off-topic paragraphs. And I did some other minor fixes. Dicklyon (talk) 07:39, 22 November 2008 (UTC)

No huge objections here. Just wondering why you think old filter technologies havelittle to do with new filter technologies? Evrything comes from somewhere? Glass filters (inorganic)? PLastic filters (Organic)? Dielectric/interference filters? All seem plausible-to-probable. Perhaps, though, cfas arose out of photographic film technology (i.e. gelatin, silve rhalide crystals, etc.?)? The dichroic cfas from ocean optics seems, to me, to be the best, most supported answer so far. Alternating layers of ZnS and cryolite thinnfims, or, at the very least, alternating thin film layers of a high-refractve index and low refractive index material! This seems to be the one undeniable case of REAL chemical constituencies of REAL CFAs. I do not believe the ocean optics patent speaks of the chemicals used, but it does describe diochroic/interference filters and we do know how those are made. As for transmissive filters, I am still in the dark. Perhaps www.gerstaltec.ch has the answer. They make photoresists and it seems some are used as color filters. —Preceding unsigned comment added by 4.154.204.64 (talk) 18:47, 29 November 2008 (UTC)


 * It's 2017 and the end of this article is still a mess... it reads like a literature review and gets close to OR with speculation on what authors are revealing or not. Stub Mandrel (talk) 23:08, 29 October 2017 (UTC)

Question
What is meant by the sentence - "Since sensors are made of semiconductors they obey 'solid-state physics'." ? All solids obey 'solid-state physics'! Am I missing something? — Preceding unsigned comment added by Iantierney (talk • contribs) 11:08, 16 January 2013 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on Color filter array. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added tag to http://www.outbackphoto.com/dp_essentials/dp_essentials_03/essay
 * Added archive https://web.archive.org/web/20110721094440/http://www.mff.ust.hk/mffdoc/safety/msds.htm to http://www.mff.ust.hk/mffdoc/safety/msds.htm

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 21:47, 10 August 2017 (UTC)

CFA on ink
The section on eink is gibberish. There is no LCD layer, and if there was that would not convert it to mono. It's LIKE an LCD, being a mono panel either with printed translucent stripes (Triton) or a 2 x 2 pattern of more translucent pastel dots (Kaleido). No eink panel in production has an LCD layer. see https://www.eink.com/brand/detail/Kaleido

and less informative https://blog.eink.com/e-ink-kaleido-plus-the-next-generation-of-print-color-displays 84.203.21.239 (talk) 12:14, 21 April 2024 (UTC)