Talk:Indexed color

merge from palette, CLUT, and list of palettes
All the information currently in palette (computing) and Colour Look Up Table belongs in this article. It should be merged, and they should be redirected here. A great deal of the information about indexed/palette systems is listed in the List of palettes, so while that article shouldn't necessarily go away, its information should be judiciously merged into a section of this article, discussing various historical indexed color schemes. --jacobolus (t) 19:53, 19 September 2007 (UTC)


 * Fancy meeting you here. :) There are some distinctions between meanings. For example, you could say a 24-bit color computer has a palette with 16 million colors, or if you were a salesperson, "an unlimited palette," but that's no longer using indexed colors or cluts. (Well, software on the computer could use cluts in the algorithmic sense). I think indexed color would better fit as a subset of palette (computing) than the other way around. Though I agree that all the info on palettes in the palette (computing), with the exception of the definition, deals with indexed color palettes only.


 * A clut is certainly related to indexed color. It's not a synonym, but the color lookup table article seems unclear, and what is there generally would fit here. In the computer science (as opposed to electronic) sense, I'd probably describe it as a data structure or area of memory used to map one color palette to another. In practice I think that overwhelmingly involves indexed color on one side or the other, for example mapping RGB to indexed, indexed to RGB, or indexed to indexed, though you could use a multidimensional array to map RGB to RGB, and that would still be a clut. In the circuitry sense, I'd describe it similarly, it's the memory referred to by the dac to generate the video signals. The article says it's a "device," which I don't think is clear or precise, but I may be wrong, or am using clut in a general sense that's not intended by the article (i.e. a look-up table that happens to involve colors, rather than the CLUT in on a video adapter, e.g. in a RAMDAC).


 * That's not to say don't merge any articles; just pointing out the different meanings, as I see it, since it's not clear in the existing articles.


 * With List of palettes, did you mean it should stay intact, and just borrow some of it for here, or mean move some of the info? It's quite a nice article as it stands! -Agyle 21:04, 23 September 2007 (UTC)


 * I think that the information about larger "palettes" more properly belongs at color depth than in its own article, and that indexed color is the proper place for the information about palettes of <256 colors. But I suppose we could call the article palette (computing)—that wouldn't really bother me too much.  I've just added the main template to the indexed color section of the color depth article, to point it here.  I think color lookup table should be one sub-section of this article—too much of its background, etc. properly belongs here to make leaving it independent worthwhile.  As for List of palettes, I just don't think that its current organization or focus is really ideal.  I tend to think that writing articles, instead of just jumbling things into choppy lists, is more useful for readers.  The information currently there should in my opinion be split up into some sub-articles about different types of palettes, and the most interesting examples should be summarized at this article (whatever its name ends up being).  --jacobolus (t) 21:38, 23 September 2007 (UTC)


 * Some amount of summary probably belongs at bitmap or raster graphics as well. --jacobolus (t) 21:41, 23 September 2007 (UTC)


 * OK, agreed on CLUT. I think if it's a topic that could warrant its own article, but doesn't need it now. Again, noting that it's talking about the hardware sense of clut (e.g. VGA RAMDAC), and this article is currently talking about software sense of indexed color (e.g. GIF header), and that either term, as I understand them, can refer to either hardware or software based palettes.


 * You mean <256 color for indexed color, or <=256? I think of VGA as the most common indexed color hardware (18-bit total palette, 8-bits 256-color subset at a time in low-res mode). I think even larger palettes were used for indexed colors (e.g. early 24-bit color cards), but I'm not sure they're covered in the List.


 * Almost all of the stuff in List's "hardware palettes" seems irrelevant here; 4-bit or 16-bit might or might not be indexed, since it's talking about maximum different colors, not the number of indexed colors that can be used at any given time. Of the manufacturer-specific info, I don't know what you have in mind, but I don't see a need to cover specific palettes here other than maybe to illustrate the concept of indexed color, or to list manufacturers to show when indexed color hardware was prevalent on personal computers. (I reckon it's still used for PDAs and cell phones.)


 * List_of_palettes and List_of_palettes are two sections that I don't think fit so well in the List article, because they're about general types of palettes rather than specific palettes or specific graphics systems.


 * I don't see anything wrong with duplicating a little info between articles, and I don't think most related articles would want as much info as List has on a trivial component like specific palettes (e.g., see the Color Graphics Adapter article compared to List_of_palettes). I think the List's shortcomings would be better addressed by improving it rather than breaking it up, or perhaps by ignoring it if you just don't like list articles. :-) I'm not planning on working on these though, just happened here via the list of palettes Merge notice and thought I'd comment. (Sorry for the verbosity!) -Agyle 00:06, 24 September 2007 (UTC)

The article needs be rewritten completely
... in my humble opinion. --SunnySideOfStreet (talk) 20:19, 27 November 2007 (UTC)


 * Go for it. :) --jacobolus (t) 21:32, 27 November 2007 (UTC)

Rewrite of "List of Palettes" article is running now!
Hey! I'm the guy who mainly built the "List of palettes" article in its current form, based on a somewhat mangled previous article. Yes, I recognize that, beyond a mere (and colorful!) list, the article should be rewriten and reordered. Right now, I'm working in a splitted series of the article, with a base article entry for the current "List of palettes" along with the following: "List of monochrome and RGB palettes", "List of software palettes", "List of 8-bit computer hardware palettes", "List of 16-bit computer hardware palettes" and "List of videogame consoles palettes". Along with proper rewriting, I did convenient survey, I provide a lot of technical info and references, and even I did a lot of samples of high-quality simulated images. These days, I'm building the samples in my Sandbox pages, so someday all of you will see the results. As you know, to do good articles requires time.

But I think these list articles should not be merged with articles depicting what a computer palette is, or a CLUT is (you know, they are not the same). I vote for revisiting these articles if you think they are somewhat ambiguous. Yours. Ricardo Cancho Niemietz (talk) 14:52, 1 February 2008 (UTC)


 * I did it! no more further discussion about "Indexed color" vs. "List of palettes" expected.Ricardo Cancho Niemietz (talk) 04:59, 2 February 2008 (UTC)


 * Okay, re: CLUT, do you really think that there's enough non-overlapping material between CLUT and indexed color, for the topics to warrant two separate articles? It seems to me that a CLUT is an essential aspect of any indexed color scheme, which means that it should be a sub-section, until its content becomes too long to fit in a section.  --jacobolus (t) 01:00, 3 February 2008 (UTC)


 * Well, I'm not actually the CLUT article's author, but my own opinion is that, although related, CLUT and Indexed colors should not be mixed together: CLUT is always a hardware resource of a given display adapter, mainly related to calibration (gamma, white temperature, etc. —although in Commodore Amiga, the color palette's hardware is also known as CLUT), while Indexed color is a generic way to handle the colors of the images in a limited fashion, and it can be implemented on hardware (as in pre-High Color graphic cards and devices) and/or be used exclusively with image files, depending on context. So, as I said before, I still prefer two separate articles, cross-linked for convenience. If all of us follow your criteria, many of the articles of the Wikipedia should be merged with the yours, from Color spaces to gamma correction, from bitmap to bit-plane, from Digital-to-Analog Converter (DAC) to palette (computing), and so on, due all of them are more or less closely related with the Indexed color concept. It's up to you to include citing sections with summaries of these related articles, and links of the style "Main article:...", but *please*, Do not predate them!Ricardo Cancho Niemietz (talk) 12:45, 4 February 2008 (UTC)

I think this whole area is up for grabs, since nothing is even sourced yet, so there's no telling how these articles relate to the scope of topics as verifiable in reliable sources. I'd be inclined to merge them all (not the lists, but the ones that are articles), but I'm not sure what the best title is. As for the lists, it's unclear to me what the categories mean; what is a "software palette" as opposed to an "RGB palette" for example? Dicklyon (talk) 06:02, 16 February 2008 (UTC)

Palette (computing) article enterely rewritten
Hi. Along with the List of palettes series of articles, I completely rewrote the Palette (computing) article, which was a mere stub. Now, it is extended, and clarifies many usages of the word "palette" in computing, so it can act as a "base article" for those related (CLUT, Indexed color, List of palettes, palette entry, color depth and so on). At proper time, these articles should be expanded as well, in order to sharp definitions around every topic. And I know, palette's article still lacks references, be patient (or include them by yourself!). -Ricardo Cancho Niemietz (talk) 19:14, 19 February 2008 (UTC)

Indexed color article completely rewritten
Hi. I rewrote the entire article almost from scratch. Now, the stub tag has been removed and the merging suggestions tags along, due to the extended and clarified article left little room for merging. Greetings. -Ricardo Cancho Niemietz (talk) 21:09, 26 February 2008 (UTC)


 * That's a good piece of work. I found a few sources, and took out the alternative terms that I couldn't find sources for, and did a few edits.  Still would be nice to see most of the contents sourced. Dicklyon (talk) 01:18, 27 February 2008 (UTC)


 * Great! Still can use some inevitable wiki touch-ups, etc., but generally quite a bang-up job.  --jacobolus (t) 01:55, 27 February 2008 (UTC)


 * Indeed, it needs more than just "touch-ups". There is a lot of good content here, but the phrasing makes it clear that the author is not a native English speaker -- word orders and idioms look very Spanish. No offense meant. I tried cleaning up two paragraphs but there's much more to do. --Rpresser 13:59, 15 July 2008 (UTC)

Lets talk
Hi, Dick. You did it again... but this time I don't want to scale. About the "named colors", well, it was never my favourite term, but it seems that it has been used. I left there simple due original author put it there. The reference I put says (I quote):


 * «Palette-based, multi-tint, named-color methods and systems utilize a pixel-by-pixel indexing technique in which individual index values into a palette of interest can be used in different ways for rendering associated images across different devices.»

The source is reliable, in the sense a US patent should be. It was not the only example; try with Google. This time, I didn't revert your deletion (again, a "citation needed" would be better), only rescued some of my words and by avoiding comparisons. I hope you'll agree this time.


 * Yes the source was reliable; but I read it, and it certainly did not support your idea that "named color" was another term for "index color"; quite the opposite; try reading it. Dicklyon (talk) 18:20, 29 February 2008 (UTC)


 * It appears to me that it denotes that "palette-based", "multi-tint" and "named-color" are equivalent, not diferent "methods and systems". I never eard about "multi-tint" before in this sense (may be it is specific of some unidentified context), but yes about "named-color". Indeed, this was the unique reason for me to leave it; personally, I dislike it. The rest of the line is the definition of "Indexed color" in itself, and that was the reason to pick that reference. It seems to me adequate for the case.

In the other side, you reallocate the "Pixel color depth" in front of the heading. I don't know which browser you use, but in IE6 the result was catastrophic: the template autoreallocates just below the first image at left, adding more blank space between the "Contents" table and the header of the first section. My revertion is intended only to repair the visual damage. Maybe other solutions would be better, but if different browsers behave different in layout, we should reach a solution of compromise.

A curiosity: Does "Dick" stands for "Richard"? "Ricardo" does... -Ricardo Cancho Niemietz (talk) 16:19, 29 February 2008 (UTC)


 * Yes, indeed, we have the same first name. I'm sometimes called Ricardo, too.  That makes us best buddies.  So stop complaining about my edits.  I don't need to go the talk page to discuss every simple edit that I can describe in an edit summary. Sorry about the IE6 problem; I don't use that; when you fix it, be sure to mention IE6 in the edit summary to let others know the reason. Dicklyon (talk) 18:19, 29 February 2008 (UTC)


 * If in the future some other layout problems will arise, I'll put the IE6 warning. In fact, the behaviour of the IE6 is the reason to put HTML separators and other layout tricks along the articles I contribute. I wasn't aware before of another browser behaviours. Both of us tended to think that everybody see the Wiki pages exactly as we see, and it is false for the both, too. I propose to you exchange snaptshots via e-mail when layout problems arise, to find best solutions. I think I'll send now the two versions, before and after of your edit. And the weekends I ever stop to wikiedit, so see you next monday. -Ricardo Cancho Niemietz (talk) 20:38, 29 February 2008 (UTC)
 * See what I mean? Here, let me translate your paragraph into normal English so you can see the difference:
 * If in the future other layout problems should arise, I'll put in an IE6 warning. In fact, IE6's behaviour is the reason I put HTML separators and other layout tricks in all the articles I contribute. I wasn't aware of other browser behaviours before now. Both of us tend to think that everybody sees the Wiki pages exactly as we do, and it is false for both of us. I propose we exchange snapshots via e-mail when layout problems arise, to find the best solutions. I think I'll send you the two versions now, before and after your edit. And on the weekends, I never stop to wikiedit, so see you next Monday.
 * This may seem trivial but your style is very, very stilted sounding to native English speakers. --Rpresser 14:14, 15 July 2008 (UTC)


 * So go fix it in the article. There’s no need to critique talk page writing style. —jacobolus (t) 18:13, 15 July 2008 (UTC)
 * I wasn't; I don't particularly care about how he writes on the talk page. If you look at the section above this one you'll see my primary complaint is about what he did to the article. Yes, he added a wealth of information -- but it's now excruciating to read the article, making his changes and now the entire article worthless. --Rpresser 19:31, 15 July 2008 (UTC)


 * Your edit summary for your comment above was “*shudder*”, which is quite insulting. Feel free to fix the writing.  This is a wiki. —jacobolus (t) 20:15, 15 July 2008 (UTC)

Palette support by GIFs
The table of file format support indicates that GIF files only support 8-bits-per-pixel. This does not seem correct to me, as Photoshop (and other tools) allow one to change the number of colors in the palette. This results in drastically different file sizes for large images, implying that not only has the color palette table been reduced, but that each pixel in the image (prior to compression) is using a smaller number of bits.

Further, the GIF87a specs (http://www.martinreddy.net/gfx/2d/GIF87a.txt) use the phrase "# bits per pixel" repeatedly, describing a range of values from 0 to 7 representing 0 to 8 bits. (For example, see page 5.)

I humbly submit that this table is either wrong or very misleading, and suggest that the GIF row be updated to show a value of "Yes" for 1-8 bits per pixel.

216.228.112.21 (talk) 18:13, 4 August 2008 (UTC)

Overlap in Framebuffer
I have removed some detail from Framebuffer that I assume is already covered here. I may have been a bit cavalier and it is possible some of this deleted material would be useful here. ~Kvng (talk) 20:34, 16 July 2018 (UTC)