Talk:List of 8-bit computer hardware graphics

Mattel Aquarius

At http://atariage.com/forums/topic/227441-aquarius-information-about-correct-ypbpr-colours-from-the-palette/ we can have accurate information about the colour palette used on Mattel Aquarius, but i have no idea how to convert it to YPbPr colourspace format - if anyone can help, please do!

IBM PC

Why is there something about the 16 bit IBM PC in here?

The reason is that the original IBM PC launched in 1981 had an Intel 8088 CPU that was 8-bit data bus technology. Now, this is noted out in the article.

To check for 16-bits IBM AT (and better) palettes, please visit List of 16-bit computer hardware palettes. And please, sign your postings.Ricardo Cancho Niemietz (talk) 09:47, 12 February 2008 (UTC)

The IBM PC had an 8 bit data bus but a 16 bit processor (and the XT a 16 bit data bus as well) so it would still seem to be better placed under the 16 bit article with the rest of the PC palettes. 58.178.196.103 (talk) 16:55, 12 February 2008 (UTC)

The main criteria to distinguish an 8-bit from a 16-bit computer architecture is its data bus, not its internal registers' size, due to the CPU doesn't live alone inside a given computer. The IBM XT was a true 16-bit architecture, but it still features the original CGA; remember that the article deals with color graphic capabilities, not general capabilities, so the IBM XT fits well here yet. The EGA and better display adapters came with the IBM AT family, which features an 80286 pure 16-bit processor and bus, so their palettes are placed under List of 16-bit computer hardware palettes. Remember: not "PC palettes" but "CGA, EGA, VGA, etc., palettes". Yours. ---Ricardo Cancho Niemietz (talk) 18:38, 12 February 2008 (UTC)

I don't agree that it's the data-bus' size because there were PC clones that used the 8086 and that had a 16-bit bus, yet using your criteria they are "8-bit" computers. In fact, using your criteria, I can name a few personal computers that had 8-bit CPUs (Z80, etc.) but 16-bit busses, and those would be misclassified as well. As previously noted, word size is the defining characteristic of "size" of a computer, and for that the PC was most definitely a 16-bit machine. CGA should really be moved to the 16-bit article. --Trixter (talk) 17:27, 13 February 2008 (UTC)

The word size has a lot more of an effect on the general purpose usage of the computer which is what most people will use to determine if a computer is 8 or 16 bit without much regard to the data bus (not many will call a Pentium II a 64 bit computer) so those looking for information on IBM PC graphics are more likely to look under 16 bit for everything related to the PC which includes CGA, MDA, Hercules, the PCjr's graphics card and not just EGA, MCGA, VGA, etc.211.26.235.163 (talk) 12:40, 13 February 2008 (UTC)

In order to avoid confussion, I had cross-linked the "IBM sections" in the two articles. Again, this series of articles aren't focused on the main systems' capabilities but in the colors in itself and the methods to produce them in every enumerated system. There are specific articles devoted to the various personal computers and display adapters cited, all of them linked from here. Also, no info is provided here about PC display hardware as Hercules Graphics Card or IBM PCjr due to they do not include different color modes (not graphic modes!) aside of those found in the MDA and CGA (the technical reason is that these adapters shared the same expensive IBM monitors). Finally, due to the technical and historical context, the CGA for the original PC/XT still belongs to the 8-bit article. So don't think about "PC palettes" anymore. And please, identify yourself next time. -Ricardo Cancho Niemietz (talk) 16:57, 13 February 2008 (UTC)

CGA needs to be moved to the 16-bit article. VGA is there, and IBM made a VGA card for the IBM PC, so are you now saying that VGA should be here in the 8-bit article as well? Please move CGA to the 16-bit article. --Trixter (talk) 17:29, 13 February 2008 (UTC)

Again, (and I hope the last time), you're still confusing the original IBM PC with any other PC compatible. The VGA was developed by IBM for its PS/2 line, which was microchannel bus, not ISA bus. Non-IBM, compatible VGA and Super VGA cards with 16-bit ISA bus cannot be attached to the original 8-bit ISA bus of the original IBM PC (as Hercules card could). So never was an original IBM PC shipped with VGA (neither from IBM or any of its competitors). Period.

Only add a last note to say that, following your criteria, many of the palettes listed under List of monochrome and RGB palettes also belong to the 16-bit PC section. Bye. -Ricardo Cancho Niemietz (talk) 17:46, 13 February 2008 (UTC)

Yes, this illustrates why your criteria is inconsistent. As previously pointed out, by people other than me, the word size of the CPU is what most people use to determine the size of the machine. Since the 8088 is a 16-bit CPU, CGA should go into the 16-bit section. Since you cannot deny that the 8088 was a 16-bit CPU (word size was 16-bits, registers are 16-bits, etc.) then how can you continue to justify CGA belongs in the 8-bit section? --Trixter (talk) 16:26, 15 February 2008 (UTC)

While I'm still unhappy about the 8-bit placement, the disclaimer you added to the article explains the 8-bit classification adequately (by time period). It's enough to leave it alone, I suppose.

Still, poll one hundred people whether or not the IBM PC was an 8-bit machine or a 16-bit machine and I *highly* doubt you'll find more than 5 or 10 people calling it an 8-bit machine. --Trixter (talk) 21:00, 20 February 2008 (UTC)

Congratulations! You just won your first citation needed tag! X-D. But seriously... For the Wikipedia, it is indifferent that a hundred or a zillion people say anything; reliable sources are always needed to backup any assertion. Please visit the 8-bit article; it states that Intel 8088 (the CPU what the original IBM PC shipped in) was a 8-bit processor (with a note that it was only 8-bit externally). Do you want to start another discussion there? Up to you. And if your hundred people read first about the original IBM PC in the Wikipedia, maybe your statistic changes dramatically... Wikipedia spreads knowledge! ;-) -Ricardo Cancho Niemietz (talk) 09:00, 21 February 2008 (UTC)

The same 8088 article states it has 16-bit registers, which is a 16-bit word size, which is a 16-bit processor. For your education, consult Word_size. Even the 8-bit article makes a note that it was the external data bus was the only 8-bit part, which does not make it an 8-bit CPU. The article states "An 8-bit processor can access 8 bits of data in a single operation, as opposed to a 16-bit processor, which can access 16 bits of data in a single operation." which obviously the 8088 does (ie. ADD AX,BX clearly accesses all 16 bits of both 16-bit registers in a single operation). So don't go throwing accusations of improper research when you yourself are contradicting that very research.

The description of CGA's capabilities in the computer hardware palettes article is very good and sound. I very much like and enjoy the information presented in the article... it's just in the wrong article, that's all. I don't want it removed, simply moved to the proper article. You have not provided sufficient justification for your assessment that the IBM PC was an 8-bit personal computer. --Trixter (talk) 19:03, 21 February 2008 (UTC)

If you count a CPU 16 bit because it has 16 bit registers, then the Motorola 68000 could also be counted as 32 bit CPU because it has 32 bit registers. However, the 68000 is considered a 16 bit cpu since it has a 16 bit data bus. --Fröhn (talk) 20:31, 8 May 2008 (UTC)

From where can access a 16-bit word the 8088 in the original IBM PC architecture? From its main heap RAM not, of course, due to the 8-bit data bus. What the processor in itself was able to interchange a 16-bit word from register to register, no doubt. We are not discussing about the 8088 (Do we are?). But a computer is much more than its processor. The Atari ST and the Commodore Amiga both shipped a Motorola 68000, which was a true 32-bit CPU, but were inserted into 16-bit data buses. So both machines (at least, their original versions) stated and rivaled as 16-bit machines in the whole press in the 16-bit age circa 1990, and this way were sold by their respective brands. Can you give similar (and reliable) proofs that the original 1981 IBM PC was considered anytime a 16-bit machine (not XT, AT, PS/2 or another compatible, only the original one)?

I lack the desire to do so, so I suppose your use of "8-bit" to denote "a range of time" will stand. --Trixter (talk) 21:11, 22 February 2008 (UTC)

And up to date, you have not filled your wiki user page, so I don't know who are you. You can visit my profile, in order to avoid nearly personal attacks. Thanks for your words about the CGA. See you ;-) -Ricardo Cancho Niemietz (talk) 19:55, 21 February 2008 (UTC)

To date I have not made a personal attack against you. If that's how you're taking this, we're done talking. We want the same thing: Accurate information. We disagree and have been discussing it. Nothing personal. I have intentionally not populated my user page because I endeavor never to make any of these discussions personal. --Trixter (talk) 21:11, 22 February 2008 (UTC)

Wow! Tracing your talk page, I discovered some facts about your contributions to the Wikipedia: you have been explicitely pointed to have a conflict of interests and you are "honored" with the most large template for deletion I ever seen! But ironically, you mantain a web about videogames in which my personal profile appears![1] Do you have granted permission to put my photo there? This photo were published by the Spanish magazine "PC mania". The website you link to, www.speccy.org, had granted this permission, and the personal mine too. I invite you to delete the image by yourself. -Ricardo Cancho Niemietz (talk) 10:55, 22 February 2008 (UTC)

Please read that more carefully; the conflict of interest and template for deletion is for MobyGames itself and other people who have acted, not me personally. I am not part of that conflict (and ironically never can be because of "conflict of interest", which prevents me from the basic civil right of defending myself). But all of that has nothing to do with our debate about whether or not the IBM PC is or was considered an 8-bit personal computer. I suppose we will never agree, as you consider the first IBM PC to be in the same class of machines as the Apple and Vic 20 and will never change your mind. If I move the information myself you will simply move it back and we will have a reversion war, which is pointless. --Trixter (talk) 21:11, 22 February 2008 (UTC)

As for your image, it has been removed at your request. It was contributed by a user, so your dispute should really be with the user and not MobyGames (just as it would be with wikipedia), but if you wish to single me out, that's your prerogative. --Trixter (talk) 21:15, 22 February 2008 (UTC)

Thanks to attend my complain about my photo. About MG website and you: I'm always curious about people who I'm discussing with. You have not filled your user page, so I followed your talk page and I found the tags there. I followed the links to MG, I found my photo there, and I became a bit furious. But both of us acted polite, so no future attacks will be expected. Sorry my parlance, I'm not a native English speaker and maybe my words were not tuned enough. I explicitely state here that there is nothing personal about our discussion. We can agree to end it in the basis: a) original 1981 IBM PC have *partially* 8-bit technology (its data bus); b) it was launched in a wave of 8-bit machines; c) all these facts are noted in summaries, with appropiate disclaimers; d) both articles are adecuately cross-linked; e) it was launch by IBM to compete against an 8-bit machine, the Apple II, and this was a brand war (the giant IBM vs. a little company named Apple), not a technology war (16-bit vs. 8-bit); f) parallelism exist between the 32-bit CPU/16-bit data bus of Atari ST and Amiga cases, and g) up to date, neither you and any of your supporters have presented any source to discuss about. But in the future, if new proofs arise to change my criteria, I'll do the change by myself. Bye. -Ricardo Cancho Niemietz (talk) 10:04, 25 February 2008 (UTC)

There should be a note (not much of one) about the fact that the PCjr could display all 16 CGA colours in 320*200 graphics mode and there probably should be short mentions of the 16 bit hardware that used more generic palettes with no implementation specific details (and a hyperlink to the generic palette). But either way, the original PC is considered by most people to have been a 16 bit machine and therefore should be in the 16 bit article, just because you think the data bus and not the machine word should be used does not mean that anyone else on the planet thinks that way.210.50.247.120 (talk) 07:32, 14 February 2008 (UTC)

IBM PCjr (and Tandy 1000 series) section added, with screen simulations. All hyperlynks already available under 8514/A and XGA and Super VGA sections of the List of 16-bit computer hardware palettes article, so no changes needed. Finally, if you'll put a phrase like "Most people considered..." in an article, you'll automatically win a "citation needed" tag. If you can give some authoritative proof or source that a given computer's architecture is determined by its processor's word size only instead of the interconnectig internal data bus (the 16-bit article doesn't clarify this point, and it doesn't cite any source or reference; the summary of Commodore Amiga will give to you a sharper idea about your intended "majority" should mean —in the Amiga's case was the whole press), and the majority of other wikipedians agree with, I'll move the original IBM PC section to its 16-bit counterpart palette's list article. -Ricardo Cancho Niemietz (talk) 09:42, 14 February 2008 (UTC)

Sam Coupé

The Sam Coupé from MGT (Miles Gordon Technology) was an 8 bit and should probably be listed. Any takers ? —Preceding unsigned comment added by Sircompo (talk • contribs) 12:06, 28 February 2008 (UTC)

In advance, any 8-bit related machine with a propietary color palette fits well here; only B&W (as ZX81) machines not. The actual list is far from being complete, and some other representative cases (as the Atari 400/800 XL) are welcome. In fact, machines not listed here are due to lack of info enough to write an enciclopedic section of the subject. If you wish to see the Sam Coupé or any other interesting machine, please provide technical info (refs, URLs, etc.), or edit the page yourself. Yours. Ricardo Cancho Niemietz (talk) 14:45, 28 February 2008 (UTC)

Ricardo, i added some information there about Sam-Coupé, and also about BBC-Micro and Thomson —Preceding unsigned comment added by Nitrofurano (talk • contribs) 15:06, 30 November 2010 (UTC)

SamCoupe has a palette of 128 colours, and their display modes can display up to 16 colours simultaneously (without line interuption). The 128 colour space is 64x2, which means it takes 1/4 of the 512 colour space. There are 2 most significant bits for each rgb primary, and a third less significant bit for all rgb primaries. —Preceding unsigned comment added by 87.196.183.225 (talk) 03:32, 26 January 2010 (UTC)

YPbPr in 8-bit computers!?

"Many of early video chips output composite video YPbPr colors, sometimes inaccurately referred as YUV colors."

This is utter and complete nonsense. There is no such thing as "composite YPbPr". Composite video is a quadrature-modulated YIQ (NTSC) or YUV (PAL) signal, nothing else. YPbPr is a component format and was introduced with SMPTE 240M in the mid/late 1980s, after the Apple II and C64. No old video game console uses YPbPr, internally or otherwise.

"YUV is the color specification used for PAL TV systems, not for composite video;"

Nonsense. PAL is broadcast as RF-modulated composite video, as is NTSC. The author of that line seems to be in complete confusion about what "composite video" is.

"Also, when seen on TV devices through a RF modulator, the perception of these colors may be not corresponding with the original YPbPr ones (most noticiable with NTSC TV color system, due to its YIQ color space plane is not one-to-one compatible with YPbPr and YUV)."

An RF modulator just modulates a baseband composite signal onto a radio frequency carrier --- the same television receiver will produce the same color no matter if fed a baseband composite NTSC or RF-modulated composite NTSC signal.

I guess I'll have to correct this once I have the time. NewRisingSun (talk) 17:47, 8 May 2008 (UTC)

I agree, it's YUV/YIQ. YPbPr is only one step in the color conversion line from YUV/YIQ to RGB. Most video IC's with a composite output do not even have RGB or YPbPr as source, because they simply phase shift the color carrier of the composite signal to produce the different chroma values. Fröhn (talk) 21:00, 8 May 2008 (UTC)

In the article, I'm talking about "color specifications", not "electrical" nor "signal" specifications. That means: YUV not equal to YPbPr not equal to YIQ. Check the articles for the conversion formulae. I already know that PAL/NTSC/SECAM are RF broadcasted, but the article isn't focused on TV. Original models of early computers (as Apple II) couldn't be plugged directly to TV sets, but to manufacturer's ad hoc monitors. That is, they didn't ship built-in RF modulators. Third party manufacturers provide cheap RF modulators, very susceptible to noise, which slighty changed colors, more or less. That kind of facts are what we're talking about. Maybe edition is needed to avoid confussion. Yours. --82.130.195.150 (talk) 14:22, 24 September 2008 (UTC)

I don't see how you justify calling them "YPbPr composite colors." They're not generated that way, and how they're decoded in the monitor is irrelevant. They're generated using phase shifts in YIQ or YUV space, or in simpler cases, just blasting raw bits out the video line. If you want to specify them in YPbPr, that's fine, but calling them YPbPr colors is bogus and deceptive. 24.4.198.162 (talk) 04:17, 18 March 2009 (UTC)

Atari 8 bit modes

The Atari 8 bit computers (400, 800, 600XL, 800XL, 1200XL, 800XE, 65XE, 130XE ...) use a composite output and have a palette of 128 colors (8 lumas * 16 chromas). Following bitmap modes exist:

CTIA/GTIA video chip:

320x192 in 2 colors (monochrome). The chroma for both colors is the same but can be any of the 16 chromas, lumas can be selected without restrictions.

160x192 in 4 colors. Those 4 colors can be chosen from the 128 color palette without restrictions.

GTIA only:

80x192, 16 lumas/1 chroma (monochrome). Unlike the normal 128 color palette this mode allows 16 lumas, but all 16 luma levels are forced to have the same chroma.

80x192, 1 luma/16 chromas. All chromas can be used, but all of them share the same luminance. Pretty useless mode :)

80x192, 9 colors. The 9 colors can be chosen from the 128 color palette without restrictions.

If you wish you can make all modes 200 pixels high, on Atari it's simply a matter of adding more entries to the display list which builds up the screen. It's just that on Atari for some reason 40x24 characters had become standard althought it could easily do 40x25 too.

Also, there is another mode which can be created by a special display list which alternates the 16 luma and the 16 chroma mode every rasterline. Because PAL decoding mixes chroma 50:50 from one rasterline to the other, a 256 color mode with 80x96 pixels resolution can be created without CPU intervention. The downsides of this mode is: Every 2nd rasterline has luminance 0 (almost black) which reduces brightness, and every other rasterline has chrominance 0 (greyscale) which reduces saturation. Fröhn (talk) 9:38, 9 May 2008 (UTC)

Commodore 128

Commodore 128, in VDC mode http://en.wikipedia.org/wiki/MOS_Technology_VDC, should be mentioned, either in a section on its onw or along the CGA discussion. Senbei64 —Preceding undated comment was added at 11:52, 20 August 2008 (UTC)

Article Classifications

At some point, the "8-bit computer", "16-bit computer", etc. denotation should be ditched and something more accurate should be used for a division, like fixed palette vs. flexible palette. That would eliminate any debate over which video subsystem belongs in which article.

One issue that illustrates this is "IBM AT" video subsystem being defined in the "16-bit" article. An assumption is made that the AT shipped with EGA (it was optional) and also that EGA was not available for the original IBM PC (it was, after the launch of the AT). A "fixed palette" vs. "flexible palette" division would eliminate this by properly documenting EGA functionality and not getting hung up on which system it could or could not be inserted into.

Also, a poster above is incorrect in stating that VGA was only available for microchannel systems. IBM, and many others, released 8-bit VGA cards suitable for the original IBM PC. So this is another example of where the system classification is simply a bad way to categorize things. --Trixter (talk) 17:51, 30 November 2009 (UTC)

No British machines!

Why is there no data for British-origin machines such as the ZX Spectrum or BBC Micro? Can anyone get data for these machines? —Preceding unsigned comment added by 82.17.144.241 (talk) 16:19, 30 March 2010 (UTC)

BBC Micro has around 8 or 9 screen modes, using 2, 4 or 16 colours, without colour clash, in resolutions like 160x256, 320x256 and 640x256 - one of the screen modes, in text mode, is very close to Teletext or Minitel - the colour palette is very similar to the 8 colours fullbright from zx-spectrum —Preceding unsigned comment added by Nitrofurano (talk • contribs) 14:12, 30 November 2010 (UTC)

The BBC Micro image is very poor as it doesn't use any form of dithering, making it look like a colour limited version of the Spectrum screen, the below three images show a better rendering in mode 0 (640x512, 2 colours), mode 1 (320x256, 4 colours) and mode 2 (160x256, 16 colours, but 8 are flashing).

(Well I would, but Wikipedia isn't letting me upload files) Tautology0 (talk) 14:47, 15 March 2011 (UTC)

TI 99/4a and TMS99x8 chips

The TI 99/4a home computer should be mentioned alongside the MSX systems, since they used the same video chip. Later MSX systems used upgraded chips, as mentioned in the article. rhyre (talk) 01:37, 10 March 2011 (UTC)

MSX "no apparent order"

Re "There is no apparent logic in this color selection."

The MSX palette looks very clearly to be bitmapped as Red:Green:Blue:Intensity, if you read "dark red" as "dark cyan" and "dark green" as "dark maganta" - which visually they are.

Jgharston (talk) 00:35, 19 December 2013 (UTC)

MSX2+ Screen12 (YJK) palette seems incomplete

The screen-12 of the MSX2+ has 19268 colors. But the palette shown at the right side of the macaw image shows only 5075 colors (counted using GIMP Colors->Info->ColorCube Analysis). Is that right? — Preceding unsigned comment added by 189.26.25.198 (talk) 23:56, 31 December 2012 (UTC)

CGA images don't match

Have a look at the black, cyan, magenta, and light grey image. Now have a look at the 16 colour image. Why is the perch light grey in one image but not in the other image? Also, why are the bright and dark versions of the palette using what seems to be the same image? — Preceding unsigned comment added by CarlKenner (talk • contribs) 08:32, 20 October 2012 (UTC)