I also noticed that the background waves seem to be dark brown on yellow-brown, which seems impossible. I think there may be color smoothing artifacts in either the picture or the monitor itself.
It may be slightly difficult to distinguish between the red and the brown of the 5153, but who would ever know since it displays yellow instead of brown. The waves may look "browner" when contrasted with the yellow than, say, with green.
I think the appearance of more than 4 colors or the wrong 4 colors is just limitations of taking digital pictures of the monitor; I had the same problem in my photos, the colors in the photos don't look exactly right as reflections off the monitor, ambient room light, moire patterns due to the course dot pitch, and a variety of other issues add to the photo what isn't really there. But it seems clear enough that the 5153 does display yellow instead of brown, very interesting!!
This thread never ceases to amaze me 😳
So my old CGA monitors were not "fully CGA-compatible" at all, or (more likely) the tubes needed recalibration (when I think about it, magenta also was much redder than what I'm used to see on a modern device).
And yes, on the photo it looks like 5 colors, but if you consider what a bit of horizontal convergence errors can do to a pattern like that, - and let's not forget the camera software trying to make the shot look "prettier" under difficult circumstances like this - the picture looks as real as it can get.
wrote:But it seems clear enough that the 5153 does display yellow instead of brown, very interesting!!
Bless my old ET4000! This of course means that all those 4-color screenshots on MobyGames are wrong and should all be replaced. 😀
I think several picture editors have batch processing functions, so that shouldn't be a problem. The real problem are the MobyGames approvers --- they seemed rather skeptical about my request to replace those incorrect RGB Ultima 3 shots with my composite shots, and if Servo didn't step in, my request probably would still be "pending".
wrote:Bless my old ET4000! This of course means that all those 4-color screenshots on MobyGames are wrong and should all be replaced. :)I think several picture editors have batch processing functions, so that shouldn't be a problem.
Yeah, I've been thinking about whether or not to worry about this palette change...it's rather annoying that that one color is so drastically different depending on the type of monitor used! I've been trying to identify and correct shots that use the completely wrong palette though, guess I can start correcting these as well... I use photoshop to do batch processing which usually works rather well.
Is it just 320x200x4 mode that color 6 shows as yellow, or does it apply to text modes as well?
The real problem are the MobyGames approvers --- they seemed rather skeptical about my request to replace those incorrect RGB Ultima 3 shots with my composite shots, and if Servo didn't step in, my request probably would still be "pending".
Actually, I didn't realize there were already RGB screenshots on file (I didn't see those, just the trivia you submitted). I'm not sure the RGB screenshots needed to be replaced, but definitely composite ones should have been added.
Yeah, I've been thinking about whether or not to worry about this palette change...it's rather annoying that that one color is so drastically different depending on the type of monitor used! I've been trying to identify and correct shots that use the completely wrong palette though, guess I can start correcting these as well... I use photoshop to do batch processing which usually works rather well.
Is it just 320x200x4 mode that color 6 shows as yellow, or does it apply to text modes as well?
No, it would have to be all modes, including text, otherwise it wouldn't make any sense. If you see or detect brown, change it to yellow. If you can't tell, then god knows what was used to take the screenshot and efforts should be made to replace the screenshot.
Actually, I didn't realize there were already RGB screenshots on file (I didn't see those, just the trivia you submitted). I'm not sure the RGB screenshots needed to be replaced, but definitely composite ones should have been added.
The RGB shots need not be omitted because they have always been accurate.
But speaking of accuracy, I would cut in half those screenshots of Hercules graphics that show 720x696. The true resolution is 720x348 and while the video display may have stretched the graphics vertically to some extent, it would double the pixels vertically. This is especially true because most Hercules cards would have been connected to the 5151 Monochrome monitor, which had an unusual aspect ratio of 5:4. Sierra's 3-D Helicopter Simulator's page is an example.
The RGB shots need not be omitted because they have always been accurate.
That's an odd standard of accuracy you have.
And if you care about the correct aspect ratio, you'll have to stretch all 320x200 shots to 320x240, and all NES shots to from 256x240 to ~288x240 as well.
otherwise it wouldn't make any sense.
well, I was quite surprised at how much of a difference there was depending on the monitor used, so even if that wouldn't make sense already stranger things have happened 😀
With hercules the display stretched the 720x384 resolution by quite a lot from what I remember; 5:4 (about 1.25:1) is closer to a square image than the squished rectangular image 720x384 would display as. Doubling the height to 696 doesn't provide exactly the right aspect ratio, but I felt it looked a lot closer than leaving the height at 384.
On a Sierra related note, do games like Mickey's Space Adventure, Troll's Tale and Oil's Well support color on RGB monitors? I can only get them to display in composite mode, the shots currently on moby are the results of a VGA card mangling the composite mode. It seems odd for Sierra to not support 4 color CGA, but I haven't figured out how to get the mode working if it's there.
Well, the RGB shots show the correct colors, but thats easy. Aspect ratio correction, especially with such low resolutions, is more difficult if you want to keep sharp, same-size pixels. To really put a 320x200 screen capture into 320x240 and still have proper pixels, you would have to upscale the screenshot to 1600x1200.
IBM did change the color from yellow to brown for the PCjr., and they had to add a circuit to their 4863 PCjr. Color Monitor to display brown when the monitor received a 0110 (IRGB) signal. Remember that the 5153 and the 4863 could only accept a 4-bit signal and therefore display 16 colors. Their VGA specifically sets the brown color in all modes. Presumably the EGA does as well for all modes, as you can see below, you can find all 16 CGA colors in the 64-color EGA color palette:
Of course, if you connect a 5153 to the EGA instead of the 5154, you must set it up on the PCB and the card should be able to show all 16 colors, but what EGA programmers intended to show as brown will be shown as dark yellow.
Sorry, but I own several 5153s and color 6 is definitely brown, not yellow. Digger is a bad example to look at because the game flashes between the RGB low-I and RGB High-I palettes during gameplay.
It's brown, and if people doubt that I guess I'll have to take some screenshots. I've owned true IBM CGA machines since 1985 -- believe me, it's brown.
If anyone needs shots to verify this, let me know. And tell the guy who took the previous 5153 shots to disable his flash -- gee, I wonder if that's why it looks lighter...
It has nothing to do with his flash and everything to do with "white balance" on the camera. Possibly. Strangely, I remember both color palettes for Digger.
More info here: http://en.wikipedia.org/wiki/White_balance
Every emulator I have tried always used the non-intense Red, Green and Brown palette and I have never seen any evidence of palette changing to the cyan, magenta, white or to the intense version, except perhaps for the New High Score screen. But consider these other two photos:
This photo shows some somewhat bright colors, I wonder if they could be the intense version of palette 1.
This photo gives another angle of the first level screen without the flash. The fellow I got the screenshots from said it looked like dark yellow to him and he said his 5153 was in good condition.
Digger alone is not enough. I will shoot a video showing the full 16-color palette, then I'll run Digger and show how it flashes palettes, and how throughout it all, #6 is brown (except the high-intensity palette).
Did you once display all 16 colors at the same time and calibrate the monitor?
I can't find either my DV camera OR my digital camera right now... dammit. When I find either one, I'll come up with a demonstration (as well as the 16-color composite palette).
Anyway, please trust me, I've been around CGA for over two decades. On a real stock IBM CGA, connected to a real 5153, it's brown. Please display all 16 colors at once using this program:
http://www.oldskool.org/pc/cgacal
Hopefully we can put this to rest. I'd be curious to see what yours looks like with all 16 displayed.
Anyway, please trust me, I've been around CGA for over two decades.
He has pictures to prove his point, and all you have to offer is "trust me"? That's a weak case. Aren't you the guy who claimed he got his CGA colors from the "6845 specification" "as a deterrent to people guessing at the colors", even though that document contains no such information? That doesn't exactly further your trustworthiness, either, just making up stuff like that.
When I find either one, I'll come up with a demonstration
Do that. Until then, quit acting as if you're the sole authority on what colors are correct. Make sure you also tell us the part and serial number of that 5153, because if it turns out that your monitor isn't just miscalibrated, then there's also the possibility that the additional circuity to turn yellow into brown (after all, if you follow the RGBI model strictly, you DO get yellow rather than brown) was added later to the same model as an "improvement", and that this revision might be noted in some way on the back of the monitor; maybe the manufacturing date is of interest as well. You've said that you've owned IBM machines since 1985, which means that it's well likely that any 5153 you've had was such a revised one, if there was one. If that is so, I'll gladly change the Wikipedia entry from "the IBM 5153 shows color 6 as dark yellow" to "early IBM 5153 models show color 6 as dark yellow".
I reject the claim that a camera's flash can turn brown into yellow but leave all other colors untouched. Therefore, any different result you might bring to the table will be considered in addition to GH's pictures; they will not be ignored.
(as well as the 16-color composite palette).
Good idea, because no one in this thread has ever seen composite colors. 😜 By the way, there is no such thing as "THE" 16-color composite palette; it depends at least on the settings of registers 0x3d8 and 0x3d9, and to a lesser extent on the YIQ-to-RGB matrices as well as the "hue" and "tint" settings in the monitor/TV's chroma decoder.
Of course, what IS entirely possible is that the additional circuity to turn yellow into brown (after all, if you follow the RGBI model strictly, you DO get yellow rather than brown) was added later to the same model as an "improvement".
I was going to suggest just that, that there could be two models of 5153. The early, pre-PCjr. or pre-EGA model would show dark yellow. Then the PCjr. comes along, and its monitor turns what should be yellow into brown, probably because brown is sort of necessary for paint programs. Then the EGA is released and with the 5154 it designates brown (using the palette registers. At that time, IBM decides to add the "browning circuit" to the 5153 design to ensure that the colors are uniform across their monitors. Also, the 5154 is expensive and many people would rather buy the cheaper 5153 and lose the 350-line modes.
Is there a manufacturing date visible on the 5153? Also, I believe the only external controls for the 5151, 5153 and 5154 are brightness and contrast. The 5151 doesn't even have an on/off button, you are supposed to plug it into the PC's power supply and it turns on when the PC does.
Unfortunately, I received the pictures from an individual who will shortly be selling off his IBM PC stuff, he took them as a favor to me and digger was the only game he could find with the proper palette. I'm not sure if I should ask him to run the program after having taken those pictures for me.
Great Hierophant: it would be great if you could ask him to make another screenshot of digger after eating a cherry. Cause after eating a cherry brown turns into yellow. So i wonder what would happen in his case, when he already has yellow before eating a cherry. 😕
I forgot about the cherry, so it would be useful to see what the colors look like once a cherry has been eaten. I know now what Trixter meant when he mentioned palette changes. We could gauge the intensity that way.
Does anyone on the forum own a 5153 that they can test with the calibration program I posted last night? I'm trying to figure out a way to prove to NewRisingSun that it really is brown short of inviting the guy over to my house (although if he lives close to Chicago he's welcome to do so). I own 5 (maybe more) CGA monitors include my PCjrs and Tandys and evan an AT&T D25 double-scanrate CGA monitor and they *all* show brown for #6. And that's the way I've always remembered it for two decades.
Here's something to chew on: If it wasn't supposed to be brown, then why did IBM make it brown in the text color section of the default VGA palette?
NewRisingSun, what would it take to convince you? Scanned photos from advertisements or IBM catalogs? If my word isn't good enough for you any more, then please tell me what is and I'll come up with proof. I'm willing to suck up all criticisms regarding the "mc6845 spec color" wikipedia comment I made if it will help set this issue straight.
