CompuServe Thread

#HAM-E and DCTV compared

6 messages in this thread
#11673From: Don Curtis/SYSOPJul 1, 1991 1:17 AM
John, I once asked the same question, and added the comment that I defy anyone to distinguish between color x and color x+1 on a 16 million color board. The answer was that in fact, no one can distinguish adjacent colors, but by using them, you get a much smoother transition between colors or shades and in essense, you end up with a much more lifelike rendering. Having thought about this for a while (and seen some 24 bit boards in action) I must agree with them. Consider using a color board where you could tell the difference between color X and color X+1. Anytime there was a transition from color X to color X+1 you would notice it (since you can tell the difference). However, if you had 10 times more colors available (and thus the original X and X+1 now become the new colors X and X+10) you now have 10 steps of changes in color between…thus you will no longer see that visible jump between adjacent colors. Don
#11680From: John PendergrassJul 1, 1991 11:37 AM
No arguments, Don. I just wonder where the 'more colors' argument starts to lose importance? Are 1million color sufficient to do the transitions, as you mention? or are 2/4/8/16 million better and better? Take two areas, ten pixels apart horizontally. The left pixel is color 50,150,200 (RGB) and the right pixel is color 60,160,210. Will any person, much less an average user, notice if the nine transition pixels are filled with ten interpolated shades, or if it just had two transition colors? I don't know. (shoot, I don't even know if the above numbers are good examples; they were just a shot at getting two nearby shades) But plan on doing some experimenting with 'average' users here with my FireCracker24. Interesting question. –jp–seattle–
#11697From: Charles BlaquiereJul 1, 1991 11:33 PM
Don, you're right: that's the reason why we need 18 to 24 bits. Mention 16 million colors to most people and they'll take you for a megalomaniac, and an ambitious one at that. B^D But, once you consider that it boils down to 256 shades of red/green/blue, you realize that in order to have smooth graduated backgrounds, you do need that many colors. With 64 shades of each primary color, people will be able to see bands in a supposedly smooth grey scale. (This is why I like having the dithered pseudo-24-bit option in HAM-E) You need 128 or 256 shades to have true smoothness, and computers being the way they are, it's basically as easy to give the user 8 bits/color as 7 bits, so we end up with 256 shades of red/green/blue and 16 million colors. Blaq! – via Whap!
#11719From: SyndesisJul 2, 1991 5:03 PM
In the PC world, it's now easy to get a super VGA board with a new kind of DAC chip that accepts 15-bit direct color, beyond the standard 256 color VGA. This add-on DAC costs about $75, added to a $400 super VGA card… it comes with Windows drivers, too. Blasting bits into 15-bit pixels is much more efficient than traditional VGA, so this board can be *faster* than VGA.
#11735From: Black Belt SystemsJul 3, 1991 1:22 AM
There are other issues… For instance, take orange. 100% red, 50% green, 25% blue. With 24 bits, you can have 64 shades of this exact orange. No more, no less. You get 256 shades of pure red, green and blue… also, the 7 brigh primaries that are equal compnents of each: cyan, magenta, yellow – white – red, green, and blue. Any other color, you have fewer shades available. Alll you have to do si think about it this way, and you come to yet another realization why you need the larger dynamic range(s). Ben Amateur Radio Callsign is A A 7 A S
#11803From: Charles BlaquiereJul 4, 1991 10:31 PM
Ben, I did realize as I was writing that you only get 256 shades of red, green, blue and the other combinations of those primary colors. Just for the record. Blaq! – via Whap!