Boreas
2006-10-23 13:31:26
I'm doing an experiement with RGB values of colors for auto color purposes and I need help getting the ranges. Note: please keep this on the DL (only SRL members) because if this has good results, we won't want jagex to know that their color shifting is less effective than they think.
Please take 10 mins and help me out here and follow these steps
1. At the moment I want to get lots of samples of the water, rock, and road colors in eastern lumby. Go Here to WT-Fakawi's Guard Fighter to see a nice picture of exactly where I mean
2. Download and open the attachment below
3. Open 3 notepads, type on the first line: water, road, and rock respectively
4. Open RS low detail, head the spot in the picture
5. Use the colorpicker on the water like in the picture
6. Copy that color to setup in line 9
7. Press play
8. Copy the output to the correct notepad so it looks like this
9. repeat 5-8 with road and rock
10. switch to a different world and repeat 5-9, keep doing this until you get bored
11. Post the 3 logs here as attachments or paste their contents as 3 quotes/codetags
12. Get a warm fuzzy feeling for contributing
Note: I am not trying to get every color they could possible be, I am getting the ranges of the red, green, and blue values. Please don't disrupt the format, or the script that analyzes the logs will screw up.
Current Results:
Note: the number of colors for each thing is found by multiplying the range of red, green, and blue values, not the range of colors, as that would be about 100 times as many
Rock:
About 40k colors, which takes about 30-120 seconds for it to to auto calibrate on my computer
It used to be like 20k colors and less time when I had a smaller sample, and its creeping up. 2 mins is still relatively small considering the script only has to do it once and then run for like 12 hours
When I get a bigger sample size I will see if the data gives a rough bell curve. If it does, I will modify my prototype auto coloring procedure to start from the middle and go out, so the times should be shorter.
Please take 10 mins and help me out here and follow these steps
1. At the moment I want to get lots of samples of the water, rock, and road colors in eastern lumby. Go Here to WT-Fakawi's Guard Fighter to see a nice picture of exactly where I mean
2. Download and open the attachment below
3. Open 3 notepads, type on the first line: water, road, and rock respectively
4. Open RS low detail, head the spot in the picture
5. Use the colorpicker on the water like in the picture
6. Copy that color to setup in line 9
7. Press play
8. Copy the output to the correct notepad so it looks like this
water
—————-
12161406
126
145
185
—————-
10844002
98
119
165
9. repeat 5-8 with road and rock
10. switch to a different world and repeat 5-9, keep doing this until you get bored
11. Post the 3 logs here as attachments or paste their contents as 3 quotes/codetags
12. Get a warm fuzzy feeling for contributing
Note: I am not trying to get every color they could possible be, I am getting the ranges of the red, green, and blue values. Please don't disrupt the format, or the script that analyzes the logs will screw up.
Current Results:
Note: the number of colors for each thing is found by multiplying the range of red, green, and blue values, not the range of colors, as that would be about 100 times as many
Rock:
About 40k colors, which takes about 30-120 seconds for it to to auto calibrate on my computer
It used to be like 20k colors and less time when I had a smaller sample, and its creeping up. 2 mins is still relatively small considering the script only has to do it once and then run for like 12 hours
When I get a bigger sample size I will see if the data gives a rough bell curve. If it does, I will modify my prototype auto coloring procedure to start from the middle and go out, so the times should be shorter.