home.social

#bluenoise — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #bluenoise, aggregated by home.social.

  1. Saturday AM, a story in two acts...

    (Incl. placing crumbs by approaching it like a #PoissonDisk / #BlueNoise problem... 😅)

    #CakeDay

  2. Saturday AM, a story in two acts...

    (Incl. placing crumbs by approaching it like a #PoissonDisk / #BlueNoise problem... 😅)

    #CakeDay

  3. Saturday AM, a story in two acts...

    (Incl. placing crumbs by approaching it like a #PoissonDisk / #BlueNoise problem... 😅)

    #CakeDay

  4. Saturday AM, a story in two acts...

    (Incl. placing crumbs by approaching it like a #PoissonDisk / #BlueNoise problem... 😅)

    #CakeDay

  5. Saturday AM, a story in two acts...

    (Incl. placing crumbs by approaching it like a #PoissonDisk / #BlueNoise problem... 😅)

    #CakeDay

  6. I also tried asking various LLM models to give me C code to generate blue noise. This didn't work. Like, at all.

    Then I just asked them to describe what blue noise is, and oh boy, did I ever get wild answers.

    One said blue noise is errors in electric signal. Another said it's blue noise because it's shades of blue. And so on.

    It's like you take a classroom of students and ask them to write an essay about something they might vaguely know about. Like blue noise.

    You get a bunch of very confident and wrong answers.

    #llm #ai #noise #bluenoise #code

  7. I also tried asking various LLM models to give me C code to generate blue noise. This didn't work. Like, at all.

    Then I just asked them to describe what blue noise is, and oh boy, did I ever get wild answers.

    One said blue noise is errors in electric signal. Another said it's blue noise because it's shades of blue. And so on.

    It's like you take a classroom of students and ask them to write an essay about something they might vaguely know about. Like blue noise.

    You get a bunch of very confident and wrong answers.

    #llm #ai #noise #bluenoise #code

  8. I also tried asking various LLM models to give me C code to generate blue noise. This didn't work. Like, at all.

    Then I just asked them to describe what blue noise is, and oh boy, did I ever get wild answers.

    One said blue noise is errors in electric signal. Another said it's blue noise because it's shades of blue. And so on.

    It's like you take a classroom of students and ask them to write an essay about something they might vaguely know about. Like blue noise.

    You get a bunch of very confident and wrong answers.

    #llm #ai #noise #bluenoise #code

  9. I also tried asking various LLM models to give me C code to generate blue noise. This didn't work. Like, at all.

    Then I just asked them to describe what blue noise is, and oh boy, did I ever get wild answers.

    One said blue noise is errors in electric signal. Another said it's blue noise because it's shades of blue. And so on.

    It's like you take a classroom of students and ask them to write an essay about something they might vaguely know about. Like blue noise.

    You get a bunch of very confident and wrong answers.

    #llm #ai #noise #bluenoise #code

  10. I also tried asking various LLM models to give me C code to generate blue noise. This didn't work. Like, at all.

    Then I just asked them to describe what blue noise is, and oh boy, did I ever get wild answers.

    One said blue noise is errors in electric signal. Another said it's blue noise because it's shades of blue. And so on.

    It's like you take a classroom of students and ask them to write an essay about something they might vaguely know about. Like blue noise.

    You get a bunch of very confident and wrong answers.

    #llm #ai #noise #bluenoise #code

  11. The blue noise texture itself I generated by generating N candidate random pixels (x/y/gray for 3d coordinate) and picking the one that's farthest away from all the existing ones. Brute force. Took maybe an hour. Fun!

    There may be better ways to generate it, and there may be better blue noise textures out there, but this one's mine. It also tiles.

    #noise #bluenoise #graphics

  12. The blue noise texture itself I generated by generating N candidate random pixels (x/y/gray for 3d coordinate) and picking the one that's farthest away from all the existing ones. Brute force. Took maybe an hour. Fun!

    There may be better ways to generate it, and there may be better blue noise textures out there, but this one's mine. It also tiles.

    #noise #bluenoise #graphics

  13. The blue noise texture itself I generated by generating N candidate random pixels (x/y/gray for 3d coordinate) and picking the one that's farthest away from all the existing ones. Brute force. Took maybe an hour. Fun!

    There may be better ways to generate it, and there may be better blue noise textures out there, but this one's mine. It also tiles.

    #noise #bluenoise #graphics

  14. The blue noise texture itself I generated by generating N candidate random pixels (x/y/gray for 3d coordinate) and picking the one that's farthest away from all the existing ones. Brute force. Took maybe an hour. Fun!

    There may be better ways to generate it, and there may be better blue noise textures out there, but this one's mine. It also tiles.

    #noise #bluenoise #graphics

  15. The blue noise texture itself I generated by generating N candidate random pixels (x/y/gray for 3d coordinate) and picking the one that's farthest away from all the existing ones. Brute force. Took maybe an hour. Fun!

    There may be better ways to generate it, and there may be better blue noise textures out there, but this one's mine. It also tiles.

    #noise #bluenoise #graphics

  16. Played around with blue noise a bit. The attached image has four different variations (using the zx spectrum next color space as target). The source image is from thispersondoesnotexist.

    First is a straight conversion with no dither, second is converted to 32 colors with pattern dither in photoshop, third is blended with blue noise texture at 10% and last has 5% uniform noise added before conversion.

    It's not exactly apples to apples comparison because these are all done in photoshop and I didn't tweak things to look as good as possible (given that tweaking a value is a undo-undo-undo, tweak, collapse, filter), but the blue noise looks fairly good.

    #noise #bluenoise #dither #graphics

  17. Played around with blue noise a bit. The attached image has four different variations (using the zx spectrum next color space as target). The source image is from thispersondoesnotexist.

    First is a straight conversion with no dither, second is converted to 32 colors with pattern dither in photoshop, third is blended with blue noise texture at 10% and last has 5% uniform noise added before conversion.

    It's not exactly apples to apples comparison because these are all done in photoshop and I didn't tweak things to look as good as possible (given that tweaking a value is a undo-undo-undo, tweak, collapse, filter), but the blue noise looks fairly good.

    #noise #bluenoise #dither #graphics

  18. Played around with blue noise a bit. The attached image has four different variations (using the zx spectrum next color space as target). The source image is from thispersondoesnotexist.

    First is a straight conversion with no dither, second is converted to 32 colors with pattern dither in photoshop, third is blended with blue noise texture at 10% and last has 5% uniform noise added before conversion.

    It's not exactly apples to apples comparison because these are all done in photoshop and I didn't tweak things to look as good as possible (given that tweaking a value is a undo-undo-undo, tweak, collapse, filter), but the blue noise looks fairly good.

    #noise #bluenoise #dither #graphics

  19. Played around with blue noise a bit. The attached image has four different variations (using the zx spectrum next color space as target). The source image is from thispersondoesnotexist.

    First is a straight conversion with no dither, second is converted to 32 colors with pattern dither in photoshop, third is blended with blue noise texture at 10% and last has 5% uniform noise added before conversion.

    It's not exactly apples to apples comparison because these are all done in photoshop and I didn't tweak things to look as good as possible (given that tweaking a value is a undo-undo-undo, tweak, collapse, filter), but the blue noise looks fairly good.

    #noise #bluenoise #dither #graphics

  20. Played around with blue noise a bit. The attached image has four different variations (using the zx spectrum next color space as target). The source image is from thispersondoesnotexist.

    First is a straight conversion with no dither, second is converted to 32 colors with pattern dither in photoshop, third is blended with blue noise texture at 10% and last has 5% uniform noise added before conversion.

    It's not exactly apples to apples comparison because these are all done in photoshop and I didn't tweak things to look as good as possible (given that tweaking a value is a undo-undo-undo, tweak, collapse, filter), but the blue noise looks fairly good.

    #noise #bluenoise #dither #graphics

  21. I never thought about the fact that random number generators in games might feel unfair to players. You might not roll that 10% chance in 20 tries while another player hits it twice in a row. So there are algorithms to make "10% chance" feel more like "about once every 10 rolls".

    mastodon.gamedev.place/@demofo

    #BlueNoise #RNG

  22. I never thought about the fact that random number generators in games might feel unfair to players. You might not roll that 10% chance in 20 tries while another player hits it twice in a row. So there are algorithms to make "10% chance" feel more like "about once every 10 rolls".

    mastodon.gamedev.place/@demofo

    #BlueNoise #RNG

  23. I never thought about the fact that random number generators in games might feel unfair to players. You might not roll that 10% chance in 20 tries while another player hits it twice in a row. So there are algorithms to make "10% chance" feel more like "about once every 10 rolls".

    mastodon.gamedev.place/@demofo

    #BlueNoise #RNG

  24. I never thought about the fact that random number generators in games might feel unfair to players. You might not roll that 10% chance in 20 tries while another player hits it twice in a row. So there are algorithms to make "10% chance" feel more like "about once every 10 rolls".

    mastodon.gamedev.place/@demofo

    #BlueNoise #RNG

  25. I never thought about the fact that random number generators in games might feel unfair to players. You might not roll that 10% chance in 20 tries while another player hits it twice in a row. So there are algorithms to make "10% chance" feel more like "about once every 10 rolls".

    mastodon.gamedev.place/@demofo

    #BlueNoise #RNG