The Powder Toy

Subframe rangefinder by LBPHacker

LBPHacker
LBPHacker
52 / 4
3rd Dec 2018
31st Jul 2019
Rangefinding by measuring roundtrip time of photons is an old idea, but this time the measurement is four times faster while still offering the same precision. Also it's subframe, which makes everything easier :P
subframe phot

Comments

  • LBPHacker
    LBPHacker     31st Jul 2019
    Sorry for the bump, I just fixed the buttons that move the obstacle left and right. For some reason I actually left a hole in them instead of colouring the arrows black.
  • SuperVaporWave
    SuperVaporWave     9th Dec 2018
    so you used a script? if so, which one? i am trying to make a logo using an image.
  • LBPHacker
    LBPHacker     9th Dec 2018
    I assume you mean the Subframe Inside logo, because that's the one that took more effort. I made it in Inkscape and exported it as a tiny bitmap, then wrote a Lua script that plotted it in TPT. The areas where the EXOT is visible now were originally transparent in the bitmap. I tried to keep as close to the style of the Intel Inside logo as I could but yeah, at this resolution it doesn't show :P The EXOT part I just filled with random lines using another script.
  • SuperVaporWave
    SuperVaporWave     9th Dec 2018
    how did you make your logo?
  • LBPHacker
    LBPHacker     9th Dec 2018
    The value is the register is sent to the display in every frame after being shifted right by 1 bit, so the display gets D instead of 2D. Initially, like I said, -2C = -6 is in the register, so it's -C that gets displayed. This is -3, but since the display doesn't understand signed integers, it just shows 32765 (and not 65533, because the right shift is not an aritmethic right shift, i.e. it doesn't preserve the sign bit).
  • LBPHacker
    LBPHacker     9th Dec 2018
    So in the case of this particular device, C turns out to be 3 (this is mainly a result of the photon emitters sitting a bit farther in from the vertical surface that faces the object). Basically the there is a 16-bit register that is initialised to -2C, and then whenever a photon returns, its roundtrip time is added to that register by a 16-bit adder. Eventually all four photons return and the register will hold -2C + R1 + R2 + R3 + R4 = 2D. (Comment too long.)
  • LBPHacker
    LBPHacker     9th Dec 2018
    It's not a coincidence. This thing basically measures the roundtrip time of four photons and then sums them up. It takes the photons about 1/4 times as many frames to travel the distance between the object and the device as many pixels there are between the two. As a result, the device will have this sum R1 + R2 + R3 + R4 = 2 * (D + C), where D is the actual distance and C is some device-dependent constant. (Comment too long.)
  • MEOWS
    MEOWS     9th Dec 2018
    32765 (the number on the meter while it is measuring) is very close to 2^15, is this a coincidence or something more?
  • Powderman773
    Powderman773     8th Dec 2018
    i feel like i probably couldnt make this so is pretty awesome +1 2344678
  • LBPHacker
    LBPHacker     7th Dec 2018
    Duh. That's a really awesome notion, but the problem with that is that there tends to be a huge gap in the terminology available to those who know subframe and those who don't, and it's not an easy task to cross this gap from either side. That those who make subframe tutorials don't really get feedback on what terms they should make more clear or explain doesn't help either.