Emulator Test (4/17)

====old paste from 4/17/15====

TODO: test fire underwater

caveat emptor:
  I only tested things in the most obvious manner, maybe Nahoc knows better.
  My math/guesstimation might be off by a little.

measurements in 30fps

base references:
  -running speed: 6 units/frame (horizontal)
  -terminal velocity: 16 units/frame (vertical)

height of 1-1: 183.8 units

===FIRE ON GROUND===

times of fire: 30, 23, 17, 17, 17... frames
distances of fire: 265.975, 190.45, 127.45 units
avg speed: 8.866, 8.280, 7.497 units/frame

===FIRE IN AIR===

times of fire+puff: 35, 28, 22, 22, 22... frames
distances of fire: 308.35, 232.825, 169.825 units
avg speed: 8.81, 8.315, 7.719 units/frame

height loss: a PERFECT fire cancel loses 0.9807 units (negligible, see terminal velocity)
  a perfect frame cancel consists of:
    -float two frames before fire animation ends
    -puff the next frame
    -fire six frames after
  to get the same height:
    -two frames to do float
    -seven frames to do puff
    -one frame to do fire
  to get the same distance:
    -frame perfect, gg

===DOUBLE FIRE===
  time: 93 frames
  distance (stand/walk/run): 734.125, 811.625, 821.124
  avg speed: 7.89, 8.727, 8.829
  (note the last two are misleading because you need to get into walking/running speed)

  It's possible that double-fire is faster for certain parts. >_>
  More testing (maybe with proper TAS tools) upcoming.

===CURLING STONE===
takes 8 frames to enter (and get its speed) after pressing B

initial velocity when
  standing/walking: 6.975
  running: 8.975
  deceleration: 0.025 units/frame^2
  it takes exactly 4 seconds from running curling stone to get back to running speed (6)