mirror of
https://github.com/ChuckBuilds/LEDMatrix.git
synced 2026-08-01 16:58:06 +00:00
Sub-pixel scrolling: motion at the frame rate, not the pixel rate
With integer positioning the number of distinct frames per second equals the scroll speed in px/s, however fast the loop renders. Measured at 50px/s and 78.7fps, 36% of frames were byte-identical: the extra frames cost work and bought no motion, and what was left was 50 discrete 1px steps a second. Two things were wrong with the pre-existing sub-pixel support. get_visible_portion never consulted sub_pixel_scrolling — it always took the integer path, so the flag and _get_visible_portion_subpixel were dead code. And that implementation needed scipy.ndimage.shift, which is not installed on the target devices (HAS_SCIPY is False there), so it would not have interpolated even if reached. Verified both: positions 1000.0 and 1000.5 produced identical frames either way. Blending is now wired up and implemented with numpy. Two details make it affordable: slice cached_array directly instead of building two PIL images only to convert them straight back (the naive version measured 15x the integer path), and use fixed-point uint16 multiply-add rather than float32, which suits the Pi's cores and gives finer weighting than the panel can resolve. Result 0.939ms against 0.237ms — 0.70ms added per frame, a 1065fps ceiling. Measured on hardware: 81.2 fps with blending on, against 78.7 with it off, so no cost within noise — and every frame is now a distinct position rather than one in three being a repeat. The trade is a slight horizontal softening of text, since each frame blends two positions. Set smooth_scroll false for maximum crispness. Also benchmarked and cleared as non-issues: extending the strip costs 9.4ms on an 11,000px strip and trimming 2.5ms, both under one frame at this rate. Co-Authored-By: Claude <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01KEZK1P1Q1fu5pcuVrkrCFZ
This commit is contained in:
@@ -243,3 +243,94 @@ class TestContinuousScrollingEndToEnd:
|
||||
|
||||
assert max(widths) < 3000, f"strip grew unbounded: max {max(widths)}"
|
||||
assert not sh.scroll_complete, "continuous strip should never complete"
|
||||
|
||||
|
||||
class TestSubPixelBlending:
|
||||
"""
|
||||
Integer positioning quantises motion to whole pixels, so distinct frames per
|
||||
second equals scroll speed regardless of frame rate — at 50px/s and 78fps,
|
||||
36% of frames were identical. Blending between neighbouring positions gives
|
||||
motion at the frame rate instead.
|
||||
"""
|
||||
|
||||
def _strip(self, width=2000):
|
||||
rng = np.random.default_rng(0)
|
||||
arr = (rng.random((H, width, 3)) * 255).astype(np.uint8)
|
||||
sh = helper()
|
||||
sh.create_scrolling_image([Image.fromarray(arr)],
|
||||
item_gap=0, element_gap=0, lead_gap=0)
|
||||
return sh
|
||||
|
||||
def _frame(self, sh, pos, subpixel):
|
||||
sh.sub_pixel_scrolling = subpixel
|
||||
sh.scroll_position = pos
|
||||
return np.asarray(sh.get_visible_portion()).astype(int)
|
||||
|
||||
def test_integer_mode_ignores_the_fraction(self):
|
||||
sh = self._strip()
|
||||
a = self._frame(sh, 500.0, False)
|
||||
b = self._frame(sh, 500.9, False)
|
||||
assert np.array_equal(a, b), "integer positioning should not move sub-pixel"
|
||||
|
||||
def test_blending_moves_within_a_pixel(self):
|
||||
sh = self._strip()
|
||||
a = self._frame(sh, 500.0, True)
|
||||
b = self._frame(sh, 500.5, True)
|
||||
assert not np.array_equal(a, b)
|
||||
|
||||
def test_zero_fraction_matches_the_integer_frame(self):
|
||||
# No interpolation to do, so it must be pixel-identical and take the
|
||||
# cheap path.
|
||||
sh = self._strip()
|
||||
assert np.array_equal(self._frame(sh, 700.0, True),
|
||||
self._frame(sh, 700.0, False))
|
||||
|
||||
def test_blend_is_monotonic_between_neighbours(self):
|
||||
# Marching the fraction from 0 to 1 should approach the next integer
|
||||
# frame, not wander.
|
||||
sh = self._strip()
|
||||
target = self._frame(sh, 501.0, False)
|
||||
dists = []
|
||||
for frac in (0.0, 0.25, 0.5, 0.75):
|
||||
f = self._frame(sh, 500.0 + frac, True)
|
||||
dists.append(np.abs(f - target).mean())
|
||||
assert dists == sorted(dists, reverse=True), f"not converging: {dists}"
|
||||
|
||||
def test_blend_endpoints_bracket_the_two_frames(self):
|
||||
sh = self._strip()
|
||||
near = self._frame(sh, 500.0, False)
|
||||
far = self._frame(sh, 501.0, False)
|
||||
mid = self._frame(sh, 500.5, True)
|
||||
# Every blended pixel must lie between its two sources.
|
||||
lo = np.minimum(near, far)
|
||||
hi = np.maximum(near, far)
|
||||
assert (mid >= lo - 1).all() and (mid <= hi + 1).all()
|
||||
|
||||
def test_output_size_and_mode_are_unchanged(self):
|
||||
sh = self._strip()
|
||||
sh.sub_pixel_scrolling = True
|
||||
sh.scroll_position = 300.4
|
||||
frame = sh.get_visible_portion()
|
||||
assert frame.size == (W, H)
|
||||
assert frame.mode == 'RGB'
|
||||
|
||||
def test_works_near_the_end_of_the_strip(self):
|
||||
# One of the two slices wraps here; must not raise or missize.
|
||||
sh = self._strip(width=600)
|
||||
sh.sub_pixel_scrolling = True
|
||||
sh.scroll_position = float(600 - W // 2) + 0.5
|
||||
frame = sh.get_visible_portion()
|
||||
assert frame is not None and frame.size == (W, H)
|
||||
|
||||
def test_works_at_the_very_last_column(self):
|
||||
sh = self._strip(width=600)
|
||||
sh.sub_pixel_scrolling = True
|
||||
sh.scroll_position = 599.5
|
||||
assert sh.get_visible_portion().size == (W, H)
|
||||
|
||||
@pytest.mark.parametrize("frac", [0.01, 0.1, 0.33, 0.5, 0.67, 0.9, 0.99])
|
||||
def test_never_raises_across_the_fraction_range(self, frac):
|
||||
sh = self._strip()
|
||||
sh.sub_pixel_scrolling = True
|
||||
sh.scroll_position = 400.0 + frac
|
||||
assert sh.get_visible_portion().size == (W, H)
|
||||
|
||||
Reference in New Issue
Block a user