feat: a plot can carry a phosphor density, one image for every acquisition

`Plot::set_density()` gives a plot a **phosphor density**: a screen-sized intensity image behind
its traces, into which each acquisition is deposited and which fades by a time constant or by a
per-acquisition factor. It is the library's and not a host's because the library is the side that
knows a trace's device geometry -- a host would be re-rasterising traces the renderer has already
staged, at a second derivation of the same mapping.

It is an **image and not a set of retained traces**, and that is the whole of its cost: one `f32`
cell per device pixel of the axes rect (`DensityGrid`), a cell's value a *number of sweeps*, so a
display of a thousand acquisitions costs what one costs and there is no depth to get wrong.
Measured on a 640x360 viewport's axes: **573x313 = 179 349 cells, 717 396 bytes** of grid, and one
byte per cell uploaded on the frame the grid moved.

- `Plot::accumulate_density()` deposits the acquisition on screen: every visible trace's window is
  mapped to sample indices and gathered exactly as `Renderer::draw_traces()` maps it, then walked
  column by column -- so the deposit is bounded by the axes' columns and not by the samples behind
  them -- and a cell takes **1.0 once per trace per sweep**, however many of that trace's own
  samples crossed it. Counting the hits instead put a single dense sweep at the ceiling of the
  brightness scale and hid the sweeps behind it, and it made a cell mean "how much the beam moved
  here" rather than "how many acquisitions were here"; the dedupe is a stamp per grid *row* (one
  `u64`, keyed by trace and column) so two sweeps agreeing on a cell still count twice. Dwell is a
  different question the library already answers (`Style::intensity_grading`).
- `decay_density(dt_s)` fades by the clock, `clear_density()` and switching the density off drop
  the image at once. `DensityConfig` carries `full_scale` and `gamma`, applied on the CPU when the
  grid changed: the value a host reads back through `ink()`/`peak()` and the picture it sees are
  one derivation, and a frame whose grid did not move skips the conversion and the upload entirely
  (`FrameStats::density_cells`/`density_upload_bytes`, the latter the *frame's* total over every
  row, which one row's assignment used to overwrite).
- The renderer draws it as one attribute-less quad over the axes rect (`kDensityVS`/`kDensityFS`,
  `DensityProgram`), over the grid and the invalid-span bands and **under the traces**, inside the
  axes' clip: it reserves no layout space and nothing on it can be picked, exactly as
  `Plot::set_invalid_spans()` does. `Style::density` is its tint.
- The two fades are the two questions a persistence display can be asked -- "still visible a second
  later" and "the N-th back is gone" -- so the choice is the caller's.

Six cases, all in the suite: four plot-level ones (the grid's resolution and memory, the
per-acquisition schedule **584 -> 876 -> 1022** ink with `peak_steady=2.000`, `exp(-dt/tau)` off
the ink, and the deposit's own geometry including a NaN as a gap) and one on the real driver
(`a_phosphor_density_is_drawn_under_the_traces`: `cells=179349 uploaded=179349` on the moved frame
and `uploaded=0` on the identical one after it, with the trace's stroke keeping its own colour over
the ink). The suite is **396 cases, 0 failed** and all four ctest entries pass.
This commit is contained in:
Martino Ferrari committed 2026-10-16 17:25:39 +02:00
1 parent 3ad02e5120
commit 6a6915e64e
17 files changed
+1199 -13

No files matched your search

+136
View File
@@ -7519,3 +7519,139 @@ PLOT_TEST(the_box_zoom_frame_lands_exactly_where_it_says) {
PLOT_CHECK(outside_after == outside_before);
p.clear_selection_rect();
}
/// The phosphor density is a *background* image drawn by the plot: it covers the axes rect in
/// the style's one tint, it is behind the traces -- a trace drawn over the ink is the trace's
/// own colour and not a blend -- and it costs an upload on the frame the grid moved and nothing
/// on the frame it did not.
///
/// The trace is moved between the two accumulations (the same `ArraySeries`, committed to new
/// values, which is how a host re-points a trace it cannot retarget) so that the image holds a
/// sweep at y = 0 *and* the trace is drawn at y = +0.5. Without that, a trace and its own ghost
/// would be in the same row and the density would be entirely hidden by the ink over it -- which
/// is exactly the ordering being asserted, so the two have to be told apart by position.
PLOT_TEST(a_phosphor_density_is_drawn_under_the_traces) {
plottest::HeadlessContext ctx(kW, kH);
plot::Status st;
PLOT_CHECK(ctx.create(&st));
PLOT_CHECK(plot::gl::load(nullptr));
plot::Renderer r;
PLOT_CHECK(r.init(&st));
plot::data::ArraySeries s(std::vector<plot::f32>{0.0f, 0.0f}, 0.0, 1.0);
plot::Plot p;
p.set_viewport({0.0f, 0.0f, static_cast<plot::f32>(kW), static_cast<plot::f32>(kH)});
const plot::u32 ax = p.add_x_axis({"t", {0.0, 1.0}, false, plot::AxisPlace::Bottom, false});
const plot::u32 ay = p.add_y_axis({"v", {-1.0, 1.0}, false, plot::AxisPlace::Left, false});
const plot::Color trace_color = plot::Color::hex(0x00E070);
PLOT_CHECK(p.add_trace({&s, trace_color, 1.5f, ax, ay, "flat"}) == 0u);
p.set_x_range(ax, {0.0, 1.0});
p.set_y_range(ay, {-1.0, 1.0});
// The baseline: no density at all, and no grid to draw one from.
r.begin_frame(kW, kH, kFrameRect);
r.draw(p);
r.end_frame();
PLOT_CHECK(!r.last_frame_had_gl_error());
PLOT_CHECK(r.stats().density_cells == 0u);
PLOT_CHECK(r.stats().density_upload_bytes == 0u);
const std::vector<plot::u8> without = read_frame(ctx);
plot::DensityConfig cfg;
cfg.enabled = true;
cfg.decay = plot::DensityDecay::PerAcquisition;
cfg.per_acquisition = 0.5f;
p.set_density(cfg);
p.accumulate_density(); // a sweep at y = 0
s.reset(std::vector<plot::f32>{0.5f, 0.5f}, 0.0, 1.0);
p.accumulate_density(); // and the next one at y = +0.5, the trace's own row
r.begin_frame(kW, kH, kFrameRect);
r.draw(p);
r.end_frame();
PLOT_CHECK(!r.last_frame_had_gl_error());
dump_if_requested(r, "density.ppm");
const plot::FrameStats first = r.stats();
const std::vector<plot::u8> with = read_frame(ctx);
// The image is drawn -- it changed pixels -- and every pixel it changed is inside the axes:
// a density reserves no layout space and paints nothing over a label strip.
const plot::Rect axes = p.layout().axes;
plot::u32 lo_x = kW, lo_y = kH, hi_x = 0u, hi_y = 0u, differing = 0u;
for (plot::u32 y = 0u; y < kH; ++y) {
for (plot::u32 x = 0u; x < kW; ++x) {
const plot::usize at = (plot::usize(y) * kW + x) * 4u;
if (with[at] == without[at] && with[at + 1u] == without[at + 1u] &&
with[at + 2u] == without[at + 2u]) {
continue;
}
++differing;
lo_x = std::min(lo_x, x);
lo_y = std::min(lo_y, y);
hi_x = std::max(hi_x, x);
hi_y = std::max(hi_y, y);
}
}
PLOT_CHECK(differing > 0u);
PLOT_CHECK(static_cast<plot::f32>(lo_x) >= axes.x - 1.0f);
PLOT_CHECK(static_cast<plot::f32>(hi_x) <= axes.right() + 1.0f);
PLOT_CHECK(static_cast<plot::f32>(lo_y) >= axes.y - 1.0f);
PLOT_CHECK(static_cast<plot::f32>(hi_y) <= axes.bottom() + 1.0f);
// The grid reached the GPU: one byte per cell, once, on the frame that moved.
const plot::DensityGrid& grid = p.density_grid();
PLOT_CHECK(first.density_cells == static_cast<plot::u32>(grid.cell_count()));
PLOT_CHECK(first.density_upload_bytes == static_cast<plot::u32>(grid.cell_count()));
// The trace is *over* the density: on the row it is drawn in, the pixel is the trace's own
// colour and not that colour blended with the image behind it.
const plot::Transform tx = p.x_transform(ax);
const plot::Transform ty = p.y_transform(ay);
const plot::u32 mid_x = static_cast<plot::u32>(tx.to_screen_x(0.5));
const plot::u32 trace_row = static_cast<plot::u32>(ty.to_screen_y(0.5));
const plot::u32 ghost_row = static_cast<plot::u32>(ty.to_screen_y(0.0));
const auto frame_pixel = [&](const std::vector<plot::u8>& f, plot::u32 x, plot::u32 y,
plot::u8 out[4]) {
const plot::usize at = (plot::usize(y) * kW + x) * 4u;
for (int i = 0; i < 4; ++i) out[i] = f[at + static_cast<plot::usize>(i)];
};
plot::u8 px[4] = {};
// The trace is drawn *over* the density, and the red channel is what says so: the trace's own
// colour is 0x00E070 and carries none, while the density's tint (Style::density, a near-white)
// at full ink over it would put the red at roughly 190. So a red near zero on the rows the
// stroke covers is the ordering, and a red near 190 would be the image painted on top. The
// stroke's centre lands between two pixel centres as often as on one and its edge is
// anti-aliased, so the *smallest* red over the rows it covers is the one to read.
int min_red = 255;
for (int dy = -2; dy <= 2; ++dy) {
const int y = std::clamp(static_cast<int>(trace_row) + dy, 0, static_cast<int>(kH) - 1);
frame_pixel(with, mid_x, static_cast<plot::u32>(y), px);
min_red = std::min(min_red, static_cast<int>(px[0]));
}
PLOT_CHECK(min_red < 40);
// ... and the row *above* it holds the ghost of the sweep before: ink that was not in the
// frame drawn without a density, and ink that is not the trace's colour.
frame_pixel(without, mid_x, ghost_row, px);
const plot::u8 ghost_without[4] = {px[0], px[1], px[2], px[3]};
frame_pixel(with, mid_x, ghost_row, px);
PLOT_CHECK(px[0] != ghost_without[0] || px[1] != ghost_without[1] ||
px[2] != ghost_without[2]);
PLOT_CHECK(!pixel_is(px, trace_color));
// A frame with nothing new to deposit: the grid did not move, so the conversion and the upload
// are skipped -- the frame costs the one draw call -- and the picture is bit-identical.
r.begin_frame(kW, kH, kFrameRect);
r.draw(p);
r.end_frame();
PLOT_CHECK(!r.last_frame_had_gl_error());
PLOT_CHECK(r.stats().density_cells > 0u);
PLOT_CHECK(r.stats().density_upload_bytes == 0u);
const std::vector<plot::u8> again = read_frame(ctx);
PLOT_CHECK(again == with);
std::printf("[density gl] cells=%u uploaded=%u differing=%u ghost row=%u trace row=%u"
" axes=%.0f,%.0f %.0fx%.0f\n",
first.density_cells, first.density_upload_bytes, differing, ghost_row, trace_row,
axes.x, axes.y, axes.w, axes.h);
ctx.destroy();
}