Differences

This shows you the differences between two versions of the page.

--- fourier:phase5_2026-04-30 [2026/04/30 10:14] – Phase 5 review handover: revised Phase 1 metric, baseline reproduction on stock danctnix, plan for chromium-fourier on ohm markus_fritsche
+++ fourier:phase5_2026-04-30 [2026/04/30 10:19] (current) – [Replacement — variant (C)] markus_fritsche
@@ Line 30: / Line 30: @@
 Two bands measured on every run:
-  * **Coarse band** (ship-gate). Per-clip realtime run, fixed length,
+  * **Coarse band** (ship-gate). Per-clip realtime run, fixed length, fixed clip, fixed framerate target. Capture: time-averaged CPU% over the run, decoder + sink drops out of total presented frames, and a path-discipline note (which surface negotiated; dmabuf-direct or GL-imported; what was actually established at both ends). "Hard facts > somewhat hard facts" — average CPU% is the hard fact, not a peak-watching estimate.
-    fixed clip, fixed framerate target. Capture: time-averaged CPU%
+  * **Latency band** (diagnosis). Per-frame distribution over the chain compressed-bytes-arrival → presented-pixel-on-display, with sub-bands queue-fill / decode / fence-publish / dmabuf-export / GL-import (when applicable) / composite / scanout. Used when (B) fails and we don't know which sub-band ate the latency.
-    over the run, decoder + sink drops out of total presented frames,
-    and a path-discipline note (which surface negotiated; dmabuf-direct
-    or GL-imported; what was actually established at both ends).
-    "Hard facts > somewhat hard facts" — average CPU% is the hard
-    fact, not a peak-watching estimate.
-  * **Latency band** (diagnosis). Per-frame distribution over the
-    chain compressed-bytes-arrival → presented-pixel-on-display, with
-    sub-bands queue-fill / decode / fence-publish / dmabuf-export /
-    GL-import (when applicable) / composite / scanout. Used when (B)
-    fails and we don't know which sub-band ate the latency.
 ==== Surface-variant clarification ====