Skip to content
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
15 changes: 12 additions & 3 deletions docs/tickets.md
Original file line number Diff line number Diff line change
Expand Up @@ -1776,9 +1776,18 @@ Two more things learned while verifying:
single instant mixes tier-warmup states — trust 60 s averages only.
- An apparent "HW loses the sun shadows" difference between screenshot pairs
did NOT survive scrutiny: the drifting cloud deck (45% shadow strength)
confounds any two captures taken minutes apart. A real GI-quality A/B needs
clouds off (`setCloudShadows(0,...)`) and same-frame capture. UNRESOLVED —
do this before ever defaulting HW on.
confounds any two captures taken minutes apart.
- **The clouds-off A/B was then done (same evening,
`setCloudShadows(0,...)` probe, 30 s settle each, gate prints verified):**
both paths render correct sun shadows; the difference is TONAL — HW GI is
mildly warmer/brighter in the indirect term (most visible on the house
facade picking up warm ground bounce; SW reads slightly greyer). At the
outdoor vantage it is a subtle grade-level difference, not a structural
quality win, for ~30% of the frame (14.4 vs 20.3 fps). Verdict: SW stays
the default; opt in per box with `BLOOM_HW_GI=1` where the GPU affords it.
Captures archived in the shooter's `tools/.testout/gi-ab-{sw,hw}.png`.
Re-open the default question for INDOOR scenes (bounce matters more there;
the menu backdrop cannot exercise INDOORCAM) or on stronger hardware.

## HW-vs-SW Lumen: MEASURED at last (2026-07-16 evening) — HW costs +20 ms/frame on the 760M

Expand Down
Loading