Refuse centering measurement on die-cut/clear-window cards
Same failure family as the edge-whitening fix: centering_profile assumed a uniform printed border, so on an E-X/SPx-style die-cut the side midpoint samples landed on clear acetate and the border-width walk measured two different materials, producing a confident 70/30 the model then treated as authoritative. Now the four side samples must read as the same material or the measurement refuses with the reason; the prompt tells the model to judge die-cut centering as design-to-cut registration instead.
This commit is contained in:
parent
bb5d50502c
commit
fa5467e55d
3 changed files with 81 additions and 6 deletions
|
|
@ -137,8 +137,10 @@ function renderGradeBlock(g, opts = {}) {
|
|||
}
|
||||
|
||||
const cm = g.centering_measurement || null;
|
||||
const centeringLine = cm
|
||||
? `measured — left/right ${cm.horizontal_label} · top/bottom ${cm.vertical_label}` : '';
|
||||
const centeringLine = cm && cm.reliable === false
|
||||
? `not measurable — ${cm.reason || "this card's cut"}`
|
||||
: (cm && cm.horizontal_label
|
||||
? `measured — left/right ${cm.horizontal_label} · top/bottom ${cm.vertical_label}` : '');
|
||||
|
||||
// Only worth surfacing when it's away from noise — most cards sit within
|
||||
// a percent or two of standard and repeating that number on every card
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue