Draw the wind-up from the photographed serve, not from prose

Oliver's idea, and it is the fix for the process as much as for the frames: give
the generator the real serve sequence as an IMAGE and let it copy the poses,
instead of describing a pose in prose and re-rolling until it lands. One
generation — a composite of his five photographed frames over two of our sprites
— returned a magenta row of the whole serve in the game's own character and kit.
The four stages are sliced straight out of it, anchored on the standing figure.

_fine.py gets --defringe: its two de-fringe passes each eat a one-pixel line, and
a racket frame IS a one-pixel line — that is what had been leaving the men holding
a grip and nothing else.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

render.js

@@ -155,7 +155,7 @@ export function proj(x, y, z) {
 
 // ---------------------------------------------------------------- sprites ----
 // assets/<name>.png; game falls back to code-drawn figures until sprites load.
-const SPRV = 71; // asset cache-buster — bump with main.js?v=N (see the img.src note below)
+const SPRV = 72; // asset cache-buster — bump with main.js?v=N (see the img.src note below)
 const SPR = {};
 for (const n of ['p_idle1','p_idle2','p_bounce','p_bounce2','p_toss','p_smash','p_run1','p_run2','p_run3','p_run4','p_swing1','p_swing2','p_die1','p_die2','p_die3','p_die4','p_die5',
                  'p_pull1','p_pull2','p_pull3','p_pull4', // the serve WIND-UP, one frame per quarter of the pull (see drawChar)