Skip to content

v1.3.2

Choose a tag to compare

@jboneng jboneng released this 30 Aug 13:30
· 6 commits to main since this release
81c2556

[1.3.2] - 2026-08-30

Fixed

  • Scan Lab Prescan crop: a click or too-small rubber-band no longer clears a committed crop (right-click to clear). Scan builds the hardware window on the GUI thread and logs crop applied vs full window so a dropped crop cannot silently scan the safe window (including ME/IR passes).
  • Scan Lab Prescan crop X mapping: the rubber-band is in mirrored display space. Mapping now flips X back to TA so Color short / ME / IR match the box (cropping the right film base no longer cuts the left).
  • Left-edge frame clip: a fixed ENDPIXEL dummy drop was removing 16/24 px of film gate after mirror_x. Assemble now trims only columns that look like USB dummy, capped at the old width.
  • 8200i SE horizontal framing: image STR/END are shifted by 96 native clocks (optical_end_inactive_native) toward ENDPIXEL so dummy-trim + mirror_x no longer clips the film gate on the left while showing extra holder on the right. USB span / line_bytes stay the same; shading still tracks the image window.
  • 8200i SE positioning slope tables: SilverFast uploads SLOPE_TABLE_SLOW for the first (reference) feed and SLOPE_TABLE_FAST for the second (final positioning) feed — 39/39 capture pairs. pyopticfilm previously used the fast ramp for both. The V2-only use_slow_final_positioning_feed flag is no longer on Model8200iSE (8100 V2 still uses fast-then-slow).

Changed

  • 8200i SE priming default: default_gl128_prime is now False (same as the 8100 V2). Scanner.scan() with gl128_prime unset skips the discarded first-scan AGOHOME-park pass. Explicit gl128_prime=True still primes.

Contributors

  • @TobbyTravel for default priming off on the 8100 V2 and for scoping the slow-second-feed slope-table fix to V2.