Skip to content
Downloads← Current beta hub

videomancer beta program

Archived RC: 1.0.0-rc.39

Dual-mode RGB 1V accuracy — EXT sync delay + encoder CSC; verify dual loopback bars at 720p60/1080p30 and no YPbPr AC / HDMI regression

Before you test

  • Install LZX Connect and enable Show pre-release firmware
  • Flash 1.0.0-rc.39 (Connect update or .uf2 from Downloads)
  • Confirm 1.0.0-rc.39 on the SYSTEM screen
  • Dual-capable cable set: HDMI in + Analog Out looped to Analog In (YPbPr AC and RGB 1V DC jacks as needed)

Testing goals

  • vmtest regress — sim rcguard + HIL acceptance

    high

    1. Connect desk HDMI loop + USB shell; run vmtest preflight (READY). 2. From repo root: PYTHONPATH=platform python -m kernel.lzx_test.vmtest regress. 3. Optional repeatability: hil run --acceptance --no-interlaced --repeat 10.

    Pass: vmtest regress exits 0; no new_fail vs baseline; sim rcguard green.

  • Dual RGB 1V — color bars levels and primaries

    high

    1. Routing = Dual; Analog Out = RGB 1V; Analog In = RGB 1V. 2. Load passthru; feed HDMI color bars (or Connect pattern). 3. Check 720p60 and 1080p30 — bars should show correct greys and primaries (no crushed pedestal, blown white, or strong green bleed into blue/red).

    Pass: Dual RGB 1V picture is usable at both timings; bar greys and primaries look correct vs HDMI direct; no sustained unlock thrash.

  • Dual YPbPr AC — control path unchanged

    high

    1. Same Dual routing; Analog Out/In = YPbPr (AC jacks). 2. Passthru bars at 720p60 (and 1080p30 if time).

    Pass: YPbPr AC dual loopback still locks and displays without obvious level or color regression vs rc.38.

  • HDMI + standalone smoke

    medium

    1. Routing = HDMI; passthru bars — lock and picture OK. 2. Routing = Standalone; colorbars (or colorbars RGB) at one progressive rate — picture present.

    Pass: No obvious blank or unlock regression on HDMI or standalone vs prior RC.

  • rc.38 YPbPr DVD / Y-replug smoke

    medium

    1. If a DVD/SD YPbPr source is available, spot-check NTSC classify and Y unplug/replug from the rc.38 goals.

    Pass: Timing stays NTSC when locked; Y replug recovers without sustained thrash (no regression from rc.38).

Report feedback