Output
π Pit-stop tire wear calls are clearer
- Delayed tire-wear calls are easier to understand and less likely to sound like your current wear.
- Fixed: Older pit-stop wear alerts are skipped once you have already made a newer stop.
β Robot Voice – Neural voice replies could switch voices or fail during busy moments
- Spoken replies are more likely to play in the right voice when several voice events happen close together.
β Short voice prompts could repeat or fail after bad cached audio
- Quick prompts and other spoken replies are less likely to repeat, break, or regenerate unnecessarily after a bad cached clip.
β Neural lap-time digits could switch to a Local voice
- Short numeric replies like lap times and gaps are less likely to fall back to a different voice just because the neural clip ran slightly long.
β Green-flag call got skipped after one lap to green
- DRE is less likely to miss the main green-flag call at race start.
Neural Recognition
β Neural commands could get rejected even when they matched
- Commands are less likely to be ignored after DRE already understood you.
β DRE could drop out after a local Neural Recognition crash
- If local Neural Recognition fails, DRE keeps listening on the safer cloud fallback for the rest of that session.
Simulators
β LMU online sessions could create ghost cars
- Nearby and position calls are steadier because LMU is less likely to duplicate opponents online.
β LMU startup calls could open with placeholder session info
- Session intro calls are less likely to open with missing LMU location details.
β LMU weather and tire-change calls could jump to the wrong warning
- Dry LMU sessions are less likely to trigger stray rain warnings.
- Nearby opponent tire-change calls are less likely to fire from garage or setup-side flickers.
β LMU car and pit data could break on incomplete simulator payloads
- DRE is more resilient when LMU sends incomplete session or pit data.
β ACC broadcasting setup could fail when your Documents folder was unavailable
- ACC is less likely to fail setup when Windows cannot resolve or write the expected broadcasting config path.
Fuel
β Auto fuel could overfill staged pit stops
- If DRE planned a smaller stop inside a multi-stop strategy, it is less likely to ask for too much fuel.
β Fuel Analytics could break on invalid fuel-unit values
- Fuel graphs are less likely to drop out when incoming fuel conversions produce invalid numbers.
Stability
β DRE startup could fail when optional speech files were blocked
- DRE is less likely to fail at startup when optional speech files are locked or blocked by Windows.
β Fuel Analytics crashed on rare redraws
- Fuel graphs are less likely to crash while live data keeps updating.
β Problem controllers could interrupt controller detection
- DRE is less likely to stop detecting your other controllers because of one problem device.
β Command setup and editing could break on missing or refreshing data
- DRE is less likely to crash or reset command setup when startup data is incomplete or search refreshes while you are editing.
β Bug reports after midnight missed the earlier race log
- If your race started before midnight and you generated a bug report after midnight, DRE includes both daily logs in the report.
π Known Issues
- None π Found a bug?