- Toggling autoplay checkbox in debug now disallows saving the score
- Fix crown of the song that was previously selected rendering above the genre
- Outline of empty crowns is slightly darker
- Fix rendering when there is a crown on oni and no crown on ura
- Fix crowns overlapping the netplay 2P icon
- Fix gauge in latency calibration
- Fix debug to work on mobile, can be toggled with ctrl+alt+`;` on hacker's keyboard with permanent notification
- Fix being unable to scroll the settings list without toggling something
- Handle KeyboardInterrupt in server.py
- Fix category jumping not working in session mode
- Fix mouse cursor being hidden at results screen in session mode
- Make "Issues" on the about screen and "An error occurred, please refresh" on loading screens translateable
- CanvasCache uses integer increments for comparison instead of Date.now()
- For imported songs, exclude song titles from genre comparisons if they appear in the name of the folder
- Fix tja files with spaces in the notation
- Fix second player's score on the results screen to have upside down appearance
- Add fixed chinese strings
Adds new settings for controlling the note offset while playing. It can be either an actual offset (it is called "Audio Latency" in the settings) or just the visual offset ("Video Latency").
With higher audio latency it means you have to press the button sooner than what you hear, similarly with higher video latency it is sooner than what you see. By offsetting these events the game would play better, however, the sound effect of you hitting the drum would still play at the wrong time, the code cannot anticipate you to hit the drum in the future so to work around this issue a new option that disables drum sounds is also included.
These settings could be set through trial and error but it would be better to get the correct values through the automated latency calibration, where you can hit the drum as you hear sounds or see a blinking animation. I tried making one by measuring latency from user input, adding all the latency up, and dividing, but that gives unreliable results. I hope someone suggests to me what I should be doing during the calibration to get better results, as I cannot figure what to do on my own.
