Patch v0.8: Music analysis overhaul
This patch features a complete overhaul of the "held voices" detection algorithms, as well a revamped look for the "Walking" intensity mode!
Vocals are an extremely important part of the game for me, as it is where my mind focuses the most during a listening experience.
I wasn't satisfied with the old algorithm and I've always wanted to improve on it. So I've spent months rewriting a completely new algorithm for detecting held human voices (starting from scratch) and I'm extremely happy with the results!
My new algorithm is now far more accurate, improves detection of deeper voice pitches, and also results in much less false positives than before (where it would confuse instruments for human voices).
Changes
- The "Held Voices" detection algorithm has been entirely rewritten, and is now far more accurate
- Changed the look of the falling bubbles in Walking intensity. The arrow indicators should now be more obvious.
- Modified the "Sky Walkway" environment side visuals to focus more on audio frequencies in the voice range, as well as modifications to the color palette
- Improved music intensity sections detection
- Improved arrow visibility on obstacles indicators
- Improved the animations of the held obstacle "sliding while standing up"
- Changed the output Music Volume curve to "ease-in" at lower values (instead of being linear), in order to provide finer control of low volumes
Bug Fixes
- Fixed the Sky Walkway visuals sometimes updating in a choppy fashion
- Fixed an aspect ratio UI issue on Steam Deck
- Audio and Video latency sliders now increment in steps of 5ms, in order to be useable with a controller and on Steam Deck
Please join our Discord or the Discussions Forum to share your feedback about this new patch!