release 2.0.0

[stas-sl]

Release notes

This release will break existing configs that’s why I’m increasing its major version to 2 (not because it includes many new exciting features). If you’re happy with your current config and functionality and don’t want to bother with updating your config, you can pin previous version explicitly and stay with it, until better time:

external_components:
  - source: github://stas-sl/esphome-sound-level-meter@v1.0.4

Changes in this release:

• New config structure. I decided to change the hierarchical YAML configuration structure for filters and sensors to a single-level flat structure. I often found it difficult to maintain the correct indentation for nested groups, so I hope this new structure will be more readable, easier to modify, and less error-prone. This change does not affect flexibility or functionality; you can accomplish the same tasks with the new structure as you could with the previous hierarchical one.
• 10 bands realtime spectrum analyser demo/config example.
• Optimised buffer memory management, which reduces memory (re)allocations and improves performance
• Ability to switch to highly optimized dsps_biquad_f32_xx functions from ESP-DSP library for SOS filter processing on ESP32 and ESP32S3. It could be 2 times faster than current C++ implementation, however there are some drawbacks: 1) only ESP32 and ESP32S3 have optimized implementations 2) it uses direct form 2 filter implementation which is considered less numerically stable than direct form 2 transposed (which is used by me), though I’m not sure if this is a big deal 3) it is not available for Arduino framework, only works with esp-idf.