Release Note
2024-2-1 : Mixed Mode for hearing aid and ble LE Audio (ASHA/MFI) . IPHONE support is still in progress
2024-1-30: automatically changing wdrc parameters ,currently attack/release time , will work on gain or knee setting
- Starry AI生成声音存在（Sound Existence）标志，取代传统VAD标志。
- StarrySoundAI Hearing Sense Enhancement Starry AI声信号前端处理算法
- WDRC 宽带动态反馈控制
- TX SRC Sample Rate Convert 输入信号的重采样，32k->16k
- ADC Gain 输入信号增益
- Analysis Filterbank 子带分析滤波器组，将信号分解成N个Band
- Synthesis Filterbank 子带合成，将信号还原成时域信号。
- Starry AI对音乐场景同样能保持极高的还原保真度
- Band Hearing Equalizer N-Band均衡滤波器组，针对助听场景生成特定的滤波器组。
- AGC-O 防破音自动增益控制，尽可能地放大声音
- RX SRC Sample Rate Convert 输出信号的重采样，16k->32k
- Frequency Shift 频移算法，<5Hz的频移防止自激振荡
- Howling Detect 啸叫检测，检测到啸叫通过改变助听器频响尽快降低影响。
- DTMF Generate 任意双频点按键音的生成
- AFC Feedback Control 声反馈处理，线性部分对回波进行处理，非线性部分交给AI。
- Play Audio Prompt (8k encoded,ulaw),saved in flash (total audio size<=200 k)
Some quick measurement data
- Audio Latency <9ms
- Working Current: 3.x mA (including DMIC microphone ,speaker ), Sleep Current: 1 uA
- Noise Reduction: great, AFC :great
- High Power Mode : gain>40db (see below diagram)
RIC Performance (Max Gain:34.10dB, Input Noise: 13.01 dbV,Tested on May/23/2023)
