Post-processing of binaural signals
By performing frequency domain analysis and separation on the binaural signals, the level and phase difference of the head-related transfer function are estimated, solving the problem of poor binaural audio processing in existing technologies and achieving a better listener experience and personalized audio processing effects.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- DOLBY LABORATORIES LICENSING CORP
- Filing Date
- 2021-12-16
- Publication Date
- 2026-07-17
AI Technical Summary
Existing audio post-processing systems struggle to effectively utilize frequency-dependent level and time differences when processing binaural audio, resulting in a poor listening experience. Furthermore, current methods typically only rotate the overall mix or coherent elements, failing to effectively separate and process binaural rendered objects.
By transforming the binaural signals to the frequency domain, the level and phase differences of the head-related transfer function are estimated, the principal and residual components are separated, and different processing parameters are used to process them to achieve frequency-related level and time difference processing.
It enhances the listener experience of binaural audio by personalizing processing based on the location and characteristics of the object, increasing speech level, attenuating background noise, and providing better dynamic range and immersion.
Smart Images
Figure CN122420733A_ABST