Spatial Audio Rendering With Object Warping Between Loudspeakers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio systems face challenges in accurately rendering audio objects across loudspeakers, particularly in configurations like Dolby Digital 5.1 and 7.1.4, leading to combing artifacts, ambiguous localization, and gain buildup, which compromise timbre and sweet spot size.
Innovation Solution
A method and system for rendering audio signals by modifying audio object positions and panning gains based on loudspeaker layout and control data, allowing objects to be positioned between or closer to loudspeakers, using techniques like clamping and warping to improve timbre and spatial accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If audio objects are panned across loudspeakers using conventional rendering methods, then spatial distribution is achieved, but combing artifacts and ambiguous localization occur
Solution Approach 1:
The patent modifies the position parameters of audio objects during rendering. Specifically, it adjusts the Y-coordinate (depth parameter) of audio objects to move them closer to the loudspeaker plane, transforming the original position metadata into modified position metadata that eliminates combing artifacts while preserving spatial intent.
Solution Approach 2:
The patent performs preliminary modification of audio object positions before the actual panning operation. By pre-adjusting positions to account for the loudspeaker configuration and listening position, it prevents combing artifacts from occurring in the first place, rather than attempting to correct them afterward.
2Measurement precision
If audio objects are positioned at intended locations, then spatial accuracy is maintained, but combing artifacts deteriorate timbre
Solution Approach 1:
The patent changes the depth parameter (Y-coordinate) of audio objects by applying an offset that moves them closer to the loudspeaker plane. This parameter modification eliminates the path length differences between loudspeakers that cause combing artifacts, while the modified positions are calculated to preserve the intended spatial accuracy in the horizontal plane.
3Area of stationary object
If audio objects are panned across front and rear loudspeakers, then spatial coverage is improved, but gain buildup increases audio level
Solution Approach 1:
By modifying the depth parameter of audio objects to position them closer to the loudspeaker plane, the patent reduces the distance between the virtual sound source and the loudspeakers. This parameter change decreases the required amplification and prevents excessive gain buildup during panning operations across front and rear loudspeakers.
4Measurement precision
If conventional rendering is used, then audio objects are reproduced at intended positions, but sweet spot size is reduced
Solution Approach 1:
The patent modifies the depth parameter (Y-coordinate) of audio objects to move them closer to the loudspeaker plane. This parameter modification creates a more favorable geometric relationship between the virtual sound sources and the loudspeaker array, which enlarges the sweet spot region where accurate spatial reproduction is achieved, while maintaining precision in the horizontal positioning.
Data Source
AI summary
Described herein is a method (30) of rendering an audio signal (17) for playback in an audio environment (27) defined by a target loudspeaker system (23), the audio signal (17) including audio data relating to an audio object and associated position data indicative of an object position. Method (30) includes the initial step (31) of receiving the audio signal (17). At step (32) loudspeaker layout data for the target loudspeaker system (23) is received. At step (33) control data is received that is indicative of a position modification to be applied to the audio object in the audio environment (27). At step (38) in response to the position data, loudspeaker layout data and control data, rendering modification data is generated. Finally, at step (39) the audio signal (17) is rendered with the rendering modification data to output the audio signal (17) with the audio object at a modified object position that is between loudspeakers within the audio environment (27).