Audio AGC Inference Using Inaudible Reference Signals
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Modern mobile devices often apply automatic gain compensation (AGC) to audio recordings system-wide, making it difficult for applications to control or undo AGC effects, which is undesirable in music recording where preserving dynamic range is crucial.
Innovation Solution
A device emits inaudible reference audio signals during recording, which are processed to infer and undo the AGC effects applied to the audio content, allowing for the restoration of the original dynamic range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If system-level AGC is applied to all audio recordings, then weak signals are enhanced and strong signals are reduced, but the original dynamic range and intended dynamics of music recordings are lost
Solution Approach 1:
The patent introduces an intermediary AGC inference module that analyzes recorded signals to extract AGC gain information without directly interfering with the audio signal path. This mediator reads the AGC effects indirectly through signal analysis and provides the gain data to consuming applications, allowing them to undo AGC effects while the system-level AGC remains active for general purposes.
Solution Approach 2:
The system implements feedback by having the AGC inference module continuously analyze recorded signals, extract AGC gain information, and provide this feedback to consuming applications. This feedback loop enables applications to adjust their processing based on the actual AGC effects applied, allowing them to compensate and restore the original dynamic range.
2Extent of automation
If system-level AGC is enabled, then audio signal levels are automatically adjusted, but applications lack control and cannot undo AGC effects
Solution Approach 1:
The AGC inference module serves as an intermediary that bridges the gap between the automated system-level AGC and applications needing control. It extracts AGC gain information from recorded signals and makes this information available to consuming applications through an interface, enabling applications to undo AGC effects while the automated AGC system continues to operate.
Solution Approach 2:
The system implements self-service by having the AGC inference module automatically analyze recorded signals, extract AGC gain information, and provide this information to applications without requiring manual intervention. Applications can then use this information to automatically compensate for AGC effects, making the system self-regulating.
3Device complexity
If AGC gains are not provided to applications, then system complexity is reduced, but applications cannot adjust or undo AGC effects
Solution Approach 1:
The AGC inference module acts as an intermediary that provides AGC gain information to applications without requiring complex system-level changes. It analyzes recorded signals to extract AGC effects and provides this information through a relatively simple interface, enabling applications to adjust their processing while adding minimal complexity to the overall system.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Methods, systems, and computer program products that infer and correct automatic gain compensation (AGC) values over time are described. A device emits a series of inaudible reference audio signals during recording. The reference audio signals have a constant amplitude. A microphone of the device records the reference audio signals while recording audio content. The device may apply AGC during the recording. An AGC inference module receives the recorded signals and extracts a representation of the reference audio signals, which have been subject to the AGC. The AGC inference module determines variations in levels of the extracted representation of the reference audio signals over time. Based on the variations, the AGC inference module infers the AGC applied to the recording over time. The AGC inference module can then provide the AGC for reading, or undo effects of the AGC applied to the audio content.