Time-domain sound field reproduction using the group Lasso

Area of Science:

• Acoustics
• Signal Processing
• Optimization

Background:

• Accurate sound field reproduction (SFR) is crucial for immersive audio experiences.
• Traditional methods often require numerous loudspeakers or struggle with under-sampled conditions.

Purpose of the Study:

• To develop a novel time-domain SFR method using group Lasso (GL).
• To optimize the number and placement of active loudspeakers for efficient broadband SFR.
• To compare the proposed GL method against least-squares approaches.

Main Methods:

• A time-domain sound field reproduction method based on group Lasso (GL) was developed.
• An iterative mixed-norm constrained optimization problem was solved to determine active loudspeaker positions and count.
• Simulations were conducted to evaluate the method's performance.

Main Results:

• The proposed GL method achieved accurate broadband SFR with a minimal number of active loudspeakers.
• The GL method significantly outperformed the least-squares method in under-sampled sound field scenarios.
• The GL algorithm demonstrated effectiveness in optimizing loudspeaker placement for sound field control.

Conclusions:

• The group Lasso-based method offers an efficient and accurate solution for time-domain sound field reproduction.
• This approach is particularly advantageous when dealing with under-sampled desired sound fields.
• The GL algorithm provides a versatile tool for optimizing loudspeaker configurations in acoustic systems.