"Subband Center Signal Scaling using Power Ratios"
C. Uhle and E. Habets
submitted to AES 53rd Conference on Semantic Audio, 2014


Abstract
--------

A method for scaling the level of the virtual center in audio signals is proposed. The input signals are processed in the time-frequency domain by applying  time and frequency dependent real-valued spectral weights that implicitly depend on the panning factors and the diffuseness of the input signal. Different methods to compute the spectral weights are presented and evaluated. In all cases, the spectral weights depend on the ratio of the power of a linear combination of the input signals and the power of a passive downmix signal. Applications of the presented method are upmixing, stereophonic enhancement, dialogue enhancement, and pre-processing for semantic audio analysis.


Sound examples
--------------

Section 5.1.:
- LPST3*.wav

Section 5.2.:
*_orig: input signals 
*_TSDR_1_beta1: Eq (16) with beta=1
*_TSDR_1_beta2: Eq (16) with beta=2 
*_SSDR_1: Eq (17) 
*_TSDR_2_beta1: Eq (18) with beta=1
*_TSDR_2_beta2: Eq (18) with beta=2
*_SSDR_2: Eq (19)
*_TSDR_alternative_1: alternative mapping of TSDR for off-center attenuation
*_TSDR_alternative_2: alternative mapping of TSDR for center attenuation

 