Audio stream mixing with dialog level normalization

A method for mixing two input audio signals into a single, mixed audio signal while maintaining a perceived sound level of the mixed audio signal, the method comprising:

• receiving a main input audio signal;
• receiving an associated input audio signal;
• wherein the associated input audio signal is coupled with the main input audio signal;
• receiving mixing metadata, which contains scaling information for scaling the main input audio signal and which specifies how the main input audio signal and the associated input audio signal should be mixed, in order to generate a mixed audio signal at the perceived sound level;
• wherein the scaling information from the mixing metadata comprises a metadata scale factor for the main input audio signal, for scaling the main input audio signal relative to the associated input audio signal;
• receiving a mixing balance input, which denotes an adjustable balance between the main input audio signal and the associated input audio signal, wherein the mixing balance input comprises scaling information which allows a deviation from a weighting of the main input audio signal and the associated input audio signal in the mixed audio signal as specified in the mixing metadata;
• identifying a dominant signal as either the main input audio signal or the associated input audio signal from the scaling information provided by the mixing metadata and from the mixing balance input, wherein the respective other input audio signal is then identified as a non-dominant signal;
• and wherein the dominant signal is identified by comparing the mixing balance input with the metadata scale factor for the main input audio signal;
• scaling the non-dominant signal in relation to the dominant signal;
• and
• combining the scaled non-dominant signal with the dominant signal to yield the mixed audio signal.

A device for mixing input audio signals into a single, mixed audio signal while maintaining a perceived sound level of the mixed audio signal, comprising:

• a receiver for receiving a main input audio signal, an associated input audio signal and mixing metadata;
• wherein the associated input audio signal is coupled with the main input audio signal;
• wherein the mixing metadata contains scaling information for scaling the main input audio signal;
• wherein the scaling information specifies how the main input audio signal and the associated input audio signal should be mixed, in order to generate a mixed audio signal at the perceived sound level;
• wherein the scaling information from the mixing metadata comprises a metadata scale factor for the main input audio signal, for scaling the main input audio signal relative to the associated input audio signal;
• a mixing balance input, which denotes the adjustable balance between the main and the associated input audio signals;
• wherein the mixing balance input comprises scaling information which allows a deviation from a weighting of the main input audio signal and the associated input audio signal in the mixed audio signal as specified in the mixing metadata;
• and
• a mixing unit configured to identify a dominant signal as either the main input audio signal or the associated input audio signal from the scaling information provided by the mixing metadata and from the mixing balance input, wherein the respective other input audio signal is then identified as a non-dominant signal;
• and wherein the dominant signal is identified by comparing the mixing balance input with the metadata scale factor for the main input audio signal;
• wherein the mixing unit is further configured to scale the non-dominant signal in relation to the dominant signal;
• and to combine the scaled non-dominant signal with the dominant signal to yield the mixed audio signal.