Audio limiter system and method
A system and method for limiting amplitude of a signal, such as limiting the loudness of an audio signal. Embodiments of the present invention allow for the most aggressive limiting by using an advanced psychoacoustic model to intelligently determine the amount of limiting that can be done to the incoming signal before producing distortion that is detectable to the human ear.
1. A method performed by a processor for limiting an audio signal, comprising:
applying a psychoacoustic model to a portion of the audio signal, to produce masking thresholds for a plurality of critical frequency bands;
generating a plurality of different amplitude envelopes for the portion of the audio signal;
for each of the plurality of amplitude envelopes:
applying the amplitude envelope to the audio signal portion to produce a limited audio signal portion;
producing a distortion signal by comparing the audio signal portion to the limited audio signal portion;
calculating distortion signal power for the plurality of critical frequency bands; and
comparing the plurality of distortion signal power for critical frequency bands against the plurality of masking thresholds to produce a distortion score; and
selecting one of the limited audio signal portions based its corresponding distortion score.
2. The method of claim 1 wherein each of the plurality of limited audio signal portions are normalized before producing a distortion signal.
3. The method of claim 1 wherein the step of selecting one of the limited audio signal portions includes selecting one of the limited audio signal portions whose corresponding distortion score is below a pre-selected threshold value.
4. The method of claim 1 wherein the audio signal is limited to not exceed a predetermined amplitude threshold that does not change.
5. The method of claim 1 wherein the critical frequency bands cover a full range of audible frequencies.
6. The method of claim 1 wherein the psychoacoustic model is configured to raise masking levels around transients in the audio signal.
7. The method of claim 1 wherein the psychoacoustic model calculates and utilizes a tonality index for the audio signal.
8. A non-transitory computer-readable storage medium having instructions, that when executed by a processor, cause the processor to perform a method comprising:
applying a psychoacoustic model to a portion of an audio signal, to produce masking thresholds for a plurality of critical frequency bands;
generating a plurality of different amplitude envelopes;
for each of the plurality of amplitude envelopes:
applying the amplitude envelope to the audio signal portion to produce a limited audio signal portion;
producing a distortion signal by comparing the audio signal portion to the limited audio signal portion;
calculating distortion signal power for the plurality of critical frequency bands; and
comparing the plurality of distortion signal power for critical frequency bands against the plurality of masking thresholds to produce a distortion score; and
selecting one of the limited audio signal portions based its corresponding distortion score.
9. The non-transitory computer-readable storage medium of claim 8 further including instructions wherein each of the plurality of limited audio signal portions are normalized before producing a distortion signal.
10. The non-transitory computer-readable storage medium of claim 8 further including instructions wherein the step of selecting one of the limited audio signal portions includes selecting one of the limited audio signal portions whose corresponding distortion score is below a pre-selected threshold value.
11. The non-transitory computer-readable storage medium of claim 8 , wherein the audio signal is limited to not exceed a predetermined amplitude threshold that does not change.
12. The non-transitory computer-readable storage medium of claim 8 , further including instructions wherein the critical frequency bands cover a full range of audible frequencies.
13. The non-transitory computer-readable storage medium of claim 8 , wherein the psychoacoustic model is configured to raise masking levels around transients in the audio signal.
14. The non-transitory computer-readable storage medium of claim 8 , wherein the psychoacoustic model calculates and utilizes a tonality index for the audio signal.
15. An audio limiter, comprising:
a psychoacoustic model component, configured to analyze a portion of an audio signal, and produce masking thresholds for a plurality of critical frequency bands;
an envelope generating component, configured to generate a plurality of different amplitude envelopes;
a distortion calculating component, configured to:
apply each of the different amplitude envelopes to the portion of audio signal to produce a limited audio signal,
create a distortion signal by comparing the limited audio signal to portion of the audio signal,
calculate distortion signal power for a plurality of critical frequency bands for the distortion signal, and
calculate a distortion score by comparing the distortion power to the masking thresholds for the plurality of critical bands; and
a selection component, configured to receive the distortion scores for the plurality of limited audio signals, and to select one of the limited audio signals based on its distortion score.
16. The audio limiter of claim 15 , wherein the psychoacoustic model is configured to raise masking levels around transients in the audio signal.