Method and a decoder for attenuation of signal regions reconstructed with low accuracy
The embodiments of the present invention improves conventional attenuation schemes by replacing constant attenuation with an adaptive attenuation scheme that allows more aggressive attenuation, without introducing audible change of signal frequency characteristics.
1. A method for a decoder for determining an attenuation to be applied to an audio signal, the method comprising:
identifying spectral regions of the audio signal to be attenuated;
grouping subsequent identified spectral regions to form a continuous spectral region;
determining a width of the continuous spectral region; and
applying an attenuation of the continuous spectral region adaptive to the width such that an increased width decreases the attenuation of the continuous spectral region.
2. The method according to claim 1 wherein identifying spectral regions to be attenuated comprises identifying the spectral regions coded with either a low number of bits or with no bits assigned.
3. The method according to claim 2 wherein identifying spectral regions to be attenuated comprises examining reconstructed subvectors to identify the spectral regions to be attenuated.
4. The method according to claim 2 wherein examining the reconstructed subvectors comprises examining the number of bits assigned to the reconstructed subvectors to determine whether the number of assigned bits falls below a predetermined threshold; and wherein a spectral region has low precision when the number of bits assigned to the corresponding reconstructed subvector falls below the predetermined threshold.
5. The method according to claim 3 further comprising encoding the subvectors with a pulse coding scheme; wherein the corresponding spectral region has low precision when comprising one or more consecutive subvectors where the number of pulses P(b) falls below a predetermined threshold.
6. The method according to claim 1 wherein identifying spectral regions to be attenuated comprises identifying the spectral regions coded with no bits assigned.
7. The method according to claim 1 wherein identifying spectral regions to be attenuated comprises identifying the spectral regions coded with a low number of bits.
8. The method according to claim 1 where the continuous spectral region further includes a region reconstructed using a bandwidth extension algorithm.
9. The method according to claim 1 wherein identifying spectral regions to be attenuated comprises identifying the spectral regions to be attenuated based on an analysis received from an encoder: wherein the analysis identifies potential candidate spectral regions for attenuation based on whether a distance measure between a reconstructed synthesis signal and an input target signal in a frequency region is above a threshold.
10. An attenuation controller of a decoder for determining an attenuation to be applied to an audio signal, the attenuation controller comprising:
an identifier unit configured to identify spectral regions to be attenuated;
a grouping unit configured to group subsequent identified spectral regions to form a continuous spectral region;
a determination unit configured to determine a width of the continuous spectral region; and
an application unit configured to apply an attenuation of the continuous spectral region adaptive to the width such that an increased width decreases the attenuation of the continuous spectral region.
11. The attenuation controller according to claim 10 wherein the identifier unit identifies the spectral regions to be attenuated by identifying the spectral regions coded with either a low number of bits or with no bits assigned.
12. The attenuation controller according to claim 11 wherein the identifier unit is further configured to examine reconstructed subvectors.
13. The attenuation controller according to claim 12 wherein a spectral region has low precision when the number of bits assigned to the corresponding reconstructed subvector falls below a predetermined threshold.
14. The attenuation controller according to claim 12 wherein a pulse coding scheme is employed to encode the subvectors; and wherein the corresponding spectral region has low precision when comprising one or more consecutive subvectors where the number of pulses P(b) falls below a predetermined threshold.
15. The attenuation controller according to claim 10 wherein the identifier unit is configured to identify the spectral regions coded with no bits assigned.
16. The attenuation controller according to claim 10 wherein the identifier unit is configured to identify the spectral regions coded with a low number of bits.
17. The attenuation controller according to claim 10 where the continuous spectral region further includes a region reconstructed using a bandwidth extension algorithm.
18. The attenuation controller according to claim 10 :
further comprising an input unit configured to receive an analysis from an encoder;
wherein the identifier unit is further configured to identify the spectral regions to be attenuated based on the received analysis; and
wherein the analysis identifies potential candidate spectral regions for attenuation based on whether a distance measure between a reconstructed synthesis signal and an input target signal in frequency region is above a threshold.
19. A mobile terminal comprising:
an attenuation controller of a decoder for determining an attenuation to be applied to an audio signal, wherein the attenuation controller comprises:
an identifier unit configured to identify spectral regions to be attenuated;
a grouping unit configured to group subsequent identified spectral regions to form a continuous spectral region;
a determination unit configured to determine a width of the continuous spectral region; and
an application unit configured to apply an attenuation of the continuous spectral region adaptive to the width such that an increased width decreases the attenuation of the continuous spectral region.
20. A network node comprising:
an attenuation controller of a decoder for determining an attenuation to be applied to an audio signal, wherein the attenuation controller comprises:
an identifier unit configured to identify spectral regions to be attenuated;
a grouping unit configured to group subsequent identified spectral regions to form a continuous spectral region;
a determination unit configured to determine a width of the continuous spectral region; and
an application unit configured to apply an attenuation of the continuous spectral region adaptive to the width such that an increased width decreases the attenuation of the continuous spectral region.