초록 ▼

Devices, software and methods are provided for encoding voice data to conceal packet loss. In addition to regular encoding, an abbreviated portion of the frame is encoded, and transmitted redundantly afterwards. Optionally the redundant encoding is in relation to an attribute of the sound represent

Devices, software and methods are provided for encoding voice data to conceal packet loss. In addition to regular encoding, an abbreviated portion of the frame is encoded, and transmitted redundantly afterwards. Optionally the redundant encoding is in relation to an attribute of the sound represented by the data. The attribute of the sound can be pitch information, peak information, or bandwidth information.

대표청구항 ▼

The invention claimed is: 1. A transmitting device comprising: a main encoder for encoding first and second frames of voice data for transmission through a network; a redundant encoder for encoding only an abbreviated portion of the first and second frames for redundant transmission through the net

The invention claimed is: 1. A transmitting device comprising: a main encoder for encoding first and second frames of voice data for transmission through a network; a redundant encoder for encoding only an abbreviated portion of the first and second frames for redundant transmission through the network; and a transmit buffer receiving inputs from the main encoder and from the redundant encoder, for transmitting the encoded first frame in a first packet, and for transmitting both the encoded abbreviated portion of the first frame and the encoded second frame in a second packet; where the encoded abbreviated portion of the first frame is usable for generating a concealment frame to play to a user as a backup for a lost frame. 2. The device of claim 1, wherein the encoded abbreviated portion of the first frame comprises attributes of sound that the voice data represent. 3. The device of claim 1, wherein the encoded abbreviated portion of the first frame comprises a pitch period of the voice data of the first frame. 4. The device of claim 1 wherein the encoded abbreviated portion of the first frame comprises a pitch cycle waveform of the voice data of the first frame. 5. The device of claim 1, wherein the encoded abbreviated portion of the first frame comprises only a selected sample data of the first frame. 6. The device of claim 5, wherein the selected sample data is data that includes values above a preset threshold. 7. The device of claim 1, wherein the encoded abbreviated portion of the first frame comprises only data with values within a passed band of a sound bandwidth. 8. A receiving device comprising: a network interface for coupling to a network; and a processor coupled with the network interface, wherein the processor is adapted to wait to receive an encoded frame of voice data and a coupled frame of voice data from the network, the coupled frame including an encoded redundant frame and an encoded second frame, and if the encoded frame is not received, to expand-decode the encoded redundant frame; wherein the encoded redundant frame comprises an abbreviated portion of the encoded frame, which is used to generate a concealment frame to play to a user as a backup for a lost frame the encoded frame that was not received. 9. The device of claim 8, wherein the encoded redundant frame is expand-decoded to generate attributes of sound that represents the data of the encoded frame. 10. The device of claim 8, wherein the encoded redundant frame is expand-decoded to generate an extracted pitch period. 11. The device of claim 8, wherein the encoded redundant frame is expand-decoded by adding filler data. 12. The device of claim 8, wherein the processor is further adapted to: synthesize a complementary band to a passed band, and combine the complementary band with the passed band. 13. An article comprising: a storage medium, said storage medium having stored thereon instructions, that, when executed by at least one device, result in: encoding a frame of voice data to form an encoded frame; transmitting the encoded frame through a network; encoding an abbreviated portion of the frame of voice data to form an encoded redundant frame encoding a second frame of voice data to form an encoded second frame; coupling the encoded redundant frame with the encoded second frame to form a coupled frame; the encoded second frame being based on a different frame of voice data than the encoded redundant frame; transmitting the coupled frame through the network. 14. The article of claim 13, wherein the instructions further result in: coupling the encoded frame with a second encoded redundant frame prior to transmitting. 15. The article of claim 13, wherein encoding the abbreviated portion of the frame of voice data to be performed by predictive coding. 16. The article of claim 13, wherein encoding the abbreviated portion of the frame of voice data to be performed by differential coding. 17. The article of claim 13, wherein the instructions further result in: encoding the abbreviated portion of the frame of voice data to be performed in relation to attributes of sound that the data represent. 18. An article comprising: a storage medium, said storage medium having stored thereon instructions, that, when executed by at least one device, result in: waiting to receive an encoded frame of voice data from the network; waiting to receive a coupled frame of voice data from the network, the coupled frame including an encoded redundant frame and an encoded second frame; and if the encoded frame is not received, expand-decoding the encoded redundant frame; wherein the encoded redundant frame comprises an abbreviated portion of the encoded frame which is used to generate a concealment frame to play to a user as a backup for a lost frame the encoded frame that was not received. 19. The article of claim 18, wherein expand-decoding is performed to generate attributes of sound that represents data of the encoded frame. 20. The article of claim 18, wherein expand-decoding includes using an extracted pitch period and an extracted pitch cycle waveform to construct the concealment frame. 21. The article of claim 18, wherein expand-decoding includes adding filler data to the encoded redundant frame. 22. The article of claim 18, wherein the instructions further result in: synthesizing a complementary band to a passed band; and combining the complementary band with the passed band. 23. A method comprising: encoding a frame of voice data to form an encoded frame; transmitting the encoded frame through a network; encoding an abbreviated portion of the frame of voice data to form an encoded redundant frame; encoding a second frame of voice data to form an encoded second frame; coupling the encoded redundant frame with the encoded second frame to form a coupled frame; the encoded second frame being based on a different frame of voice data than the encoded redundant frame; and transmitting the coupled frame through the network. 24. The method of claim 23, further comprising: coupling the encoded frame with a second encoded redundant frame prior to transmitting. 25. The method of claim 23, wherein encoding the abbreviated portion of the frame of voice data to be performed by predictive coding. 26. The method of claim 23, wherein encoding the abbreviated portion of the frame of voice data to be performed by differential coding. 27. The method of claim 23, wherein the abbreviated portion of the frame of voice data comprises attributes of sound that the frame represents. 28. The method of claim 23, wherein encoding the abbreviated portion of the frame of voice data includes performing pitch detection to extract a pitch period and a pitch cycle waveform. 29. The method of claim 28, wherein performing pitch detection includes: selecting a plurality of pitch candidates, determining an autocorrelation pattern for each of the pitch candidates, determining which one of the autocorrelation patterns is optimum, and selecting the pitch candidate with the optimum autocorrelation pattern. 30. The method of claim 23, wherein the encoded redundant frame contains less data than the encoded frame. 31. The method of claim 30, further comprising: selecting a peak threshold value; encoding data with values larger than the peak threshold value; and setting data with values less than the peak threshold value equal to a preset value. 32. The method of claim 31, wherein encoding the abbreviated portion of the frame of voice data includes run length encoding the data with the preset value. 33. The method of claim 31, wherein the preset value is zero. 34. The method of claim 23, wherein encoding the abbreviated portion of the frame of voice data includes selecting and encoding only data with values within a passed band of a sound bandwidth. 35. The method of claim 34, wherein the passed band is a low-frequency band. 36. The method of claim 35, further comprising: down-sampling the passed band prior to encoding it. 37. A method comprising: waiting to receive an encoded frame of voice data and a coupled frame of voice data from a network, the coupled frame including an encoded redundant frame and an encoded second frame; and if the encoded frame is not received, expand-decoding the encoded redundant frame; wherein the encoded redundant frame comprises an abbreviated portion of data of the encoded frame, which is used to generate a concealment frame to play to a user as a backup for the encoded frame that was not received. 38. The method of claim 37, wherein expand-decoding is performed to generate attributes of sound that represents the encoded frame. 39. The method of claim 37, wherein expand-decoding includes using an extracted pitch period and an extracted pitch cycle waveform to construct the concealment frame. 40. The method of claim 39, wherein expand-decoding is performed by repeating the pitch cycle waveform. 41. The method of claim 39, wherein expand-decoding is performed by: constructing a weighted average waveform from the pitch cycle waveform; and repeating the weighted average waveform. 42. The method of claim 41, further comprising: extracting a second pitch cycle waveform from data of at least one other frame, and wherein the weighted average waveform is also constructed from the second pitch-cycle waveform. 43. The method of claim 37, wherein expand-decoding includes adding filler data. 44. The method of claim 43, further comprising: determining optimized filler data. 45. The method of claim 37, wherein expand-decoding includes: synthesizing a complementary band to a passed band; and combining the complementary band with the passed band. 46. The method of claim 45, wherein synthesizing the complementary band is performed by analyzing data from other received frames. 47. The method of claim 46, wherein synthesizing is performed by: determining a complementary band information synthesis shift from the passed band; and using the synthesis shift to derive the complementary band from the passed band. 48. The method of claim 47, wherein the synthesis shift is determined by: cross-correlating the frame of the passed band with data from the other received frames. 49. The method of claim 47, wherein the synthesis shift is determined by: analyzing the encoded frame prior to transmitting it. 50. A receiving device comprising: coupling means for coupling to a network; and processing means adapted to wait to receive an encoded frame of voice data and a coupled frame of voice data from the network, the coupled frame including an encoded redundant frame and an encoded second frame, and if the encoded frame is not received, to expand-decode the encoded redundant frame; wherein the encoded redundant frame comprises an abbreviated portion of the encoded frame, which is used to generate a concealment frame to play to a user as a backup for a lost frame the encoded frame that was not received. 51. The device of claim 50, wherein the encoded redundant frame is expand-decoded to generate attributes of sound that represents the encoded frame. 52. The device of claim 50, wherein the encoded redundant frame is expand-decoded by using an extracted pitch period and an extracted pitch cycle waveform to construct the concealment frame. 53. The device of claim 50, wherein the encoded redundant frame is expand-decoded by adding filler data. 54. The device of claim 50, wherein the processor is further adapted to: synthesize a complementary band to a passed band, and combine the complementary band with the passed band. 55. The method of claim 37 further comprising playing the concealment frame to a user in an attempt to conceal to the user the fact that the encoded frame was not received.