Audio encoding device, method, and program which controls the number of time groups in a frame using three successive time group energies

An audio encoding method of conducting grouping of a subband signal obtained by frequency-converting an input audio signal in a time direction and a frequency direction for each frame unit, and encoding the signal for each unit of the grouping, characterized in:

• calculating a candidate for a group boundary in the time direction;
• generating a maximum value of a group number in the time direction;
• employing said candidate, thereby to conduct the grouping in the time direction so that a space of the boundary of the group being generated in the time direction becomes equal to or more than a constant and yet the group number in the time direction becomes equal to or less than said maximum value, and to generate time direction group information;
• and conducting the grouping in the frequency direction by employing said maximum value and said time direction group information,
• wherein said maximum value of said group number is made variable according to coding quantity information,
• wherein a maximum time group quantity is determined in response to a buffer residual quantity, where the coding quantity information is the buffer residual quantity;
• wherein the candidate for the group boundary in time direction is selected by comparing change in energy information of three successive time groups.

A non-transitory computer-readable data storage medium storing an audio encoding program for conducting grouping of a subband signal obtained by frequency-converting an input audio signal in a time direction and a frequency direction for each frame unit, and encoding the signal for each unit of the grouping, characterized in causing an information processing device to execute the processes of:

• calculating a candidate for a group boundary in the time direction;
• generating a maximum value of a group number in the time direction;
• employing said candidate, thereby to conduct the grouping in the time direction so that a space of the boundary of the group being generated in the time direction becomes equal to or more than a constant and yet the group number in the time direction becomes equal to or less than said maximum value, and to generate time direction group information;
• and conducting the grouping in the frequency direction by employing said maximum value and said time direction group information,
• wherein said maximum value of said group number is made variable according to coding quantity information,
• wherein a maximum time group quantity is determined in response to a buffer residual quantity, where the coding quantity information is the buffer residual quantity,
• wherein the candidate for the group boundary in time direction is selected by comparing change in energy information of three successive time groups.

A non-transitory computer-readable record medium having a program recorded therein for conducting grouping of a subband signal obtained by frequency-converting an input audio signal in a time direction and a frequency direction for each frame unit, and encoding the signal for each unit of the grouping, said record medium characterized in having a program recorded therein for causing an information processing device to execute the processes of:

• calculating a candidate for a group boundary in the time direction;
• generating a maximum value of a group number in the time direction;
• employing said candidate, thereby to conduct the grouping in the time direction so that a space of the boundary of the group being generated in the time direction becomes equal to or more than a constant and yet the group number in the time direction becomes equal to or less than said maximum value, and to generate time direction group information;
• and conducting the grouping in the frequency direction by employing said maximum value and said time direction group information,
• wherein said maximum value of said group number is made variable according to coding quantity information,
• wherein a maximum time group quantity is determined in response to a buffer residual quantity, where the coding quantity information is the buffer residual quantity;
• wherein the candidate for the group boundary in time direction is selected by comparing change in energy information of three successive time groups.

An audio encoding device for conducting grouping of a subband signal obtained by frequency-converting an input audio signal in a time direction and a frequency direction for each frame unit, and encoding the signal for each unit of the grouping, characterized in comprising:

• a candidate extraction unit for calculate a candidate for a group boundary in the time direction;
• a maximum value generation unit for generating a maximum value of a group number in the time direction;
• a time group generation unit for employing said candidate, thereby to conduct the grouping in the time direction so that the group number in the time direction becomes equal to or less than said maximum value, and to generate time direction group information;
• and a frequency group generation unit for conducting the grouping in the frequency direction by employing a difference between said maximum value and the number of said groups generated in the time direction,
• wherein said maximum value of said group number is made variable according to coding quantity information,
• wherein a maximum time group quantity is determined in response to a buffer residual quantity, where the coding quantity information is the buffer residual quantity;
• wherein the candidate for the group boundary in time direction is selected by comparing change in energy information of three successive time groups.

An audio encoding method of conducting grouping of a subband signal obtained by frequency-converting an input audio signal in a time direction and a frequency direction for each frame unit, and encoding the signal for each unit of the grouping, characterized in:

• calculating a candidate for a group boundary in the time direction;
• generating a maximum value of a group number in the time direction;
• employing said candidate, thereby to conduct the grouping in the time direction so that the group number in the time direction becomes equal to or less than said maximum value, and to generate time direction group information;
• and
• conducting the grouping in the frequency direction by employing a difference between said maximum value and the number of said groups generated in the time direction,
• wherein said maximum value of said group number is made variable according to coding quantity information,
• wherein a maximum time group quantity is determined in response to a buffer residual quantity, where the coding quantity information is the buffer residual quantity;
• wherein the candidate for the group boundary in time direction is selected by comparing change in energy information of three successive time groups.

A non-transitory computer-readable data storage medium storing an audio encoding program for conducting grouping of a subband signal obtained by frequency-converting an input audio signal in a time direction and a frequency direction for each frame unit, and encoding the signal for each unit of the grouping, characterized in causing an information processing device to execute the processes of:

• calculating a candidate for a group boundary in the time direction;
• generating a maximum value of a group number in the time direction;
• employing said candidate, thereby to conduct the grouping in the time direction so that the group number in the time direction becomes equal to or less than said maximum value, and to generate time direction group information;
• and
• conducting the grouping in the frequency direction by employing a difference between said maximum value and the number of said groups generated in the time direction,
• wherein said maximum value of said group number is made variable according to coding quantity information,
• wherein a maximum time group quantity is determined in response to a buffer residual quantity, where the coding quantity information is the buffer residual quantity;
• wherein the candidate for the group boundary in time direction is selected by comparing change in energy information of three successive time groups.

A non-transitory computer-readable record medium having a program recorded therein for conducting grouping of a subband signal obtained by frequency-converting an input audio signal in a time direction and a frequency direction for each frame unit, and encoding the signal for each unit of the grouping, said record medium characterized in having a program recorded therein for causing an information processing device to execute the processes of:

• calculating a candidate for a group boundary in the time direction;
• generating a maximum value of a group number in the time direction;
• employing said candidate, thereby to conduct the grouping in the time direction so that the group number in the time direction becomes equal to or less than said maximum value, and to generate time direction group information;
• and
• conducting the grouping in the frequency direction by employing a difference between said maximum value and the number of said groups generated in the time direction,
• wherein said maximum value of said group number is made variable according to coding quantity information,
• wherein a maximum time group quantity is determined in response to a buffer residual quantity, where the coding quantity information is the buffer residual quantity;
• wherein the candidate for the group boundary in time direction is selected by comparing change in energy information of three successive time groups.