Coding apparatus for digital signal

• US 5,264,846 A
• Filed: 03/27/1992
• Issued: 11/23/1993
• Est. Priority Date: 03/30/1991
• Status: Expired due to Term

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Oikawa, Yoshiaki
• Primary Examiner(s): Hoff, Marc S.
• Application Number:
  US07/859,116
• Time in Patent Office: 606 Days
• Field of Search: 341/76, 341/77, 381/29-40
• US Class Current: 341/76
• CPC Class Codes: H04B 1/665 using psychoacoustic properties of the ear, e.g. masking effect

Abstract

An apparatus for compressing a digital signal in which a block of a spectral components in the frequency domain is divided in frequency into at least two sub blocks. The peak value for each sub block is determined, and the difference between the peak values is determined. Block floating is applied to each sub block individually when the difference between the peak values is above a fixed value. This prevents quantizing noise from being heard when the input signal is highly tonal.

Claims

1. An apparatus for compressing a digital input signal, the apparatus comprising:

a deriving mans for deriving from the digital input signal plural spectral components in the frequency domain, the spectral components being divided into blocks corresponding to critical bands;

means for dividing one of the blocks of spectral components corresponding to critical bands in frequency into sub blocks;

peak value determining means for determining a peak value for each sub block;

comparing means for determining a difference in the peak values of the sub blocks; and

block floating means for applying block floating processing to the sub blocks individually when the difference from the comparing means is above a fixed value.

2. The apparatus of claim 1, wherein, when the difference from the comparing means is within the fixed value, the block processing means applies block floating processing to the sub blocks using a common scale factor.

3. The apparatus of claim 1, wherein

the peak value determining means determines the peak value for each sub block by determining a maximum absolute value of all th spectral components in the sub block;

and

the block floating means applies block floating to the spectral components in each sub block using the peak value determined by the peak value determining mans as a scale factor common to all the spectral components in the sub block.

4. The apparatus of claim 3, wherein the fixed value is 6 dB.

5. The apparatus of claim 3, wherein the block floating means includes a requantizing means for requantizing the spectral components in the sub blocks individually when the difference from the comparing means is above the fixed value.

6. The apparatus of claim 1, wherein the deriving means includes:

a means for dividing the input signal into a signal component in each of plural frequency ranges;

a means for dividing in time the signal in each frequency range into blocks;

an orthogonal transform means for orthogonally transforming each block of each frequency range signal to provide the spectral components in the frequency domain; and

a means for dividing the spectral components in the frequency domain into blocks corresponding to critical bands.

7. The apparatus of claim 6 wherein the orthogonal transform means applies a faster Fourier transform (FFT).

8. The coding apparatus of claim 6 wherein the orthogonal transform means applies a discrete cosine transform (DCT).

9. An apparatus for compressing a digital input signal, the apparatus comprising:

a deriving mans for deriving plural spectral components from the digital input signal, the plural spectral components being divided into blocks, each block corresponding to a critical band;

block dividing means for dividing one of the blocks of spectral components into two sub blocks;

tonality signal generating means for generating a tonality signal indicating when part of the input signal represented by the spectral components in the block is tonal; and

a block floating means for applying block floating processing to both sub blocks using a common scale factor, the block floating means applying block floating to each sub block individually in response to the tonality signal.

10. The apparatus of claim 9, wherein the tonality signal generating means includes:

a means for determining a peak value for each sub block; and

a means for determining a difference between the peak values of the sub blocks, and for generating the tonality signal when the difference is greater than a predetermined value.

11. The apparatus of claim 10, wherein:

the block floating means includes a means for selecting the greater of the peak values of the sub blocks as the common scale factor; and

the block floating means applies block floating processing to each sub block individually using the peak value for each sub block as a scale factor for the sub block in response to the tonality signal.

12. The apparatus of claim 9, additionally comprising a requantizing means for requantizing the spectral components in both sub blocks using a common step size, the requantizing means requantizing the spectral components in each sub block using a different step size in response to the tonality signal.

13. The apparatus of claim 12, wherein:

the apparatus additionally comprises a means for determining an allocated bit number for each critical band; and

the requantizing means includes a step size determining means for determining the common step size by dividing the common scale factor by a number of steps corresponding to the allocated bit number, the step size determining a step size for each sub block by dividing the scale factor for each sub block by the number of steps corresponding to the allocated bit number in response to the tonality signal.

14. The apparatus of claim 13, wherein:

the block floating means includes a means for selecting the greater of the peak values of the sub blocks as the common scale factor; and

the block floating means applies block floating processing to each sub block individually using the peak value for each sub block as a scale factor for the sub block in response to the tonality signal.

15. The apparatus of claim 9, wherein the deriving means includes:

a means for dividing the input signal into a signal component in each of plural frequency ranges;

a means for dividing in time the signal in each frequency range into blocks;

an orthogonal transform means for orthogonally transforming each block of each frequency range signal to provide the spectral components in the frequency domain; and

a means for dividing the spectral components in the frequency domain into blocks corresponding to critical bands.