Encoder and decoder

Details Encoder and decoder Encoder and decoder

1999-01-19

2001-05-08

JeanPierre, Peguy (Department: 2819)

Coded data generation or conversion

Digital code to digital code converters

To or from mixed base codes

C708S190000

Reexamination Certificate

active

06229461

ABSTRACT:

BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an encoder for dividing data Y into a scale factor SF, a word length WL, and a data value X and for encoding the divided data, and a decoder for decoding the encoded data Y to obtain decoded data Y.
2. Description of the Related Art
Conventionally, audio or image data is encoded into a compressed form for storage or transmission. One such data encoding methods may be one in which data is divided into a scale factor SF, a word length WL, and a data value X before encoding. With this method, a word length WL and a scale factor SF are obtained from the original value of data Y. Moreover, a data value X is also obtained from the Expression below.
X=Y·{2
(WL−1)
−1}/SF
wherein the scale factor SF is divided into SFN and SFO based on the relationship expressed as SF=2
(SPN/3-SFO)
.
Also, for decoding, the original value of data Y must be calculated using the inputted SFN, SFO, WL, and X, based on the following expressions.
Y=SF·X/{2
(WL−1)
−1},
wherein
SF=2
(SPY/3-SFO)
Conventionally, the above values are calculated by a microcomputer, or the like, using appropriate software.
However, such calculation using software may take time. In particular, when an MD (mini-disk) playback machine is used, it is preferable to use a single integrated circuit for the above decoding operation, and the entire circuit is therefore desired to have a simple structure and capable of high speed calculation. This leads to a demand for a hardware calculation circuit having a simple structure.
SUMMARY OF THE INVENTION
The present invention aims to provide an encoder and a decoder capable of the above calculations, and having a relatively small size circuitry structure.
According to an encoder of this invention, a scale factor SF is divided into an index value SFF and a mantissa value SFV. {2
(WL−1)
−1}/SFV, which is determined based on a mantissa value SFV and a word length WL, is stored in the ROM. With this arrangement, the number of values to be stored in the ROM can be reduced according to the number of values which the mantissa value SFV may be. This makes it possible to reduce the capacity of the ROM.
Also, according to a decoder of the invention, a scale factor SF is divided into an index value SFF and a mantissa value SFV. {2
(WL−1)
−1}/SFV, which is determined based on a mantissa value SFV and a word length WL, is stored in the ROM. With this arrangement, the number of values to be stored in the ROM can be reduced according to the number of values which the mantissa value SFV may be. This makes it possible to reduce the ROM capacity.
Preferably, a mantissa value SFV may be either 1, 2
⅓
, or 2
⅔
so that 2
(WL−1)
−1, {2
(WL−1)
−1}/2
⅓
, and {2
(WL−1)
−1}/2
⅔
are stored in the ROM. Since there are only three values for a mantissa value SFV, a large capacity ROM is not required.
The encoder calculates Y·2
SFF
by shifting the bits of data Y. Since a calculation based on 2
SFF
is a multiplication of 2
−SFF
with respect to either data Y or an output from the ROM, such a calculation is achievable by shifting the bits of data Y according to an index value SFF. This enables reducing the number of multiplications to only one. Note that multiplication of an index value of data Y can be calculated through addition/subtraction with respect to an index value SFF and bit-shifting. This enables a high speed calculation to be made using a simple structure unit.
The decoder calculates X·2
SFF
by giving an index value to the data value X or shifting the bits of the data value X. A calculation based on 2
SFF
is a multiplication with respect to either data X or an output from the ROM. This calculation, however, can be made, when data is handled as floating point data, by giving an index value SFF as an index value of data X. Also, when data is handled as fixed point data, the above calculation can be made by shifting data X according to an index value SFF. Therefore, the number of application of a multiplication can be reduced to only once. This makes it possible to achieve high speed calculation using a unit with a simple structure.


REFERENCES:
patent: 4001566 (1977-01-01), Stevenson
patent: 4100602 (1978-07-01), Shapiro
patent: 4458325 (1984-07-01), Nakata et al.
patent: 5642111 (1997-06-01), Akagiri
patent: 5682461 (1997-10-01), Silzle et al.

Affiliated with

Also associated with

Rating

Encoder and decoder does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Encoder and decoder, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Encoder and decoder will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2504243