|
|
PDF STV3208 Data sheet ( Hoja de datos )
| Número de pieza | STV3208 | |
| Descripción | 8 x 8 DISCRETE COSINE TRANSFORM | |
| Fabricantes | ST Microelectronics | |
| Logotipo |  |
|
Hay una vista previa y un enlace de descarga de STV3208 (archivo pdf) en la parte inferior de esta página. Total 17 Páginas | ||
|
No Preview Available !
STV3208
8 x 8 DISCRETE COSINE TRANSFORM (DCT)
. 0 TO 27MHz PIXEL RATE IN SINGLE
PRECISION MODE,
. 0 TO 20 MHz PIXEL RATE IN DOUBLE
PRECISION MODE
. FORWARD AND INVERSE 8 x 8 TRANS-
FORM
. 9-BIT TWO’S COMPLEMENT PIXEL FORMAT
. 12-BIT TWO’S COMPLEMENT COEFFICIENT
FORMAT
. OPTIMIZED ACCURACY FOR 8-BIT TWO’S
COMPLEMENT PIXEL FORMAT
. SELECTABLE SCANNING OF COEFFICIENT
BLOCKS
. FULLY TTL AND CMOS COMPATIBLE
. CMOS TECHNOLOGY
. SINGLE +5 VOLT POWER SUPPLY
. MAXIMUM POWER DISSIPATION : 750mW
AT 27MHz
DIP40
(Plastic Package)
DESCRIPTION
The STV3208 is a dedicated circuit for the 8 x 8
discrete cosine transform(DCT) computation.Two-
dimensional forward DCT (FDCT) or inverse DCT
(IDCT) is performed for 8 x 8 block sizes and a pixel
rate up to 27MHz. The circuit architecture is fully
bidirectional with 9-bit magnitude pixel data bus
and a 12-bit magnitude coefficient data bus pro-
grammed as input or output depending on the
selection of FDCT or IDCT.
Pixel Bus
Coefficient Bus
FDCT
Input
Output
IDCT
Output
Input
Data Format
9-bit 2’s
Complement
12-bit 2’s
Complement
For the forward transform, the input pixels are
coded on 9-bit 2’s complement and the output
coefficients are coded on 12-bit 2’s complement.
For the inverse transform, the data format is iden-
tical with the coefficients used as input and the
pixels used as output.
Two operating modes are provided : single preci-
sion mode at a pixel rate up to 27 MHz, and double
precision mode at a pixel rate up to 20 MHz.
PQFP44
(Plastic Package)
ORDER CODES
Part
Number
STV3200CP
STV3200CV
Temperature
Range
0 to 70oC
0 to 70oC
Package
DIP 40
PQFP44
July 1992
1/17
Free Datasheet http://www.datasheet4u.net/
1 page  
STV3208
If CSS is low, the coefficient block is scanned line by line and the pixel block is scanned column by column,
or the coefficient block is scanned column by column and the pixel block is scanned line by line.
Figure 2b : Data Ordering (CSS low)
ORDER 1
PIXEL ORDER
COEFFICIENT ORDER
1 9 17 25 33 41 49 57
12345678
2 10 18 26 34 42 50 58
9 10 11 12 13 14 15 16
3 11 19 27 35 43 51 59
17 18 19 20 21 22 23 24
⇔4 12 20 28 36 44 52 60
25 26 27 28 29 30 31 32
5 13 21 29 37 45 53 61
33 34 35 36 37 38 39 40
6 14 22 30 38 46 54 62
41 42 43 44 45 46 47 48
7 15 23 31 39 47 55 63
49 50 51 52 53 54 55 56
8 16 24 32 40 48 56 64
57 58 59 60 61 62 63 64
ORDER 2
12345678
9 10 11 12 13 14 15 16
17 18 19 20 21 22 23 24
25 26 27 28 29 30 31 32
33 34 35 36 37 38 39 40
41 42 43 44 45 46 47 48
49 50 51 52 53 54 55 56
57 58 59 60 61 62 63 64
⇔
1 9 17 25 33 41 49 57
2 10 18 26 34 42 50 58
3 11 19 27 35 43 51 59
4 12 20 28 36 44 52 60
5 13 21 29 37 45 53 61
6 14 22 30 38 46 54 62
7 15 23 31 39 47 55 63
8 16 24 32 40 48 56 64
3. DATA FORMAT
Coefficients format is 12-bit 2’s complement, cor-
responding to the range -2048 to 2047.
There are 2 possible ranges for pixel data :
9-bit two’s complement magnitude
(see Figure 3)
The pixel data range is -256 to +255. In this case
the PR pin must be set to 0 for IDCT. D8 is the most
significant bit and D0 the least significant bit for the
pixel data. A clipping to the range -256 to +255 is
performed before outputting reconstructed pixels
after an IDCT.
8-bit two’s complement magnitude
(see Figure 4)
Figure 3
Pixel data range is -128 to +127. In this case D0
must be set to 0 and the PR Pin must be set to 1
for IDCT. D8 is the most significant bit and D1 the
least significant bit for the pixel data. A clipping to
the range -128 to +127 is performed before output-
ting reconstructed pixels after an IDCT.
This mode may be used for intra picture coding. In
this case, pixel data range is 0 to 255. For a FDCT,
the most significant bit of input pixel data (D8) must
be invertedbefore entering the chip. This isequiva-
lent to substract 128 to the input pixel data. Note
that this operation will only have effect on the DC
value F(0,0). For an IDCT, the most significant bit
of output pixel data (D8) must be inverted. This is
equivalent to add 128 to the output pixel data.
5/17
Free Datasheet http://www.datasheet4u.net/
5 Page 
STV3208
A:
Characteristics of IDCT. Error between the IDCT computed with 64-bit floating point accuracy and the IDCT
computed by the STV3208 is measured. Measures have been done according to the CCITT WGXV method
Peak Error
Peak Mean Square Error
Overall Mean Square Error
Peak Mean Error
Overall Mean Error
Single Precision
1
0.0403
0.0287
0.0125
0.0050
Double Precision
1
0.0258
0.0200
0.0041
0.0000062
Exact value
Errors of ± 1 LSB
Errors of ± 2 LSB
9-bit pixels
Single Precision Double Precision
97.1 %
98.0 %
2.9 %
2.0 %
0% 0%
8-bit pixels
Single Precision Double Precision
99.96 %
99.97 %
0.04 %
0.03 %
0% 0%
B:
Characteristics of FDCT. Error between the FDCT computed with 64-bit floating point accuracy and the
FDCT computed by the STV3208 is measured.
Exact value
Errors of ± 1 LSB
Errors of ± 2 LSB
9-bit pixels
Single Precision Double Precision
93.6 %
96.9 %
6.4 %
4.1 %
0% 0%
8-bit pixels
Single Precision Double Precision
93.6 %
96.9 %
6.4 %
4.1 %
0% 0%
C:
Characteristics of FDCT followed by an IDCT. Error between the source picture and the FDCT computed
by the STV3208 followed by an IDCT computed by the STV3208 is measured.
Exact value
Errors of ± 1 LSB
Errors of ± 2 LSB
9-bit pixels
Single Precision Double Precision
89.3 %
90.6 %
10.7 %
9.4 %
0% 0%
8-bit pixels
Single Precision Double Precision
99.88 %
99.92 %
0.12 %
0.08 %
0% 0%
11/17
Free Datasheet http://www.datasheet4u.net/
11 Page | ||
| Páginas | Total 17 Páginas | |
| PDF Descargar | [ Datasheet STV3208.PDF ] | |
Hoja de datos destacado
| Número de pieza | Descripción | Fabricantes |
| STV3208 | 8 x 8 DISCRETE COSINE TRANSFORM | ST Microelectronics |
| Número de pieza | Descripción | Fabricantes |
| SLA6805M | High Voltage 3 phase Motor Driver IC. |
 Sanken |
| SDC1742 | 12- and 14-Bit Hybrid Synchro / Resolver-to-Digital Converters. |
 Analog Devices |
|
DataSheet.es es una pagina web que funciona como un repositorio de manuales o hoja de datos de muchos de los productos más populares, |
| DataSheet.es | 2020 | Privacy Policy | Contacto | Buscar |