Renesas SH7781 - Manual (Page 1055)

20. Graphics Data Translation Accelerator (GDTA)
Rev.1.00 Jan. 10, 2008 Page 1025 of 1658
REJ09B0261-0100
No. Operation Description
(4) Calculation of
input position
(nth row and
below)
The following formulae are used to compute input positions (nth row and
below) (DDR2-SDRAM input address).
Input address formulae for nth row and below
• Y input comparison address
Calculation formula: Past frame Y point value (base point) + [(mbrow
× 16
+ (Recon_down>>1) + n - 1)
× (width + Y padding)] + [mbcol × 16 +
(Recon_right>>1)]
Past frame Y pointer value (base point): MCPYPR setting address
mbrow: Calculated from MCCF setting
mbcol: Calculated from MCCF setting
Recon_down: Calculated from MCCF setting
Recon_right: Calculated from MCCF setting
width: Calculated from MCWR setting
Y padding: Calculated from MCYPR setting
Depending on the motion vector value, 16-dot data (16 bytes) or 17-dot data
(17 bytes) is processed in succession.
Future frame Y pointer address is calculated using a formula similar to that for
the past frame.
• U/V input comparison address
Calculation formula: Past frame U point value (base point) + [(mbrow
× 8 +
(Recon_down/2>>1) + n - 1)
× (width/2 + U padding)] + [mbcol × 8 +
((Recon_right/2)>>1)]
Past frame U pointer value (base point): MCPUPR setting address
mbrow: Calculated from MCCF setting
mbcol: Calculated from MCCF setting
Recon_down: Calculated from MCCF setting
Recon_right: Calculated from MCCF setting
width: Calculated from MCWR setting
U padding: Calculated from MCUVPR setting
Depending on the motion vector value, 8-dot data (8 bytes) or 9-dot data (9
bytes) is processed in succession.
Future frame U pointer address is calculated using a formula similar to that for
the past frame.
V pointer address is calculated using a formula similar to that for the U pointer
address.
(5) Data reading Input data stored in DDR2-SDRAM is read into the GDTA at the address
computed in (3) and (4).