Visible to Intel only — GUID: myt1639406085026
Ixiasoft
Visible to Intel only — GUID: myt1639406085026
Ixiasoft
20.4. Color Space Converter IP Registers
Address | Register | Access | Description | |
---|---|---|---|---|
Lite 52 | Full | |||
Parameterization registers | ||||
0x0000 | VID_PID | RO | RO | Read this register to retrieve the Color Space Converter product ID. This register always returns 0x6AF7_022F. |
0x0004 | VERSION | RO | RO | Read this register to retrieve the version information for Color Space Converter. |
0x0008 | LITE_MODE | RO | RO | Read this register to determine if Lite mode is on. This register returns 0 when Lite mode is off and 1 when Lite mode is on. |
0x000C | DEBUG_ENABLED | RO | RO | Read this register to determine if Debug features is on. This register returns 0 for off and 1 for on. |
0x0010 | BPS_IN | RO | RO | Read this register to determine the bits per symbol configured for the input data. Range [8 : 16]. |
0x0014 | BPS_OUT | RO | RO | Read this register to determine the bits per symbol configured for the output data. Range [8 : 16]. |
0x0018 | FRAC_BITS | RO | RO | Read this register to determine the number of fractional bits used by the fixed-point data type to store the coefficients and summands Range [0 : 31]. |
0x001C | COEFFS_SIGNED | RO | RO | Read this register to determine if the coefficient fixed-point type has a signed bit. (0 – unsigned, 1 - signed). |
0x0020 | COEFFS_INT_BITS | RO | RO | Read this register to determine the number of integer bits used by fixed-point data type to store coefficients. Range [0 : 16]. |
0x0024 | SUMMANDS_SIGNED | RO | RO | Read this register to determine if the summand fixed-point type has a signed bit. (0 – unsigned, 1 - signed). |
0x0028 | SUMMANDS_INT_BITS | RO | RO | Read this register to determine the number of integer bits used by fixed-point data type to store summands. Range [0 : 20]. |
0x002C | BINARY_POINT_RIGHT_MOVE | RO | RO | Read this register to determine the number of places by which the binary point moves to the right. Use to scale the result of the calculation. Range [-16 : 16]. |
0x0030 | ROUND_METHOD | RO | RO | Read this register to determine the method to remove the fractional bits when converting output result back to integer format. 1: Round half up 2: Round half even 3: Truncate to integer |
0x0034 to 0x011F | - | - | - | Unused. |
Control and debug registers For more information, refer to Control Packets |
||||
0x0120 | IMG_INFO_WIDTH | RW | RO | When Lite mode is on, the expected width of the incoming video fields. When Lite mode is off, the received width in image information packets. |
0x0124 | IMG_INFO_HEIGHT | RW | RO | When Lite mode is on, the expected height of the incoming video fields. When Lite mode is off, the received height in image information packets. |
0x0128 | IMG_INFO_INTERLACE | RW | RO | When Lite mode is on, the expected interlace information of the incoming video fields. When Lite mode is off, the received interlace information in image information packets. |
0x012C | RESERVED | RW | RO | Unused. |
0x0130 | IMG_INFO_COLORSPACE | RW | RO | When Lite mode is on, the expected color space of the incoming video fields. When Lite mode is off, the received color space in image information packets. |
0x0134 | IMG_INFO_SUBSAMPLING | RW | RO | When Lite mode is on, the expected chroma sub-sampling of the incoming video fields. When Lite mode is off, the received chroma sub-sampling in image information packets. |
0x0138 | IMG_INFO_COSITING | RW | RO | When Lite mode is on, the expected chroma co-siting of the incoming video fields. When Lite mode is off, the received chroma co-siting in image information packets. |
0x013C | IMG_INFO_FIELD_COUNT | - | RO | The received field count field in image information packets. |
0x0140 | STATUS | RO | RO | Bit 0: Status bit. 1 means Color Space Converter is processing a video field, 0 otherwise. When Lite mode is off: Bit 1 : Pending register updates bit. Any writes to the color space converter specification registers (0x0148 - 0x0178) cause the IP to raise the pending register updates bit, to indicate outstanding changes to the color space converter settings. The IP lowers this bit at the start of the first packet for a new frame (first packet after the end of packet) after a write to the COMMIT register. |
0x0144 | COMMIT | RW | RW | The IP uses this register with the STATUS register pending register updates bit. Write any value to this register after writing new values to any of the color space converter specification registers (0x0148 - 0x0178). The color space converter switches to the new color space converter specification values at the start of the next video frame after you write to this register. The commit register ensures atomic register updates, avoiding the IP only applying some of the updated register values before the next incoming frame |
Specification registers | ||||
0x0148 | Coefficient A0 | RW | RW | Write to these registers to set the coefficient and summand values. The coefficients and summands are stored in the format indicated by the IP’s parameters. The coefficient and summand registers use integer numbers, with specified number of integer and fractional bits. If signed representation is on, the IP uses signed 2’s complement. |
0x014C | Coefficient B0 | |||
0x0150 | Coefficient C0 | |||
0x0154 | Coefficient A1 | |||
0x0158 | Coefficient B1 | |||
0x015C | Coefficient C1 | |||
0x0160 | Coefficient A2 | |||
0x0164 | Coefficient B2 | |||
0x0168 | Coefficient C2 | |||
0x016C | Summand S0 | |||
0x0170 | Summand S1 | |||
0x0174 | Summand S2 | |||
0x0178 | Output Color Space | RW | RW | Only applicable when Lite mode is off. Write to this register to set the value of color space field in outgoing image information packets from the Color Space Converter. Refer to the Intel FPGA Streaming Video Protocol Specification for more information about color space encoding and image information packets. |
Register Bit Descriptions
Name | Bits | Description |
Color Space Converter version ID and product ID | 31:0 | This register always returns 0x6AF7_022F.
|
Name | Bits | Description |
Register map version | 7:0 | Register map version. Returns 0x01. |
QPDS patch revision | 15:8 | Returns 0x00. |
QPDS update revision | 23:16 | Updated for each release. For 21.4 returns 0x04 |
QPDS major revision | 31:24 | Updated for each release. For 21.4 returns 0x15. |
Name | Bits | Description |
Lite mode parameterization bit | 0 | Returns 1 if lite mode is on. |
Unused | 31:1 | Unused. |
Name | Bits | Description |
Debug features parameterization bit | 0 | Returns 1 if debug features is on. |
Unused | 31:1 | Unused. |
Name | Bits | Description |
BPS_IN | 15:0 | Returns the value of the BPS_IN parameter, indicating the bit per symbol the input interface is configured for. |
Unused | 31:16 | Unused. |
Name | Bits | Description |
BPS_OUT | 15:0 | Returns the value of the BPS_OUT parameter, indicating the bit per symbol the output interface is configured for. |
Unused | 31:16 | Unused. |
Name | Bits | Description |
Coefficient and summand fractional bits | 15:0 | Returns the value of the coefficient and summand fractional bits parameter, indicating the number of fractional bits used by the fixed-point data type to store the coefficients and summands. |
Unused | 31:16 | Unused. |
Name | Bits | Description |
Coefficients signed | 15:0 | Returns the value of the coefficient signed parameter indicating coefficients are stored in a signed format. |
Unused | 31:16 | Unused. |
Name | Bits | Description |
Coefficient integer bits | 15:0 | Returns the value of the coefficient integer bits parameter, indicating number of integer bits used by fixed-point format to represent coefficients. |
Unused | 31:16 | Unused. |
Name | Bits | Description |
Summands signed | 15:0 | Returns the value of the summands signed parameter indicating summands are stored in a signed format. |
Unused | 31:16 | Unused. |
Name | Bits | Description |
Summand integer bits | 15:0 | Returns the value of the summand integer bits parameter, indicating number of integer bits used by fixed-point format to represent summands. |
Unused | 31:16 | Unused. |
Name | Bits | Description |
Binary point right move | 15:0 | Returns the value of the move binary point right parameter, indicating the number of places the binary point is shifted by when scaling the result. |
Unused | 31:16 | Unused. |
Name | Bits | Description |
Round method | 15:0 | Returns the value of the round method parameter, indicating method used to round excess fractional bits from the result. 1: Round half up 2: Round half even 3: Truncate to integer |
Unused | 31:16 | Unused. |
Name | Bits | Description |
Width bits | 15:0 | When Lite mode is on, write to this register to set the expected width of the incoming video fields. When Lite mode is off with Debug features on, this register returns the width field from the most recently received. |
unused | 31:16 | Unused. |
Name | Bits | Description |
Height bits | 15:0 | When Lite mode is on, write to this register to set the expected height of the incoming video fields. When Lite mode is off with Debug features on, this register returns the height field from the most recently received. image information packet. |
unused | 31:16 | Unused. |
Name | Bits | Description |
interlaced bits | 3:0 | When Lite mode is on, write to this register to set the expected interlacing of the incoming video fields. When Lite mode is off with Debug features on, this register returns the intlaceNibble field from the most recently received image information packet. |
unused | 31:4 | Unused. |
Name | Bits | Description |
Colorspace code bits | 6:0 | When Lite mode is on, write to this register to set the expected color space of the incoming video fields. When Lite mode is off with Debug features on, this register returns the CSP field from the most recently received image information packet. |
unused | 31:7 | Unused. |
Name | Bits | Description |
Subsampling code bits | 1:0 | When Lite mode is on, write to this register to set the expected chroma sub-sampling of the incoming video fields. When Lite mode is off with Debug features on, this register returns the SUBSA field from the most recently received image information packet. |
unused | 31:2 | Unused. |
Name | Bits | Description |
Cosite code bits | 1:0 | When Lite mode is on, write to this register to set the expected chroma co-siting of the incoming video fields. When Lite mode is off with Debug features on, this register returns the COSITE field from the most recently received image information packet. |
unused | 31:2 | Unused. |
Name | Bits | Description |
Status bit | 0 | 1 means the Color Space Converter is processing a video field, 0 otherwise. |
Pending register updates bit | 1 | 1 means the Color Space Converter has pending updates, 0 otherwise. |
Name | Bits | Description |
Commit bits | 31:0 | Write to any bits to trigger a commit. |
Name | Bits | Description |
Coefficient_value | (Coeff_signed + coeff_int_bits + Coeff_sum_frac_bits):0 | This block of registers allows the setting of coefficient. The coefficients are represented by fixed-point format, and are determined by the IP’s parameterization. |
Name | Bits | Description |
Summand_value | (Summand_signed + summand_int_bits + Coeff_sum_frac_bits):0 | This block of registers allows the setting of summand. The summands are represented by fixed-point format, and are determined by the IP’s parameterization. |
Name | Bits | Description |
Output Colorspace | 6:0 | Write to this register to set the color space code for outgoing image information packets when you turn off Lite mode. Refer to Intel FPGA Streaming Video Protocol Specification for more details on color space field of image information packets. |
When you turn on lite mode, registers are RW only if you turn on Debug features, otherwise they are WO. For full, turn off lite mode.