
CXD1922Q
MPEG-2 Video Encoder
Description
The CXD1922Q is a highly-integrated device includ-
ing MPEG-2 encoder, motion estimation, and system
and rate controller on a single chip. The device is ideal
for low-cost solutions in consumer products, storage
media, communications and authoring applications.
The CXD1922Q supports MPEG-2 MP@ML encoding
compliant with ISO/IEC 13818-2 for frame structure
encoding.
Features
MPEG-2 Encoding
High quality, real-time encoding (I, B, P)
Supports MP@ML, SP@ML
— Image size:
up to NTSC 720x480 @ 30fps
up to PAL 720x576 @ 25fps
— Up to 2 “B-frame” (M
≤
3)
— Frame structure
Encoding rates:
— Max. bit-rates:
15Mbps (P, B)
25Mbps (I only)
MPEG-2 Video Elementary Stream Output
— 4:2:0 chroma format encoding
Motion Estimation
Adaptive motion estimation algorithms
Wide motion vector search range:
— Horizontal:
— Vertical:
Frame/field motion estimation
Half-pel accuracy
Dual-prime encoding support
–288 to +287.5
–96 to +95.5
–1–
2/9
9
Sony reserves the right to change products and specifications without prior notice. This information does not convey any license by any implication
or otherwise under any patents or other right. Application circuits shown, if any, are typical examples illustrating the operation of the devices. Sony
cannot assume responsibility for any problems arising out of the use of these circuits.
Encoding Controller (Closed)
Programmable rate and system control
Host independent operation
CBR and VBR encoding support
Constant length coding for storage media
Adaptive frame/field MC/DCT coding
Programmable parameters including:
— Bit rate
— GOP structure (M&N value)
— Encoding image size
— Quantization matrix tables, etc.
Interfaces
Host: 16-bit controller
Input video: 4:2:2 YUV CCIR 656 (8-bit parallel)
Output video: 8-bit video elementary stream (8-bit
parallel)
External memory: 32Mbits
— Supports SDRAM (e.g. 512K x 16 x 2-chips)
Other:
Supports MPEG-1 encoding
Multi-level temporal noise reduction
Image conversion for free size encoding
4:2:2 to 4:2:0 video conversion
Automatic scene change detection
Automatic 3:2 pull-down inversion
Error concealment I-frame motion vectors
Support for user data insertion in encoded stream
208-Pin QFP (Plastic)