TANDBERG E5780, Справочное руководство

Страница: 13 / 298

Introduction to the Basic Encoder
Reference Guide: evolution 5000 E5780 and E5782 Encoder Page 1-5
ST.RE.E10135.4
Video Input Types
The video input types which are supported are:
•
625-line composite PAL-B, -D, -G, -H or -I (ITU-R BT. 624-4)
•
525-line composite NTSC-M (with and without pedestal) or PAL-M (ITU-R BT. 624-4)
•
Serial digital (ITU-R BT.656-4, part 3) input (D1 serial format) and (ANSI/SMPTE 259M)
(component only)
•
Internal test pattern function
•
480p59/60, 576p50, 720p50, 720p59/60 (SMPTE 296M), 1080i25, 1080i29/30 (SMPTE
274M), 1080pSF/24, 1080pSF/25 and 1080pSF/23.976
Serial Digital Video Input Error Detection and Handling (EDH) (SD Only)
The serial digital video input supports error detection and handling (EDH) as defined by the
specification SMPTE RP 165-1994, ‘Error Detection Checkwords and Status Flags for Use in
Bit Serial Digital Interfaces for Television’.
Video Encoding Functions
The standard video encoding functions include:
•
Support for all MP@ML and 422P@ML
2
standard coding modes
•
Support for some MP@HL and 422P@HL standard coding modes
•
Selectable bitrate operation, <1.5 Mbit/s - 90 Mbit/s (see Table 1.3 )
•
Support for various video picture resolutions (HD and SD)
•
An internal frame synchroniser (see Internal Frame Synchroniser on page 1-6 )
•
Support for Active Format Descriptor (AFD)
•
Support for a variety of Group of Pictures (GOP) structures with a variable number of B
frames
•
Built-in patented adaptive noise reduction circuitry
3
•
A logo overlay facility whereby the Encoder is able to overlay broadcasters
trademarks/logos onto the active video
•
The ability to generate internal video test patterns. These can be moving, with the ability
to load two frames (SD only)
•
With software option M2/ESO2/HD-MPASS enabled, a multi-pass encoding mode is
available which results in more efficient use of bandwidth
•
Support for ZigZag scan which statistically produces the same or more efficient coding on
most slow moving material
Motion Estimation
For SD fully Exhaustive motion estimation is used. It takes a macro block of 16 pixels x 16
pixels and then performs an exhaustive search without subsampling.
For HD a hierarchical motion estimation scheme is used. It creates a hierarchical set of
pictures, each layer containing a vertically and horizontally sub-sampled version of the picture
in the layer above. The motion estimation search begins in the lowest layer that contains the
coarsest image. An exhaustive search of this layer produces an initial location for the best
match.
3
Noise reduction is only available when software option M2/ESO2/NR or M2/ESO2/HDNR is purchased.