User`s manual

User’s Manual v.13 QVidium
®
H.264 HD Video Codecs™
Copyright 2011-2013 QVidium
®
Technologies, Inc. Page 23 of 67
At the receiver, the ARQ error correction mechanism first buffers incoming packets in a
synchronized network de-jitter buffer that maintains a constant, configurable delay for incoming
packets. The de-jitter buffer processes RTP sequence numbers and reorders packets as
necessary to restore proper packet sequence and to buffer against erratic network delays. Its goal
is to output a smooth, consistent flow of packets with increasing sequence numbers. This output
feeds the ARQ error correction mechanism.
The ARQ unit at the receiver watches for gaps in RTP sequence number from the de-jitter buffer
output, and immediately sends out retransmission requests when necessary. Incoming
retransmitted ARQ packets bypass the de-jitter queue and find their correct place in the ARQ
buffer, so that by the time packet are output by the ARQ buffer, the ARQ error correction
mechanism should no longer contain any missing packets.
When the encoder and decoder are both started, the decoder automatically measures the round-
trip delay and packet jitter to the encoder and automatically configures ARQ for optimal operation
according to a few basic user-specified parameters, as described in Section 3.5.2.
Note: For testing ARQ error correction, select ARQ transport at both the encoder and decoder and
use the default ARQ parameters at the decoder (receiver).
3.6 Configuring Advanced Video Encoding Profiles
The encoder has a set of advanced configuration settings that you can configure from the
Advanced Encoder Profile tab as shown in Figure 1b below. These parameters are divided into 3
sections: Advanced Encoding parameters, Output parameters, and Network parameters.
Warning: Some combinations of parameters can cause the encoder to malfunction. QVidium does
not test nor can it support all possible combinations of parameters. However, our philosophy is to
allow users to have the flexibility of setting these parameters to suit their needs and make the
decision for themselves as to whether the resulting operation meets the needs of his application.
To get the system into a known reliable configuration, highlight the profile you wish to reset and
click Default. This will reset all the parameters to a known set of pre-tested values.
Advanced Encoding Parameters:
Stream Mode: This selects the type of bit-rate shaping the encoder uses to create the
multiplexed audio/video MPEG Transport Stream. You can select between Constant Bit Rate
(CBR), which is now the default, Capped Variable Bit Rate (Capped VBR), and Low Delay
VBR. MPEG compressed video normally arrives in bursts. CBR inserts extraneous null data to
fill the gaps and maintain a constant bit rate for the output stream. This helps maintain
interoperability with legacy video decoders. In Capped and Low Delay VBR, the Stream Bit
Rate only sets a peak value for the output, and the encoder does not add any padding to the
stream, making it more bandwidth efficient.
Closed Caption (CC): This allows you to select whether Closed Captions are passed-through
into the output stream, and the type of Closed Captioning, You can select among Line 21,
CEA-608 and CEA-708. For CEA-608 or CEA-708, the QVENC will add Closed Caption data
from the Vertical Ancillary Data (VANC) in the SDI input. (CEA-608 and CEA-708 are only
available for the QVENC.)
File Size Max (Bytes): This parameter only applies when writing the output stream to a file
(File TS and File MP4). When writing an unlimited, continuous video stream, this parameter
invokes a stream segmentation mechanism that breaks the output files into series of files, each
limited to the size specified by this setting. This can be useful for indexing the video stream. It
also helps when writing to a FAT32 (MSDOS) file system, which has a 4 Gigabyte limit on the
maximum file size. In addition, the current version of the firmware does not support MP4 files
larger than 4 Gigabytes. Setting this to “Unlimited” disables this mechanism, resulting in a
single large output file.