configuration

Toolchain configuration

Before a chain process can be started, a configuration has to be made. The better this configuration is prepared, the less efforts have to be done if changes have to be done later. In the following the configuration concept is presented in short words and some additional tipps are given to keep an overview.

The concept

After the setup has been successfully executed, the configuration files in the config folder have to be filled out. In its basic structure, the configuration is build as a tree. The root configuration file is the PVS matrix: pvs.cvs. This file connects the sources from the srcVid folder with different settings listed in the HRC table, which can be found in the src.csv resp. hrc.csv. Each dataset contains an individual and unique ID. In each configuration file further explainations are given for the appropriate use-cases. It has to be stated out that the configuration is splitted up in further sub-configuration files, where further settings (which are also reuseable), can be declared. The appropriate files are located in the hrc folder.

In basic each table is saved in the csv-format, splitted with semicolons (;). In the first column each time an ID has to be defined, as modeled in the following table:

id_field_name	field_1_name	field_2_name	...	field_n_name
id_1	field_1,1_value	field_1,2_value	...	field_1,n_name
id_2	field_2,1_value	field_2,2_value	...	field_2,n_name
...	...	...	...	...
id_m	field_m,1_value	field_m,2_value	...	field_m,n_name

pvs.csv

The pvs.csv file links the settings from the hrc.csv with the src.csv file.

Fields:

• [int] pvs_id: the id of the PVS dataset
• [int] src_id: the id of the SRC dataset
• [int] hrc_id: the id of the HRC dataset

src.csv

Fields:

• [int] src_id: the id of the SRC dataset
• [string] src_name: the name of the source file located in the srcVid folder

hrc.csv

Fields:

• [int] hrc_id: the id of the HRC set
• [int] encoding_id: the id of a referenced encoding dataset, described in hrc/encoding.csv
• [int] packet_loss_id: the if of a reference packet loss setting, described in the hrc/packet_loss.csv
• [string] coder: the name of the coder used in this chain. Possible values are:
• ffmpeg: www.ffmpeg.org
• [string] stream_mode: the mode used for streaming. Possible values are:
• mpegts-udp: for UDP streaming with MPEGTS container
• mpegts-rtp: for RTP streaming with MPEGTS container
• raw-rtp: for RTP streaming without a container

Tips & Tricks

• With # in front of a single line in a configuration file it is possible to write comments and/or to comment out existing configurations
• The used ids do not have to be exactly increasing values. You are free in combining values, i.e. an ID 240 could indicate that the setting referenced with this id has an fps value of 24 and a packet_loss_rate of 0%.

Example

A detailed example about the project configuration is given here.