SVS-Vistek CameraLink svs16000, Руководство пользователя

Страница: 14 / 97

Users Manual Area Scan Cameras SVSGigE Page 14
Users Manual –svs11002 and svs16000
© SVS-VISTEK GmbH
3.3.2 Persistent Network address assignment
Open the dialog by selecting “Network settings” in the context menu that comes up
when right-clicking on an entry in the discovery dialog:
All settings in the “Network configuration”
dialog will be transferred to the camera and
they will be made persistent on the EEPROM
inside the camera. Therefore this dialog
allows for assigning a non-volatile network
address to a camera which it will maintain
also over power off and restart cycles.
Network settings performed in this
dialog become valid but after a reboot
of the camera.
In addition to a persistent network address it
can be decided whether the camera will try
to obtain a dynamic network address from a
DHCP server on start-up. If this option is
used the camera can no longer be identified
uniquely by IP address. Instead, other items
like MAC address, serial number or the user
defined name can be used for this purpose.
Usually the DHCP method is convenient for
involving a camera in to an already existing
network of computers which all obtain their IP addresses from a DHCP server. The
camera will well behave in this environment when adjusting it also to the DHCP mode.
In some cases it may be that a camera will be connected for performance reasons
directly to a dedicated network card. Given that this network card has not got
assigned a valid IP address there is a fall-back procedure defined which takes
automatically a network address from a special range that has been assigned for this
purpose. This is the so called “Link Local Address (LLA)” behavior that constitutes
the last item in a chain of fallback cases. It has been defined in the GigE Vision
standard document that the LLA fall-back procedure can never be deactivated.
Therefore the LLA checkbox is deactivated but always in a checked state.
Currently the default gateway setting is not used and therefore deactivated. This is
related to performance considerations which recommend to operate GigE cameras in
the same network segment where the clients exist and not to route the data stream in
to other networks.
3.3.3 Using jumbo frames
The transport efficiency in the streaming channel can be improved when using “jumbo
frames” in network transport. This will reduce the overhead which is caused by
maintaining header data with each data packet. A network packet has usually a size of
about 1500 bytes which can be increased to e.g. 16112 by switching “jumbo frames”
on. In addition to the network card “jumbo frames” have to be supported also by a
switch that forwards the image data stream from a camera to the PC.
Adjusting higher packet sizes requires network cards that support jumbo packets, e.g.
Intel PRO/1000 PT which offers a single network port or Intel PRO/1000 PT which
offers two network ports. Other cards have to be checked whether they contain an
adjustment which allows for switching “jumbo frames” on.