Working around performance limitations – PRG Node Plus User Manual
Page 14
12
PRG NODE PLUS USER MANUAL
Working Around Performance Limitations
The embedded processor in the Node has limitations with its buffering and communication processing, and often
cannot keep up with modern computers that send data packets very close together.
In a Virtuoso or Vx76 system, the Node is doing all the calculations to generate DMX512 for the control channels
patched to it. In cases where large numbers of fixtures generating intensity-based traffic - such as LEDs - are patched
to the Node, the processing limitations will become evident during fast effects across many channels. A typical
symptom would be effects not keeping up or continuing to run after being stopped on the console.
In an Art-Net system, the Node is receiving network DMX512 packets and outputting the information on the DMX
ports. The limitation here is in the Node's limited available network packet buffering. If the network packets for
consecutive Art-Net universes are spaced less than about 100 microseconds apart, the Node may drop some of the
packets since they are being sent faster than the Node can receive and process them. A typical symptom would
appear as temporary freezes in the DMX outputs. (Note that this is generally not a problem when the universes are
input to the network by multiple sources.)
Suggestions:
+
Virtuoso/V676 Control Systems - The maximum number of control channels patched to a given Node should not
exceed 250 where there are a large number of LED fixtures patched to the Node.
+
Art-Net Systems - Configure the Art-Net source to increase the time between consecutive network packets, or to
interleave/alternate the universes in a different order. Most of the fast Art-Net devices allow adjustment of their
Art-Net output timing, so this can help to resolve any problems.
+
Spread out the universes mapped to a single Node, since the Node can discard unneeded universes more quickly
than processing mapped universes. For example, it is better to map every third universe to a Node instead of a
string of six consecutive universes.
+
As a general rule, limit the maximum number of universes a Node must process to four (4).
Note: If merging is enabled, the Node must listen to two universes (or more) per output, so the number of output
universes may need to be reduced to one or two with a high-speed Art-Net source. (Refer to
CAUTION!
Since there can be quite a bit of variation in Art-Net controllers, it is always a good idea to test the
controller with all networking gear and Art-Net devices prior to a show situation!