Sound Reinforcement - Forums for Live Sound Professionals - Your Displayed Name Must Be Your Real Full Name To Post In The Live Sound Forums > Console Connectivity Solutions

Yamaha CL series

<< < (4/4)

Mac Kerr:

--- Quote from: Jim Wilkens on July 12, 2014, 03:32:45 pm ---Kieran, how exactly do you do that camera phone trick?

--- End quote ---

No trick involved. The sensors in digital cameras as sensitive in the infra-red range, the lasers that illuminate fiber are in the IR range. Just look into the end of an illuminated fiber with a digital camera and you will see the light if it's there. You can try this out with the IR remote for your TV.

Mac

Jim Wilkens:

--- Quote from: Mac Kerr on July 12, 2014, 04:05:21 pm ---No trick involved. The sensors in digital cameras as sensitive in the infra-red range, the lasers that illuminate fiber are in the IR range. Just look into the end of an illuminated fiber with a digital camera and you will see the light if it's there. You can try this out with the IR remote for your TV.

Mac

--- End quote ---

Got it.I think the "trick" I needed is to just unplug the LC connector from the transceiver and point it at the camera.

Kieran Walsh:
For the purposes of completeness for the switch torture bit...

The last tranche covered "how do I tell if I have QoS setup ad working... and more importantly... do I even need it?"

The other major consideration is Multicast.

In the interests of "doing it backwards" - good multicast management is more important than QoS... the reason for doing this backwards is that often times good multicast management will mask bad QoS setup, so in testing, always stress the system to the maximum extent of the least significant action first.

Now Multicast pruning is achieved on a Dante network using the IGMP standard. We have found that some switches just don't have the horsepower to deal with a lot of real-time multicast packets. The tricky part of this is that if you look on the datasheet it will say IGMP snooping... and a nice tick... It is of course a fair claim... the switch will have this feature if it says so...

Most lower cost switches manage IGMP functions in a microcontroller inside the switch... this is a fair approach, as its a lower cost processing platform, and in 99% of cases it is perfectly adequate.

So - once again... smash as much multicast as you can into the switch... get the system so that it is stable (no pops and cracks for at least 15 minutes)... then if your switch has it... press the IGMP snooping enable button.

You should hear absolutely no difference. NB you should allow the system to run for at least double the IGMP lease time to ensure something funny is not going on... also at present this experiment should not be conducted with Apple computers sending multicast - Part of Apples IGMP implementation at present can be misinterpreted by some switches, and lead to blame being wrongly apportioned - Windows (or Dante-Enabled Hardware) does not suffer from this at the moment.

Again, if there is a drop-out, or some other undesirable operation... try reducing the bandwidth (channel count) that you are hitting the switch with to establish a SWL.

The case I mentioned before had us in the situation that at no channel count was the performance acceptable for IGMP snooping to be enabled. We still "passed" the switch for the desired application because the only reason we wanted to filter multicast was to stop the wireless access point from being flooded (this would render the Wifi controller useless). As it so happened the Cisco Aironet AP had an ingress multicast filter, so filtering in the switch (whilst the "best" method) was unnecessary as the overall system application was still fulfilled.

In this circumstance relaxing the constraints of the project saved in the order of $15,000-$20,000 whilst still actually delivering what was really required.

Navigation

[0] Message Index

[*] Previous page

Go to full version