E681 setup - High number of sequence errors.

[Barnabybear]

Just done more tests.

Each packet contains a sequence number that increments from 1 to 255 then loops back to 0. If a received packet doesn't have the expected sequence number, the sequence error count is incremented.

Using Windows 7 32 bit - Vixen 2.1.4.0. - J1Sis 1.0.0.1. - E681 2.20. if I send 1 universe no errors if I send 2 or more unverses the error count is = packets recieved -1.
Some how it looks like the packet addressing or the order is getting messed up when I send more than one universe. Is there an easy way to view the data thats being sent?

[budude]

Just done more tests.


Using Windows 7 32 bit - Vixen 2.1.4.0. - J1Sis 1.0.0.1. - E681 2.20. if I send 1 universe no errors if I send 2 or more unverses the error count is = packets recieved -1.
Some how it looks like the packet addressing is getting messed up when I send more than one universe. Is there an easy way to view the data thats being sent?

A couple choices - - SACNView does have a viewer - might have a logger as well but haven't checked. Another choice would be to download WireShark and kick off a packet capture and then start off your sequence - stop when you're satisfied and then take a look at what came out. Maybe compare between the single and 2x runs.

[j1sys]

OOOPS!!

I just did a wireshark dump of the output of the plugin. It is doing a single sequence number across all universes. The spec calls for a sequence number PER universe.

Wrong (current):

Univ 101, Seq 0
Univ 102, Seq 1
Univ 103, Seq 2
Univ 104, Seq 3
Univ 101, Seq 4
Univ 102, Seq 5
Univ 103, Seq 6
Univ 104, Seq 7
...

Should be (according to my reading of the spec):

Univ 101, Seq 0
Univ 102, Seq 0
Univ 103, Seq 0
Univ 104, Seq 0
Univ 101, Seq 1
Univ 102, Seq 1
Univ 103, Seq 1
Univ 104, Seq 1
...

I will issue a revise set of plugins for Vixen 2.1, 2.5, 3.0 ASAP. And will also advise the other software people who have based some of the driver on my code of the errata.

So for now just ignore those error counts.

My apologies,

-Ed

[budude]

Wow - wasn't expecting that to be the outcome! Nice follow-up Ed!

[pmcpa]

This MIGHT explain some of the seq errors I see in LSP... it isn't an issue though, as the sequence still plays back! Thanks for looking into this Ed!

For others future reference, all of the above troubleshooting steps that everyone has posted are great places to start!

[rokkett]

I use WireShark for debugging my E1.31 packets. It has an ACN decoder to help find these types of errors. Open Edit->Preferences->Protocols->ACN and enable the "Decode ACN" and "Streaming DMX" options and restart WireShark.

[Barnabybear]

Hi, all it looks like we've bottomed this one.
Thanks for all the advice and comments. I have a couple of new bits of software in the toolbox and a greater understanding of my setup - I did wonder how I'd managed to break something that everyone else had got to work though.
It's good to know that Ed is sorting the indexing - thanks Ed it's an invaluable plugin. I'm on holiday but will look forward to giving it a try when I get home.

[RPM]

I have always noticed these errors as well, but always figured it was something in the sending app for the same reasons noted... this just confirms it.

Since the sequence number field is 1 byte in the E1.31 framing layer, this can only be 0-9 so I can see where this could cause problems with multiple universes as well.

Ed, which document shows the sequence details? The ESTA BSR-E1.31 I have doesn't show this detail.

OOOPS!!

I just did a wireshark dump of the output of the plugin. It is doing a single sequence number across all universes. The spec calls for a sequence number PER universe.

Wrong (current):

Univ 101, Seq 0
Univ 102, Seq 1
Univ 103, Seq 2
Univ 104, Seq 3
Univ 101, Seq 4
Univ 102, Seq 5
Univ 103, Seq 6
Univ 104, Seq 7
...

Should be (according to my reading of the spec):

Univ 101, Seq 0
Univ 102, Seq 0
Univ 103, Seq 0
Univ 104, Seq 0
Univ 101, Seq 1
Univ 102, Seq 1
Univ 103, Seq 1
Univ 104, Seq 1
...

I will issue a revise set of plugins for Vixen 2.1, 2.5, 3.0 ASAP. And will also advise the other software people who have based some of the driver on my code of the errata.

So for now just ignore those error counts.

My apologies,

-Ed

[jstjohnz]

I have always noticed these errors as well, but always figured it was something in the sending app for the same reasons noted... this just confirms it.

Since the sequence number field is 1 byte in the E1.31 framing layer, this can only be 0-9 so I can see where this could cause problems with multiple universes as well.

Ed, which document shows the sequence details? The ESTA BSR-E1.31 I have doesn't show this detail.

From ANSI E1.31 2009:

"Sources shall maintain a sequence for each universe they transmit. The sequence number for a universe
shall be incremented by one for every packet sent on that universe."

It's a 1-byte field so sequence numbers range from 0-255.

[RPM]

Ahhh.... I just noticed the spec shows it as an octet, so this would be correct. My bad.

From ANSI E1.31 2009:

It's a 1-byte field so sequence numbers range from 0-255.