[aravis] Re: Debugging latency & long pause issues with Blackfly camera...
- From: Patrick Doyle <wpdster@xxxxxxxxx>
- To: aravis@xxxxxxxxxxxxx
- Date: Mon, 14 Apr 2014 20:12:39 -0400
Thank you Emmanuel,
I've poked around a bit today (well, all day, actually), and it
appears that, in my system, packets are being dropped for some reason.
I don't know if they are never sent by the camera, are dropped by the
switch, or dropped by my embedded board. Your integration of
ARV_DEBUG in the source code looks like the right tool (along with a
packet sniffing switch) to help me figure out what's going on. Also,
I have learned that the NIC on my board does not support jumbo
packets. It seems that, once one packet is lost, a whole storm of
them get lost following that. I hope to learn more on Wednesday. If
there is any data I can collect or trials you would like me to run, I
will do whatever I can on my end.
Thank you again.
--wpd
On Mon, Apr 14, 2014 at 3:53 PM, Emmanuel Pacaud <emmanuel@xxxxxxxxx> wrote:
>
> Hi Patrick,
>
> Le lundi 14 avril 2014 à 09:38 -0400, Patrick Doyle a écrit :
>> This shows a rather long pause at the beginning (137 seconds, which
>> correlates with the wall clock time I measured) getting the pipeline
>> up and running, followed by 8 frames whose timestamps differ by 34e6
>> nanoseconds (34 ms, which is sort of correct for 30 frames/sec),
>> followed by a 19 second pause (again, correlates well with the wall
>> clock), followed by more frames.
>>
>> I'm not asking that anybody solve this (although, if you've seen this
>> sort of thing with the Blackfly, or other cameras and can tell me what
>> to fix, you would have my eternal gratitude), but if anybody has some
>> suggestions as to what I might look at, or what tools I might use, to
>> diagnose these pauses, I would be grateful.
>
> I have a blackfly camera with which I face some troubles similar to
> yours. I think it's due to the fact the register giving the Time Tick
> Frequency returns a wrong value. And this value is used to compute the
> image buffer timestamps.
>
> I'll try to fix this issue by tomorrow.
>
> Cheers,
>
> Emmanuel.
>
>
>
Other related posts:
