The Pulsar Engineering Community Board

Dear all,

does somebody face similar problem and have solution?

Have flir bfs-pge-31s4m camera which is recognised in Spinnacker Spectrum analyser sees the camera but during connect have error 'Invalid node'
Also the same behaviour in successful fresh x64 Spectrum Analyser build in VS2022 Any help is appreciated

Regards
Oleg

I don't know why it would happen with a fresh build but I do know you have to install an old version of the api (Spinnaker SDK 2.0.0.147) to connect with the provided build.

Could you run the program evemon.exe first, then launch spectrumanalyzer and try to reproduce the bug? It should give you a detailed log of everything happening.

From the name it looks like it's trying to access something that doesn't exist in the camera. I'll check in more depth when I'm in front of my computer today.

Here is the screenshot with attached zipped err file

I confirm that it is coming from a portion of the code that should never trigger (fail safe). For some reason, one of the mandatory parameter of the camera (exposure, gain, ROI...) cannot be accessed.

Before we go further into the debug, I will ask you to perform a few operations and tell me if it's still not working.

• Close all softwares (Spectrum Analyzer, SpinView etc.)
• Change your adapter IP to 192.168.100.1 with subnet mask 255.255.255.0
• Start SpinView and update camera IP; check that you can access camera frames (start acquisition)
• Close SpinView
• Start SpectrumAnalyzer and try if it connects

I noticed that your current IP is in the "invalid DHCP" range. I don't know if that can trigger an issue with the camera so it's better to check that first.

If it still not work, could you tell me if you know how to use the debugger in Visual Studio ? If not we'll do a bit differently (I'll send you a modified dll that display messages)

Hi Luc,

unfortunately it shows the same error despite IP address of the camera in correct DHCP range.
DLL would be preferable option for the troubleshooting
Thanks a lot

Hi soleg,

I'm running a bit out of time these days so I can give you a procedure that involves no coding to already dig a bit without having to wait for me.

It's possible that the nodes names changed during the transition from BFLY series to BFS. I don't know if other people are using the BFS as well but I developed the software for the BFLY.

Long story short, in the camera dll project you will find plenty of lines that initialize the nodes such as FloatProperty followed by strings like "Analog", "Gain". These are the addresses of the nodes inside the camera and should match the names in spinview.

The tricky part is that you should not be looking at the node label in SpinView ("AnalogGain") but click (right-click?) on it to display a full dialog with more information. There you will find the full node access path -- it's how I did it in the first place.

If you could already compare the names that would help! This weekend I will allocate a couple of hours to create a debugging program to track the faulty node.

Hi Luc,

please fined attached file with comparison : most of the nodes have same names but there are also some with different names and missing ones.

Regards
Oleg

p.s. bfly camera was not so easy to get and vendor recommended successor model bfs

Quick update : I was able to find a BFS camera and reproduce the bug. That will ease up the correction process!

I was able to come up with a quickfix. It should be working with BFS cameras now except you won't be able to save wavelength calibration settings. Note that I did not have a BFS-31S4M camera so I made the fix using another BFS model.
Tell me if it works! It would also be worth to send me a capture of the evemon program if you spot silent "invalid nodes" errors.