Need help with Point Grey Camera and NI DAQ Card control

[Tony]

Hi, I am a new user of Bonsai. I've gone through the examples come with it, but still have no idea about how to start. My experiment is as follows: I am using two Point Grey Flea3 USB3.0 cameras to record behavior of a mouse getting a laser stimulation. There is also an analogue input channel from an NI DAQ card for recording the laser pulses. The cameras and the analogue input channel are synchronized with a digital output channel of the NI card. I've attached an image to show the flow of the experiment. For now I am using the Matlab to control the NI DAQ card and would like to use Bonsai for the video recording and some marker tracking analysis or use Bonsai to take control of the whole process. The experiment is basically like a for loop, but I don't know how to do it with Bonsai. Can someone send me a workflow to start with? Thank you so much!

[goncaloclopes]

Hi Tony and welcome to the forums!

Are you using the PointGrey cameras in triggered mode? That is, are the outputs from NI-DAQ set to trigger when each frame from the cameras is captured? If so, then a simple Bonsai workflow to record the data would be the following:

This basically sends every frame from the camera to to a movie file and all the GPIO pin data (the sync pulses) to a text file. You can copy paste the nodes for the second camera. Don't forget to tick the checkbox "GPIO Pin State" under Advanced Camera Settings (see also this thread).

I'm attaching the workflow as well. This should get the raw data out successfully.

However, from your description I get the impression you have some kind of trial structure as well, aligned to the laser stimulation. Do you want independent videos to be created aligned on these stimulation pulses? One way to do this would be to use the SelectMany operator as explained in the Bonsai paper Fig 3E.

This basically checks for a rising edge of one of the GPIO pins to start a new trial and chops up a new video at that frame.

Let me know which scenarios are you interested in an we can work out more details.

Best,

[Tony]

Hi Goncalo,
Thank you so much for your quick reply. The experiment flow I proposed is the second scenario. I tested the workflow you attached today. It works well with small number of triggers like 30 or less, but when the trigger number increases, the program starts to miss triggers. For example, for 160 triggers, it only saves ~153 videos. This maybe because the frame rate is high (100 Hz) and the camera is continuously running and logging without free intervals. One solution I can imagine is to introduce an interval allowing the logging process to finish. For example, when rising edge detected in the camera GPIO pin, the camera starts to record 3.5 s but only the first 2.5 s data is logged into disk. I am not sure whether this can solve the problem, but is this possible in Bonsai? Thanks.

[goncaloclopes]

Hi Tony,

It should work with an arbitrary number of triggers, so this may be indicative of a problem. When you say a "large number of triggers" do you mean larger frequency or simply total count (i.e. 160 triggers at 1 Hz or 160 Hz triggers)? The logging process should already finish by its own as long as you don't shutdown Bonsai prematurely. Every frame goes to its respective write buffer in parallel.

Best,

[goncaloclopes]

Another explanation for the "missing" videos could be that some of the interval boundaries had no frames.

Videos can only be created if there is at least one frame. If your inter-trigger interval is fast enough that no frame arrived from the camera in that period of time, no video would be created for that interval. Again, the critical bit of information is knowing how frequently are you sending your trigger.

Hope this helps,

[Tony]

Hi Goncalo,

I found the source of the problem which is I checked the 'Enable/disable trigger' in the trigger control panel of Flycapture2. When I unchecked it, everything works beautifully.

Thank you so much for your help!

Tony