Hi, hoping someone can help me here. I have an Arducam IMX219 Wide Angle Camera Module for the NVIDIA Jetson Nano (got it via UCTRONICS) and I’ve honestly never gotten it to work. The camera does seem to be detected - the video device shows up when I list the devices in the terminal. But when I run nvgstcapture-1.0 it prints some output, never captures an image, and exits with an error. A couple of times the whole Nano just froze after running that command and rebooted itself.
Setup: Jetson NANO 4GB Developer Kit. I followed the NVIDIA first-picture CSI/USB camera tutorial.
Weird thing is I also bought an Arducam 4K 8MP IMX219 Autofocus USB Camera Module, and that one works fine. Running nvgstcapture-1.0 --camsrc=0 --cap-dev-node=1 pops up the preview and captures no problem. So the USB camera works with the Nano but the CSI camera does not.
I googled some of the error messages but didn’t find anything useful. Any idea what I’m doing wrong?
Hey, thanks for the details. Before we go guessing, can you run our diagnostic bug report tool so we can see what the system actually reports for the CSI camera? Please run these commands in order:
That will drop an arducam-bug-report.txt in the current directory. Please attach that file here so we can take a look. Just to confirm, this is the CSI IMX219 module on the Jetson Nano 4GB dev kit, and the USB camera working via --camsrc=0 --cap-dev-node=1 while the CSI one fails - is that right?
Yep that’s right, CSI camera fails and the USB one is fine. I ran the tool and it produced the arducam-bug-report.txt file - I’ve attached it to this post. Let me know what you find.
Thanks. Looking at it, the I2C communication looks unstable - that would explain why the camera is detected but nvgstcapture-1.0 can’t actually pull frames and sometimes hard-locks the Nano.
A few things to try:
Can you try a different connection cable? The CSI ribbon can cause flaky I2C even if it looks fine.
Please take a picture of the hardware connection (how the camera and cable are seated) and attach it here. It helps to see if the connector or orientation is the issue.
OK, I swapped in a different connection cable and re-ran the bug report tool with it. New arducam-bug-report.txt is attached, plus a couple of pictures of the hardware connection. Still the same behavior with nvgstcapture-1.0 - no capture, and the same errors. Let me know what you see.
Thanks for the new report and the photos. From the picture it looks like you’re using a third-party development/carrier board rather than the official dev kit. Some of our cameras aren’t fully compatible with certain third-party boards, and that can cause exactly this kind of I2C/capture failure. Because of that we can’t tell yet whether the issue is with the camera itself or with the hardware platform.
Can you test the camera on the official Jetson Nano development kit? If it works there, then the problem is the third-party board. If it still doesn’t work on the official board, then it points to the camera, and we’ll take it from there. Please report back with the result.
Alright, I tested it on the official Jetson Nano board today and unfortunately I get the same error messages when I try to capture with nvgstcapture-1.0 - exactly like on the third-party board. So it fails on the official dev kit too. I generated a new bug report file on the official board and attached it, along with fresh photos of the hardware connections. I really can’t get this camera working - what else can I try?
Thanks for testing on the official dev kit - that was the important part. Since the same nvgstcapture-1.0 errors show up on the official Jetson Nano board, the problem isn’t the third-party carrier board. Combined with the unstable I2C reading and the cable swap not helping, this points to a hardware fault with the camera module itself rather than anything in your setup.
I’ve sent you a private message with the next steps to get this sorted out. If you don’t see it, let me know here and I’ll follow up.