With Bullseye (5.15 kernel), there is supposed to be a “speed boost” for Raspberry Pi 4 owners, I was wondering if there was a plan to provide a new image using this version? I did attempt to upgrade, but this broke Python app (LightburnBridge)
What advantage would that add? It only needs walking speed and it’s running on a rocket now…
It would be nice if they could just package it as a .deb installation. Then we could put it on any linux type build. I have an old mesh server I could have it on and used the pi for stuff that needed the horsepower.
If we’re dreaming…
It is due to fact that I am testing using it with a mpg-streamer installed on the Pi (with a lightburn camera), and I am having occasional failures while sending a file to the laser. It will get near the end, and then suddenly, the laser is busy (which it is not). I end up having to send via USB
You want a camera which does MJPEG or something like that. I had put a camera and streamer on my CNCjs rPi and it was killing performance. Got a better camera which did compression and resource usage went way down for the streamer.
1 Like
interesting thought, I’ll look into that. Do you have any suggestions on a small camera that would fit the bill? I’ll do some looking around as well. Thanks for the idea
I’m not familiar with the specific camera streaming use case but I do know that Raspberry Pi 4 has native hardware encoding of H264. If you’re getting sluggish performance you may want to make sure you’re using a hardware path for encoding.
Hardware Accelerated Video Encoding on the Raspberry Pi 4 on Ubuntu 20.04 64-bit (willusher.io)
It’s been a while but IIRC the setup used FFMPEG for the streaming server and when I saw the process loading the system I googled “ffmpeg excess processing” or something like that and that’s where I found references to using a camera which had M-JPEG support. I’d found a pretty inexpensive one, one which swiveled and the camera was on a “Y” fork and would also tilt(manually) which worked well for aiming at my CNC.
1 Like