3 ms·
Is it really a big deal? How many people are doing encoding on the device and can't spare a core for it? I'm sure they exist, but the move from 4->5 will leave
by dgacmu 3y ago
Is it really a big deal? How many people are doing encoding on the device and can't spare a core for it? I'm sure they exist, but the move from 4->5 will leave them with plenty of headroom, so it's not clear this changes anything, even if it makes 5 potentially slightly less attractive for some.
(I say this using a cm4 for image acquisition commercially. Because of the way our pipeline works, I skip the hardware encoder anyway - first, it's a complete pain in the ass to use, and second, it can't handle full-resolution images from a pi camera. We're slower than video because of the resolution, but still capture between 1fps and 1/10fps depending on configuration, with enough headroom to do some on-device image classification using tflite, rotation and JPEG compression, etc.)
- msh 3y agoPeople transcode video in Plex or jellyfin
- dgacmu 3y agoSure, but it kind of sounds like the 5 will still be a superior solution, even without hardware support: https://www.raspberrypi.com/news/introducing-raspberry-pi-5/#comment-1594055 https://www.raspberrypi.com/news/introducing-raspberry-pi-5/... Obviously, it would need benchmarks, but taking that post at its word, the 5 can transcode twice as fast in software as the 4b could do in hardware. That doesn't mean some other hardware platform wouldn't be drastically better, I just don't get the extreme reaction about the 5 when compared to the 4.