I'm awfully new to this, so bear with me if this is a dumb question, but ...
I'm in the market for a video capture device so I can use my DSLR for some live streaming video. In investigating external (USB) video capture devices, I found that some of the more advanced models have built-in encoding (e.g. H.264), while others (e.g. Elgato's Cam Link 4K) don't do any on-device encoding.
For capturing directly to storage, I imagine the external hardware-specific solution is great as the bitrate of the captured stream can be maximized (or at least held constant) independent of the controlling computer.
For live streams, however, where the video signal will be modified and muxed by/with some other sources using OBS (e.g. additional on-screen visual displays or effects), my (perhaps naïve) assumption is that any such live alterations require raw frames anyhow, and thus using external encoding is a waste as the signal with have to be de-coded by OBS anyhow, then re-coded for the eventual output signal (e.g. H.264 as part of an HLS or MPEG-DASH live stream). This assumption is informed mostly by how ffmpeg works, where all alterations take place on demuxed & decoded frames (which then get encoded and muxed).
Is my reasoning above even remotely close to being accurate? Is there any advantage to a video capture device with its own lossy encoding when mixing signals prior to forwarding that (now mixed) video to another destination?