Post #2593822

2026-05-11 13:50 UTC

I'm honestly quite disappointed with /r/vtubertech. I expected better than this from the VTuber community. I get that this is an *extremely* widespread myth, but people are getting irrationally angry at the idea that maybe Apple just has a really good face tracking model and it's not the camera. Literally all they have to do is just... cover up the depth camera on an iPhone and see how the tracking still works practically the same, with just the front selfie cam. But oh well, I guess I'll just write and release "VBridger for macOS except it's using the webcam" instead 😇

Replies (4)

@becomethewaifu@tech.lgbt 2026-05-11 14:03

@lina@vt.social I wonder if the 'raw' depth camera is even accessible to regular software? If it works in NIR, I'm half tempted to play around with it simply for that... (In MacOS that is. I'm not allowed to put Asahi on my work mac for obvious reasons, and my only "personal mac" is an old Core2-era mini I use for firewire DVcam stuff)

Open ##2593826
@qgustavor@urusai.social 2026-05-11 15:16

@lina@vt.social Maybe that's just outdated info from when they didn't have a good RGB-only model and had to rely on the depth camera. This reminds of "the dead of the power user" blog post I just read: people are not doing research, they are not trying to understand how things work anymore.

Open ##2593829
@DenJohn@mas.to 2026-05-11 15:39

@lina@vt.social yeah, everyone acts like apple has some unmatched hw tech for facial recognition. Meanwhile it's basically 1st gen kinect(360).

Open ##2593830
@vineyardsiren@mastodon.social 2026-05-13 05:42

@lina@vt.social id love to see if this could be the start of better webcam based tracking in general

Open ##2593831