5 ms·
Here's one view for the device - FaceTime++ If it's a gamechanger for remote working, then I'll buy them for my company and have my remote employees use them
by tikkun 3y ago
Here's one view for the device - FaceTime++
If it's a gamechanger for remote working, then I'll buy them for my company and have my remote employees use them
If it's a gamechanger for communicating with family overseas, then I might see if I can convince one of my relatives to get one (though the price will be harder to justify for personal use - but not impossible if it's a gamechanger)
Both of those are contingent on how good it is. If it's something where I do a shared presence FaceTime call using it, and I feel like I can't go back to not having it, then I'll buy them and I expect it'll be a huge success. If I/people try it and it's not much of an improvement over regular FaceTime, then it won't be a big success.
- ajmurmann 3y agoI like the idea of the shared presence, but don't see how that doesn't come at the cost of losing the actual video off the person you are talking to. I want to see my family in higher fidelity, not their Mii version or have them look like Zuckerberg in the Meta demos.
- smoldesu 3y agoThe original iPhone was an iPod, a phone and an internet communicator. It was a hard sell at $599, and didn't take off until the price was lowered and the App Store started proliferating. If it had an external battery pack or cost $3,000, nobody would have bought it. If the big killer app for this headset is "calling people", then Mark Zuckerberg is probably doubled-over in hysterics right now.
- bullfightonmars 3y agoThe will have to make presence _really_ good for this to be worthwhile. The closest thing I have seen to next level is Google's Project Starline [1]. I would spend so much money to get one of these for my home and one for my parent's home halfway across the country. 1. https://blog.google/technology/research/project-starline/ https://blog.google/technology/research/project-starline/
- Invictus0 3y agoVideocalling barely works at 1080p because of network issues. I don't know how much bandwidth FaceTime++ would need but I am not optimistic on this usecase at all.
- SXX 3y agoMotion capture processed data and pre-processed 3D models are far less bandwidth-intensive than a video feed. The only problems that can't be really solved by super realistic 3D is latency.
- zmmmmm 3y agoThe rumors that have come out say that Apple has achieved 1:1 3d video chat where there avatars are almost completely photo realistic. I'm very interested to see this especially. Apparently this is a big reason why the device has to bundle 2 X M2 processors and also limits it to one on one calls. I think it's going to be amazing and significantly shift people's perceptions here with how good it is. They should be able to demonstrate hyper realistic pass through as well so essentially you are looking at an almost completely realistic image of the other person teleported into your own surroundings.
- smoldesu 3y agoIt looks like Facebook has been playing with the idea internally for a while too: https://twitter.com/AiBreakfast/status/1652121065194409985 https://twitter.com/AiBreakfast/status/1652121065194409985
- tikkun 3y agoCould you please send a link on the rumors? Big if true!
- ben_w 3y agoThis isn't particularly surprising if true, as I've seen several demos of similar on Two Minute Papers, and even before them considered a 3D video chat system which just sent depth info from the depth sensor along with the normal RGB info. I even made a proof-of-concept of my own at some point in… I think 2018?… which was just recording locally and playing back with Apple's own AR libraries. Honestly, it was so easy I'm surprised it isn't already standard the same way custom backgrounds and 3D overlays (hats, face-paint, etc.) already are.