You can see the end of the handle there. The rest of the stick is automatically removed by the camera software.
![]() You can see the end of the handle there. The rest of the stick is automatically removed by the camera software. Regards, -scott Welcome to Rivendell, Mr. Anderson. |
|
![]() 360 cameras capture all the way around and you select POV in post to generate the final. -- Drew |
|
![]() But that's much easier, because they don't edit in a realtime pic of themselves. They only use a pre-built 3-D model of themselves; possibly with slight variations for doors open etc. And I suppose if you have it re-sprayed in another colour it'll still show the factory-original hue on-screen. Check any (semi-)recent video of a Mercedes S-class,BMW 7-series, etc by Doug DeMuro or some other such car YouTuber. Yeah, no way was I suggesting he had that kind of processing power on top of his helmet. If it were done that way (seems less likely now that people are pointing out the selfie stick), it would have to be post-processing. But this was no super-amateur video; I assumed all more or less professional YouTubers do quite a lot of that. -- Christian R. Conrad The Man Who Apparently Still Knows Fucking Everything Mail: Same username as at the top left of this post, at iki.fi |
|
![]() The hardware is dead simple to use, and the software is more tedious than difficult. Just stumbled across one where you can see the shadow from it. https://youtube.com/shorts/axY472ZdUSY -- Drew |
|
![]() ...off-the-shelf for YouTuber hat-cams in what, three-four years? I saw this review of, think it was an S-class, where DeMuro (or whoever) fiddled around while standing still: Pinch-to-zoom, swipe to pan, etc etc, on the touchscreen, and he got a made-up perspective from some point of view where he wasn't and no cameras were. "OK, let's fly up and look down on ourselves", like. Can't recall if he demonstrated or just said that "On the freeway, we can 'drive behind ourselves and look at ourselves from the rear' -- or from in front or any side." In that courtyard demo, it was as if the camera was on a drone, buzzing around the car. OK, so I was apparently (or at least most likely) wrong in thinking that had trickled down to YouTubers' hat-cameras already, but for all I can see, it will. Soon. (Wow, isn't the IT industry great? We're really contributing to making the world a better place! Soon narcissistic exhibitionists in rich countries won't have to schlep around selfie-sticks any more! What, me cynical?!? [For further reading: Search "bullshit jobs" in any search engine of your choice.]) -- Christian R. Conrad The Man Who Apparently Still Knows Fucking Everything Mail: Same username as at the top left of this post, at iki.fi |
|
![]() Cars are easy. You know what they look like from the outside. The selfie-stick footage shows faces and bodies in real time. There's no software yet doing real time generation of human movement. -- Drew |
|
![]() Changing perspective on the movement of limbs seen from on top of someone's head into seeing them from in front or behind is a big enough ask, but I suppose it can be done... But from up there, the camera doesn't see his face at all. Duh! -- Christian R. Conrad The Man Who Apparently Still Knows Fucking Everything Mail: Same username as at the top left of this post, at iki.fi |
|
![]() Stick it out on an arm like a bike mirror except have it within the camera's view to get the face and body. Four of them with four cameras (or two of them with really wide angle but that would add to the computer stitching so I would use as many cameras and mirrors as possible) and you can get mostly everything. Software should be able to do the rest but that's got to be some serious compute power and algorithmic brilliance as it joins it all together, reverses for front camera, looking back at your face viewpoint, combined with flipping all the traffic and joining it with the other camera viewpoints in the non-body portion and upside down flipping your body and stitching it all together. Hey Drook, it's a gleam in my eye. Nah, that's the glassiness from smoking pot. It's a pipe dream. But someone else could do it. I don't actually do anything anymore, I'm more of an idea rat. |
|
![]() It's probably been 10-15 years since NBA teams starting ringing the arena with still cameras. When it looked like something good was happening, they'd hit the button and they'd all fire at once. Then they could rotate the view and see it from all angles. I immediately imagined putting 4-8 cameras - video, not still - around the court and use software to simulate intermediate views. Picture they can spin a dial to select a location for a virtual camera. -- Drew |
|
![]() ...but for this guy and his "fans" (to the extent he has any? Apparently so), the hat-cam, possibly with mirrors, is the right use case. -- Christian R. Conrad The Man Who Apparently Still Knows Fucking Everything Mail: Same username as at the top left of this post, at iki.fi |
|
![]() -- Drew |
|
![]() |
|
![]() It's software tech that been around for years but only in the last decade has it started being widespread as the hardware gets better and the software gets better (and faster). Wade. |
|
![]() Software never gets faster. Only hardware does, (barely) compensating for software getting slower. -- Christian R. Conrad The Man Who Apparently Still Knows Fucking Everything Mail: Same username as at the top left of this post, at iki.fi |
|
![]() I was thinking of things like better algorithms being developed and implemented. Wade. |