What do you mean by "video mesh"?
He's probably referring to some experimental tech to extract geometry from moving pictures or photos, which so far is neither in any way related to Adobe or Autodesk and is based on indepenedent research by other companies or universities.
"..experimental tech to extract geometry from moving pictures or photos, which so far is neither in any way related to Adobe or Autodesk.."
- your not even half right:
goto http://www.ulozto.net/xBEeM58/adobe-max-video-meshes-mp4, and you'll see it was debuted a an "ADOBE" Max, Adobe Systems conference,
and if you Google: "video mesh""CS6" you'll see it was recently schedualed for release with both After Effects and Premiere CS6, ie: "..I for one cannot wait to upgrade my Production Premium to CS6, if only for the new “video mesh” feature previewed at Adobe Max 2011. With this new tool, after adding some roto-scoping and meshes to the video footage, an editor can “look around” inside the video in 3D. That gives a whole new meaning to “we can just fix it in post-..”
Adobe support suggested, and the aforementioned is why i asked here since development staff monitor and answer questions, but i'll have to ask elsewhere Dev. staff are privy to latest research and more complete breadth of ADOBE products.
But any informed answer from this forum would be appreciated,
Never ever. Sorry, but you are falling for speculation and gossip. Whenever Adobe shows such tech demos they never make any promises that it will ever be a product - be it just for legal reasons. If ever, this may show up in the next big release of CC which you can expect in May after first demos being shown on NAB in April - just like every year. Again, you are operating on wrong assumptions - a lot of them, beginning with the one that a tech demo would actually be a usable product.
But you see now it is directly related and self associated with Adobe.
So, possibly in next CC or never, your right, sounds liike a typical Adobe debut, excite and forget commitment. But did you watch the demo?, and if so what about it suggests an unusable product?
Well, lets begin with this not being any plug-in to one of their programs and running as a separate app with a debug terminal open. Second, quality issues, quality issues, quality issues. Third, "test footage" that of course works splendidly in the demo, but says nothing about realworld use cases. It's built to impress kids and has a long way to go before becoming a product that everyone can use. Fourth, let's be honest here: Without AE getting a proper native 3D space what would be the point attempting to reconstruct a 3D scene this way? As long as you break your fingers with AE's camera navigation and the programs less than glorious 3D engine, serious use cases like plate matching, set extensions, scene relighting and what have you will be a major pain. And finally, fifth, even if it once comes to be, it doesn't eliminate a ton of otehr steps liek carefully planning your shoot, possibly using multiple cameras and so on. So for what it's worth, and you can call me an eternal pessimist, but I think AE has other issues to resolve long befor this. Not saying that it couldn't be useful, but I'd take stuff that could be useful on a broader level like better masking over this any time, if the decision depended on which feature the devs should focus their not infinite resources. And make no mistake: You are going to need good masking even with this when it can't figure out obscurations properly in the automatic algorithm.
I highly doubt Adobes' target demographic was "kids", more likely 2to3D conversion artists amongst many other interested professionals, but i agree that masks or Roto's as the demonstrator showed are still a necessity. As to the other problems in this eventual development, the benefits of VMesh capabilities obviously outweight any of the typical implementation difficulties of such high demand add ons.
You must be new to AE. If you think these things happen overnight, don't hold your breath.
Think of the presentation you saw as as a concept car being show by a car maker. Then think of the last time you saw the concept car's features incorporated in the next model year.
Agree with Dave - you sound like one of those tech kids yourself. In the grander scheme of things this is just one solution to a very specific problem that may only affect 10% of all users. Stereo 3D wouldn't even be its primary application, since that has by far other requirements and is not about creating a "real 3D world" in the first place - many Stereo 3D effects could never happen in a physical world and are just used as a means of artistic expression, but are otehrwise painstakingly constructed from multi-layering planes in compositing or 3D renderings. And contrary to what you may believe I don't expect to see 3D-flyarounds becoming a mainstream thing any time sone, with the main limitation being that most people will not have access to a suitable set of cameras to reconstruct a whole scene. I see this more as a way of doing subtle corrections like shifting the viewing ungle by just that half meter because the camera could not pe placed this way on set or you could use the technology to gnerate depth and motion based mattes for further processing. in any case, no matter how smart and great it will be, it wil lrequire manual labor, planning and meticulous execution just as much has hitting that "make magic" button in the plug-in/ effect.
... and if you really, Really, REALLY want to see such a feature happen, I encourage you to file a Feature Request.
Feature requests work a lot like the old joke about voting in Chicago: do it early and often. Adobe appears to work off the principle that the squeaky wheel gets the grease.
Just back, points taken,
There is a lot of theoretical stuff that is experimented with all the time. Some of it makes it into software that will actually do more than demonstrate a principal. There are amazing things that can be done to recreate a 3D world from footage taken from a moving camera if the camera moves in the right way and everything is perfect.
I clearly remember that demo and that's all it was. Just a demo of some interesting technology that may be useful in a production pipeline. Funny thing is that you can do everything that was demonstrated in that video in AE right now. It just requires some basic 2.5D layering techniques and something that is sometimes called projection mapping or camera mapping. You could have done it using the same techniques when 3D was first introduced in AE, you could have even done it before AE had 3D layers, it just would have taken a different technique and involved a lot of hand work.
In the mean time you can composite any effect that you have seen in ANY movie or commercial using After Effects. You may have to create some of your building blocks with other applications or with a camera, or a laser scanner, or motion capture equipment, but waiting for a miracle plug-in to make your work stand out is going to take a very long time.
Just back, points taken, CORRECTIONS: except for the ones you put words in my mouth IE:
"..You must be new to AE. If you think these things happen overnight,."
"..Stereo 3D wouldn't even be its primary application,."
"..And contrary to what you may believe I don't expect to see 3D-flyarounds becoming a mainstream thing any time sone,."
- i said None of these things. Plus i appreciate but am more than familiar with your stated conversion methods in addition to:
- auto/manual depth map generation,
- semi-automatic rotoscope dimensionizing procedures/software,
- warp 2d to 3d,
- symetric morphing to 3d views,
- photogrammetric point cloud extraction with surface generation,
..the list is still growing. All of these are effective 2to3d methods that vary greatly in time and effort required, but Video Meshes from pro converters perspective "apparently!" represent a significant leap forward in speed of process and great reduction (NOT ellimination) in required manual labour. And this is what appeals to my industry peers and any pro's whom could benefit from a "comparatively!" instantaneous 2to3d environment.
But all i really asked is When if! known and thats been answered re: next CC's or never.
Yeah, I hear you: that stuff would be great. It would make everyone's lives a lot easier. We'd be able to concentrate more on composition, visual communication and artistry and a lot less on technique. That's a teriffic thought.
But as you can see, no one else in the thread is turning handsprings over the prospect of it showing up any time soon. Where your reaction to that demo seems to be, "WOW! When can I get my hands on that?" we're thinking, "Uh-huh, that's nice."
To summarize the feedback you've gotten, I'd say it would be this: take that video with a grain of salt, and don't get your hopes up.
Thats also my take on the average of feedback at this forum, plus i obviously wish VM was anywhere to try and varify wether my hopes are dashed or substantiated.