Posts
Cameron
Would it be possible to make a 3d scanned video, not just a picture? Where you can see all 360 degrees of someone for example dancing. Like could you get a ton of camera's and put them in all different directions and then be able to the footage together. Does this technology exist?
Thanks
Answer
They already have this technology to film in this way, it is no different than the technology to do the stop motion rotation. The problem is we do not have the technology to actively use anything generated like this. The problem is this...
36 cameras at 10 deg increments will give you a 360 view of the person, so what this means is you would need a monitor which has 18 double sided monitors at 10 deg increments around a central shaft (or 36 screens total) with 36 separate inputs just to view this video in a fixed location monitor, these do exist by the way, and you would see the video output based upon where you are located around the monitor.
The more cameras and monitors you have set up like this the more it will appear to be 3D.
Now for the problem of your idea. You are wanting to generate a real time 3D video out of the footage, this is physically impossible with our technology level, we can't even generate a true 3D image in the first place, much less a video of it for play back. People who work with 3D graphics software and play 3D games have seen what you are wanting to do, this is because they are working with 3D objects being generated on the fly, but when they output this information it is 2D video, there is no way to output a real time 3D object which can be viewed from any angle at the viewers request, the data generated in this manner would be insane to try to have a permanent copy generated for reviewing, when this is done in games and software once it has been viewed it is lost.
They already have this technology to film in this way, it is no different than the technology to do the stop motion rotation. The problem is we do not have the technology to actively use anything generated like this. The problem is this...
36 cameras at 10 deg increments will give you a 360 view of the person, so what this means is you would need a monitor which has 18 double sided monitors at 10 deg increments around a central shaft (or 36 screens total) with 36 separate inputs just to view this video in a fixed location monitor, these do exist by the way, and you would see the video output based upon where you are located around the monitor.
The more cameras and monitors you have set up like this the more it will appear to be 3D.
Now for the problem of your idea. You are wanting to generate a real time 3D video out of the footage, this is physically impossible with our technology level, we can't even generate a true 3D image in the first place, much less a video of it for play back. People who work with 3D graphics software and play 3D games have seen what you are wanting to do, this is because they are working with 3D objects being generated on the fly, but when they output this information it is 2D video, there is no way to output a real time 3D object which can be viewed from any angle at the viewers request, the data generated in this manner would be insane to try to have a permanent copy generated for reviewing, when this is done in games and software once it has been viewed it is lost.
Why don't video games use 3D scans of real objects?
jackalante
Wouldn't that make the object look just like real life? I know (Since you don't actually see objects in video games look just like real life) that there has to be some type of limitation as to why this can't happen, so what is it?
Answer
Games are bound to limitations (specifically polygon count) 3d scans are very sloppy in topology and high resolution in poly count. They may use a 3d scan if they really want to but it would need to be re-topologiezed and cleaned up before it would be of any use. This is a great deal of work and really takes away the benifit of a scan vs simply using photographs for textures and having an experienced 3d modeler make a perfect model from scratch that won't have issues that 3d scanned models will.
Games are bound to limitations (specifically polygon count) 3d scans are very sloppy in topology and high resolution in poly count. They may use a 3d scan if they really want to but it would need to be re-topologiezed and cleaned up before it would be of any use. This is a great deal of work and really takes away the benifit of a scan vs simply using photographs for textures and having an experienced 3d modeler make a perfect model from scratch that won't have issues that 3d scanned models will.
Powered by Yahoo! Answers
Title Post: Would it be possible to make a 3d scanned video?
Rating: 100% based on 99998 ratings. 5 user reviews.
Author: Yukie
Thanks For Coming To My Blog
Rating: 100% based on 99998 ratings. 5 user reviews.
Author: Yukie
Thanks For Coming To My Blog
No comments:
Post a Comment