| Printable Version of Topic
Click here to view this topic in its original format |
| Unofficial VirtualDub Support Forums > Advanced Video Processing > 3d Video Question |
| Posted by: rjisinspired Mar 16 2011, 10:57 PM |
| I started playing with some 3D techniques. In one example you lay a second copy of a video track underneath the first video track then on the first video track you apply two levels filters for the green and blue channel then on the second video track you apply a red levels channel onto it. Going back to the first video track to change the composting to "screen" then you go into the pan and crop from the top video and nudge the video position about 8-10 nudged to the right to offset the video a little. If you nudge to the left you get a sinking effect though I have noticed that the opposite positioning is true sometimes? The technique recommends having the red portions to the left to make things pop out but sometimes the reverse does this? The effect seems ok but there are some problems I have noticed. 1. The effect isn't as eye popping as it could be and... 2. HD doesn't seem to work well with 3D because when you put the 3D glasses on and you hit the really hot colors my eyes start to go crazy. You got the reds and pinks being filtered in from the left eye, red allowing it's same color in, and partially blocked from the right eye and there's that conflict which makes a scene difficult to watch. Does anyone have a guide on how to get decent 3D videos? I am using only one camera because I only have one but from examples I have seen, 3D can be done with only one camera. I guess maybe contrast and other adjustments are needed to pull this off? |
| Posted by: evropej Mar 17 2011, 05:29 PM |
| My experience with 3d at the PC level has not been good. The stress on the eyes is to great. Single camera with two lenses =] I believe you need 4 inches or so of separation to create the effect of eye separation. |
| Posted by: rjisinspired Mar 17 2011, 06:49 PM |
| See this is what I don't understand: I keep hearing you "need" two cameras but what difference does it really make having video from two cameras as opposed to just using a copy from one camera? Shifting one of the two videos and of the perspective in post edit, to me, is the exact same thing regardless of which method is used. In both methods you working with two videos, one of them is offset but the same is being done if you import from only one camera, copy the first video track, offset it a little in both field and perspective through pan and crop; viola, same thing. What is so different? I have been looking around and I only found one poster saying that you are only getting one level of 3D using the one camera method and just duplicating the first video. What does that mean? An offset is an offset, lol. |
| Posted by: dloneranger Mar 17 2011, 07:15 PM |
| Well, imagine a simple clip Empty room, and two people enter at the same time One from the left and one from the right They both walk at the same speed to the opposite side of the room Your eye's are seeing the two people at two different angles, giving you the sense of perspective But the left and right eye are seeing different images and angles for both people With only 1 image, that's just moved over a bit, both eyes are seeing with the same angle It doesn't give the correct perspective effect It's a bit like sitting 6 foot to the right of the tv and looking at it by just moving your eyes Or Imagine looking into an empty box. Move the box until your left eye can see the inside, but the right eye can only see the outside With two camera's you'd get the same effect But with only one camera, you'll have the effect of both eyes either being able to only see the inside or the outside You're brain knows that this is 'wrong' and can even give some people eyestrain/headaches If I can only see the outside then I should be further to the side, if I can see the inside, then I should be more to the centre etc |