I thought this news might be of interest to the readers of this forum. There is a press release on this web site:
http://www.21stcentury3d.com/3dvx.html
http://www.21stcentury3d.com/3dvx.html
3D camera based on DVX100A
- Thread starter 21c3d
- Start date
Re: 3D camera based on DVX100A
The 3D footage can be viewed in a variety of ways. Our preferred method is dual progressive scan genlocked DVD players with dual DLP projectors and polarized glasses, but many other systems can work. There is no 24p right now becuase as suspected it has proved to be too difficult to syncronize (so far). I happen to believe that the progressive scan contributes more to the "cinema" look than the lower frame rate. Why throw away 20% of the temporal data? People like 24p however, and it is always good to have options, so we are exploring ways to integrate that feature. Stay tuned!
The 3D footage can be viewed in a variety of ways. Our preferred method is dual progressive scan genlocked DVD players with dual DLP projectors and polarized glasses, but many other systems can work. There is no 24p right now becuase as suspected it has proved to be too difficult to syncronize (so far). I happen to believe that the progressive scan contributes more to the "cinema" look than the lower frame rate. Why throw away 20% of the temporal data? People like 24p however, and it is always good to have options, so we are exploring ways to integrate that feature. Stay tuned!
Re: 3D camera based on DVX100A
Your post opens up a really broad discussion that is probably beyond the scope of this particular forum and the interest of most readers here, but let's just say that there is much more behind the lens of a human eye than there is behind the lens of a camera. Historically, many 3D cameras have done what you describe, but there are a number of problems that this technique introduces and it is not the method by which we make our 3D images.
Your post opens up a really broad discussion that is probably beyond the scope of this particular forum and the interest of most readers here, but let's just say that there is much more behind the lens of a human eye than there is behind the lens of a camera. Historically, many 3D cameras have done what you describe, but there are a number of problems that this technique introduces and it is not the method by which we make our 3D images.
XCheck
Veteran
Re: 3D camera based on DVX100A
[quote author=21c3d link=board=nws;num=1102610892;start=0#6 date=12/09/04 at 13:32:47]...let's just say that there is much more behind the lens of a human eye than there is behind the lens of a camera...[/quote]
Not sure what's behind mine... (eye, I mean). Whatever it is, sometimes it hurts terribly.
Okay, so the answer is NO, and that's fine, too. I was just curious.
[quote author=21c3d link=board=nws;num=1102610892;start=0#6 date=12/09/04 at 13:32:47]...let's just say that there is much more behind the lens of a human eye than there is behind the lens of a camera...[/quote]
Not sure what's behind mine... (eye, I mean). Whatever it is, sometimes it hurts terribly.
Okay, so the answer is NO, and that's fine, too. I was just curious.
Re: 3D camera based on DVX100A
From the pictures of the rig, it appears that the interocular is a bit wide for converging anything closer than about 6-8 feet,
but it looks like the stuff you're shooting doesn't require close up convergence.
As you indicated, there are problems with attempting to "track convergence" while shooting by adjusting the lens axes, especially background fringing at wide angles when creating anaglyph 3D. Some purists advocate only parallel cameras; others want variable convergence.
You've done an admirable job with the syncing of cameras and also of the playback devices to create a 3D experience.
Congratulations!
Ken
From the pictures of the rig, it appears that the interocular is a bit wide for converging anything closer than about 6-8 feet,
but it looks like the stuff you're shooting doesn't require close up convergence.
As you indicated, there are problems with attempting to "track convergence" while shooting by adjusting the lens axes, especially background fringing at wide angles when creating anaglyph 3D. Some purists advocate only parallel cameras; others want variable convergence.
You've done an admirable job with the syncing of cameras and also of the playback devices to create a 3D experience.
Congratulations!
Ken
Re: 3D camera based on DVX100A
Ken,
Thanks for the kind words. You are right the IPD is a bit wide right now. We are hard at work on a second rev of the camera that will reduce the IPD to 3.5", still a bit wide, but better. Also, it would be possible to toe in converge the cameras if absolutle necessary. Right now, an ultra-light weight aluminum bar holds the two together. I have an older mount which does allow for toe in and the camera could certainly be mounted on that. I prefer the current mount. We are able to shoot objects closer than 6 - 8 feet. We control parallax with axial offset in post. As you said, purists prefer parallel cameras, and we have come to find this to be the best way to go. In a few weeks we will have more images and videos on our website. Please check back for 3D samples.
Ken,
Thanks for the kind words. You are right the IPD is a bit wide right now. We are hard at work on a second rev of the camera that will reduce the IPD to 3.5", still a bit wide, but better. Also, it would be possible to toe in converge the cameras if absolutle necessary. Right now, an ultra-light weight aluminum bar holds the two together. I have an older mount which does allow for toe in and the camera could certainly be mounted on that. I prefer the current mount. We are able to shoot objects closer than 6 - 8 feet. We control parallax with axial offset in post. As you said, purists prefer parallel cameras, and we have come to find this to be the best way to go. In a few weeks we will have more images and videos on our website. Please check back for 3D samples.
Re: 3D camera based on DVX100A
[quote author=21c3d link=board=nws;num=1102610892;start=0#6 date=12/09/04 at 13:32:47]a really broad discussion that is probably beyond the scope of this particular forum and the interest of most readers here[/quote]
This thread is preposterous. Are you suggesting some sort of "third" dimension?
[quote author=21c3d link=board=nws;num=1102610892;start=0#6 date=12/09/04 at 13:32:47]a really broad discussion that is probably beyond the scope of this particular forum and the interest of most readers here[/quote]
This thread is preposterous. Are you suggesting some sort of "third" dimension?
Re: 3D camera based on DVX100A
Hi,
Sorry I missed a few of those responses, I certainly didn't mean to imply that anyone here is dumb. If you guys want to hear a long winded explaination of why I don't opt for toe in convergence I can provide it. I didn't mean to sound smug, but obviously the way our eyes see images and the way the brain interprets them is substantially different than the way a camera operates. Many 3D cameras in the past have attempted to emulate the mechanical operation of the eyes, converging on close objects and even actively converging on objects as the move around the scene. This is of course what your eyes do, but keep in mind that there is only one audience member at the movie theater in your mind. A person can really only gaze upon a single point at any given moment. When a 3D movie is viewed, there is no way to know where any particular audience member might be looking at any given moment. Many stereographers assume that the "object of interest" will be the only thing viewed, but clearly this is an inaccurate assumption. The viewer is free to look at anything on screen at any time. If you converge cameras on very close objects you introduce the very likely possibilty of creating a situation in which background objects are unviewable. This can cause many of the famous 3D headaches we've all heard about. The other problem is toe in convergence creates keystone distortions in the images much in the same way a video projector does when not pointed squarely at the screen. Lastly, and perhaps most importantly, a 3D movie is an illusion, not an entirely accurate simulation of human visual perception. It is therefore important to understand the methods by which that illusion operates and its limitations. When viewing real world objects, our eyes accomodate (focus) and converge (target) on the object we are viewing. When viewing objects in a simulated 3D environment like a 3D movie, our eyes focus on a screen that is at a fixed distance. They converge on objects at various distances depending on what is happening on screen. This is not normal for the visual system and must be repsected. Most people with normal healthy eyesight will have no problem with this abnormal situation, but pushing this can also cause eyestrain. Parallel cameras minimize these distortions and seem to be a better way to create 3D images.
Hi,
Sorry I missed a few of those responses, I certainly didn't mean to imply that anyone here is dumb. If you guys want to hear a long winded explaination of why I don't opt for toe in convergence I can provide it. I didn't mean to sound smug, but obviously the way our eyes see images and the way the brain interprets them is substantially different than the way a camera operates. Many 3D cameras in the past have attempted to emulate the mechanical operation of the eyes, converging on close objects and even actively converging on objects as the move around the scene. This is of course what your eyes do, but keep in mind that there is only one audience member at the movie theater in your mind. A person can really only gaze upon a single point at any given moment. When a 3D movie is viewed, there is no way to know where any particular audience member might be looking at any given moment. Many stereographers assume that the "object of interest" will be the only thing viewed, but clearly this is an inaccurate assumption. The viewer is free to look at anything on screen at any time. If you converge cameras on very close objects you introduce the very likely possibilty of creating a situation in which background objects are unviewable. This can cause many of the famous 3D headaches we've all heard about. The other problem is toe in convergence creates keystone distortions in the images much in the same way a video projector does when not pointed squarely at the screen. Lastly, and perhaps most importantly, a 3D movie is an illusion, not an entirely accurate simulation of human visual perception. It is therefore important to understand the methods by which that illusion operates and its limitations. When viewing real world objects, our eyes accomodate (focus) and converge (target) on the object we are viewing. When viewing objects in a simulated 3D environment like a 3D movie, our eyes focus on a screen that is at a fixed distance. They converge on objects at various distances depending on what is happening on screen. This is not normal for the visual system and must be repsected. Most people with normal healthy eyesight will have no problem with this abnormal situation, but pushing this can also cause eyestrain. Parallel cameras minimize these distortions and seem to be a better way to create 3D images.
Re: 3D camera based on DVX100A
[quote author=disjecta link=board=nws;num=1102610892;start=0#13 date=12/10/04 at 09:08:55]I have to say, when I read the article, I thought this was a joke...interesting but very nichy...[/quote]
Maybe I need to change the way it is worded. What would about it made you think it was a joke. I appreciate the input and I want to make sure people don't get that impression.
[quote author=disjecta link=board=nws;num=1102610892;start=0#13 date=12/10/04 at 09:08:55]I have to say, when I read the article, I thought this was a joke...interesting but very nichy...[/quote]
Maybe I need to change the way it is worded. What would about it made you think it was a joke. I appreciate the input and I want to make sure people don't get that impression.
Re: 3D camera based on DVX100A
[quote author=Jerry XCheck link=board=nws;num=1102610892;start=0#3 date=12/09/04 at 11:11:09]I'd suspect it's harder to synchronize a 2:3 or 2:3:3:2 pulldown. But I am just guessing...[/quote]
We've just had a breakthrough that I believe will allow us to do 24p in 3D. I will keep you posted.
[quote author=Jerry XCheck link=board=nws;num=1102610892;start=0#3 date=12/09/04 at 11:11:09]I'd suspect it's harder to synchronize a 2:3 or 2:3:3:2 pulldown. But I am just guessing...[/quote]
We've just had a breakthrough that I believe will allow us to do 24p in 3D. I will keep you posted.
Rich Lee
I shoot Pirates.
Re: 3D camera based on DVX100A
[quote author=disjecta link=board=nws;num=1102610892;start=15#17 date=12/10/04 at 11:08:08]I'd probably validate it by mentioning the DVX right up front. It's not that it would come across as a joke to people not associated with this board but after Rich Lee's joke about the DVX200, I was skeptical[/quote]
is the end product of this parallel dvx 3d setup similer to the 3d lens that canon makes for xl1/2?
also...how is it recoreded to tape? are you combining the images somehow? or does each cam record to tape...and you capture 2 tapes and sync them up somehow later?
anyway, i have messed around with stereo images for a bit, and find them interesting to work with and view....
heres a little test i did with a 3d app...right click and save target as....you will need those red/blue glasses to view...
http://www.rich-lee.com/temp/StereoStudy.mov
also, where can we get some sample video from your setup?
[quote author=disjecta link=board=nws;num=1102610892;start=15#17 date=12/10/04 at 11:08:08]I'd probably validate it by mentioning the DVX right up front. It's not that it would come across as a joke to people not associated with this board but after Rich Lee's joke about the DVX200, I was skeptical
[/quote]is the end product of this parallel dvx 3d setup similer to the 3d lens that canon makes for xl1/2?
also...how is it recoreded to tape? are you combining the images somehow? or does each cam record to tape...and you capture 2 tapes and sync them up somehow later?
anyway, i have messed around with stereo images for a bit, and find them interesting to work with and view....
heres a little test i did with a 3d app...right click and save target as....you will need those red/blue glasses to view...
http://www.rich-lee.com/temp/StereoStudy.mov
also, where can we get some sample video from your setup?
Re: 3D camera based on DVX100A
I acutally tried this myself.
I acutally tried this myself.