1+1=3D Tutorial
|
For the purposes of this demonstration, I'll be describing a method that produces a 400 pixel wide viewing area comprised of two 200-pixel wide views, side by side. 200 pixels is within acceptable parameters for the average human eye (about 2.5 inches) and should be something most people can adapt themselves to, if sufficiently motivated.
|
1+1=3D A new point of view
A new point of view
As you have most likely customized your settings to your liking, it's time to ditch all that and start fresh. Delete any models, resize the ground to default settings for scale continuity and set the background color to black, or any suitably dark hue. Then go to the menu: Window>Document Window Size… |
Set Width to 400 and height to 480. If you select Lock window, the window can still be moved within the Poser environment but not resized. If you select Constrain aspect ratio, you'll be able to resize the window. This allows you to scale the window down, to relieve eyestrain, for example, while maintaining the relative framing of your animation. But don't do that yet. Just be aware that you can adjust the viewing window without changing your movie settings.
OK |
Decide at this time if you want to display camera names in their panes while you work. You may want to, initially, as you become more oriented to the motion of 'flying' stereo cameras. Once you 'get your wings', you'll certainly find all non-image information a distraction and long to be free of it. There are ways and there are others ways but they are all still very far away.
|
1+1=3D Getting the 1st 1 right
Getting the 1st 1 right
On the left side of the window (in version 5 and at the bottom in version 6) are the Window Pane Presets that allow for various multi-angle views of the Poser world. 2nd from the bottom is the vertical splits screen. Select it. Select the right pane with a single click of your mouse. |
Reproduce these settings:
Focal set to 55mm is a recommendation to get started but this is a variable that is worth investigating. Do not forget to match the Focal settings in the left view! |
1+1=3D Getting the 2nd 1 Left
Go to the menu and select the Aux Camera: Display | Camera View | Aux Camera.
In order to appreciate the stereoscopic view once it has been finally set up, there must be something on view to appreciate.
I suggest you start simply and develop your static spatial senses before you start moving the cameras and going nuts flying around animated worlds. Or not. Personally, I wasn't able to restrain myself once I got going. I'm using a simple model for these illustrations. As a stereo view, this should look flat because both cameras are in the same location. |
1+1=3D
In case you've been busy with extra-curricular activities, select the left pane again.
Select the Aux Camera from the menu. Change the DollyX setting to something between -0.010 and -0.040. This moves the Aux Camera to the left and produces a left eye view in the left pane to compliment the right eye view of the Main Camera in the right pane. I've been told that the setting of -0.012 approximates the distance between adult human eyes in the Poser world. If you want to see the world at the same scale as your figure, measure the distance between the eyes of the figure using a third camera centered on one eye and move it whatever distance to the center of the other eye. Note the distance and use that to separate your two views. A larger negative number will exaggerate the apparent distance between views creating the effect of miniaturizing the world. Imagine being a giant in the world whose eyes would naturally be further apart and you may understand the effect without seeing it, though by all means, see it for yourself. Under no circumstances change the values of DollyY, which would have the effect of off-setting one view above or below the other, or DollyZ, which would have the effect of off-setting one view in front or behind the other. Any of that will impair the illusion. |
In this shot, I have repositioned the Main Camera to get a better view of the figure. Viewed in parallel, the figure should be fully stereoscopic and appear to extend in front of the screen.
Though the two views are some distance apart on the X-axis, they both point in parallel which is not the way your eyes behave, as they would normally converge on an actual object you were looking at. The fix is to slightly rotate the Aux Camera yOrbit to a negative number; -2 for an example. This has the effect of turning the left view towards the object. Use the framing of the object in the Main Camera right pane to frame the object in the Aux Camera left pane and the figure will recede in the view. Welcome to real 3D in 3D. It should go without saying, but here I go, the camera can move during your animation and that means 1 + 1 + t = 4D! |
comments on the tutorial from renderosity.com
kuroyume0161 10:48PM Sun, 01 June 2008
Too bad that these two views cannot be combined via OpenGL 3D Stereoscopics to 'flicker-glasses' or a headset so that you can truly see the stereo depth while working!
Too bad that these two views cannot be combined via OpenGL 3D Stereoscopics to 'flicker-glasses' or a headset so that you can truly see the stereo depth while working!
EyeSoare [Michael Woody]11:34PM Sun, 01 June 2008
The complaint escapes me because this functions perfectly. Too bad kuroyume0161 missed the point. Working in Poser I can see a fully stereoscopic view while posing AND while animating WITHOUT any extra technology/glasses/etc. cluttering up the process. The suggestion that OpenGL 3D is somehow more desirable is confusing to me but then I don't generally try to solve problems that aren't problems with solutions that aren't solutions. If anyone has anything to say about the tutorial that relates to the tutorial, I would be interested in that.
The complaint escapes me because this functions perfectly. Too bad kuroyume0161 missed the point. Working in Poser I can see a fully stereoscopic view while posing AND while animating WITHOUT any extra technology/glasses/etc. cluttering up the process. The suggestion that OpenGL 3D is somehow more desirable is confusing to me but then I don't generally try to solve problems that aren't problems with solutions that aren't solutions. If anyone has anything to say about the tutorial that relates to the tutorial, I would be interested in that.
kuroyume0161 4:28PM Tue, 03 June 2008
No, my point stands. I'm rather very well versed in 3D Stereoscopics. If I understand you, you are referring to cross-eyed viewing - I don't remember reading anything about anaglyphs or so-called 3D monitors so that's my assumption. Cross-eyed viewing can lead to headaches, migranes, dizziness, nausea. Not that flicker glasses can't do the same under circumstances (such as low refresh rates) or prolonged use. If that is your intention, you should provide a warning concerning that as some people are prone to these symptoms while others just can't cross their eyes (at all or sufficiently) to bring the images together completely. I can do it only for very short stints. I'm not knocking your idea - I love 3D Stereoscopics - but the currently cheapest and best way to do it is with flicker glasses on a sufficiently high refresh rate CRT monitor. Anaglyph glasses aren't bad but you do lose some color information. Headsets and 3D monitors are usually prohibitively expensive.
No, my point stands. I'm rather very well versed in 3D Stereoscopics. If I understand you, you are referring to cross-eyed viewing - I don't remember reading anything about anaglyphs or so-called 3D monitors so that's my assumption. Cross-eyed viewing can lead to headaches, migranes, dizziness, nausea. Not that flicker glasses can't do the same under circumstances (such as low refresh rates) or prolonged use. If that is your intention, you should provide a warning concerning that as some people are prone to these symptoms while others just can't cross their eyes (at all or sufficiently) to bring the images together completely. I can do it only for very short stints. I'm not knocking your idea - I love 3D Stereoscopics - but the currently cheapest and best way to do it is with flicker glasses on a sufficiently high refresh rate CRT monitor. Anaglyph glasses aren't bad but you do lose some color information. Headsets and 3D monitors are usually prohibitively expensive.
EyeSoare 8:03PM Wed, 04 June 2008
When I set up the stereoscopic view, I normally prefer parallel viewing (or diverged eye viewing or the "wall-eyed" method) over cross-eyed, as parallel viewing produces no headache for me. I can not, of course, speak to other peoples experience. Some say it is more difficult than cross-eyed but that is not my experience. It's really no different to me than looking through a parallel view stereoscope with out the device. Suffice it to say, you can set up the view in Poser to be either parallel or cross-eyed as you see fit by simply reversing which camera is in which viewing pane.
If you, or any reader, is unfamiliar with the parallel viewing method, it may be understood as having your eyes converge on a distant point beyond the screen while focusing on the closer plane of the two images on the screen, thus, each eye sees a different part of the screen in focus allowing unique views to enter each eye producing the illusion.
The cross-eyed method forces your eyes to converge between you and the screen and is notorious for causing eye strain headaches. If there is still a danger of eye fatigue from the parallel viewing method, it is easy enough to resize the window to be smaller, allowing your eyes to converge closer to but still behind the screen. The, for me, persistent superiority of this method is it requires NO ADDITIONAL TECHNOLOGY (lenses or devices of any kind) to model or animate meaning anyone with Poser can do it. Your point is your opinion and that's cool but that this works and is already possible in Poser (and has been for a decade) is a fact which makes it a tool ANYONE CAN USE NOW. In my opinion, that makes it better than state-of-the-art technology I can't afford. I appreciate your comments.
When I set up the stereoscopic view, I normally prefer parallel viewing (or diverged eye viewing or the "wall-eyed" method) over cross-eyed, as parallel viewing produces no headache for me. I can not, of course, speak to other peoples experience. Some say it is more difficult than cross-eyed but that is not my experience. It's really no different to me than looking through a parallel view stereoscope with out the device. Suffice it to say, you can set up the view in Poser to be either parallel or cross-eyed as you see fit by simply reversing which camera is in which viewing pane.
If you, or any reader, is unfamiliar with the parallel viewing method, it may be understood as having your eyes converge on a distant point beyond the screen while focusing on the closer plane of the two images on the screen, thus, each eye sees a different part of the screen in focus allowing unique views to enter each eye producing the illusion.
The cross-eyed method forces your eyes to converge between you and the screen and is notorious for causing eye strain headaches. If there is still a danger of eye fatigue from the parallel viewing method, it is easy enough to resize the window to be smaller, allowing your eyes to converge closer to but still behind the screen. The, for me, persistent superiority of this method is it requires NO ADDITIONAL TECHNOLOGY (lenses or devices of any kind) to model or animate meaning anyone with Poser can do it. Your point is your opinion and that's cool but that this works and is already possible in Poser (and has been for a decade) is a fact which makes it a tool ANYONE CAN USE NOW. In my opinion, that makes it better than state-of-the-art technology I can't afford. I appreciate your comments.
molsmith 6:36AM Sun, 27 July 2008
I think anything useful which enables others to explore REAL 3D is a mighty plus. I am an oldie and true 3D has fascinated me since a little boy. I notice recently that many younger people engaged in art, poser, Geometry 3D, etc., are unaware of stereoscopic art and technique, which is a shame. Tutorials like this one, helps new people explore old ideas with a fresh mind and perspective. Well done! regards Mol http://www.3dcreator.net/ Renderosity Vendor and passionate about visual arts.
I think anything useful which enables others to explore REAL 3D is a mighty plus. I am an oldie and true 3D has fascinated me since a little boy. I notice recently that many younger people engaged in art, poser, Geometry 3D, etc., are unaware of stereoscopic art and technique, which is a shame. Tutorials like this one, helps new people explore old ideas with a fresh mind and perspective. Well done! regards Mol http://www.3dcreator.net/ Renderosity Vendor and passionate about visual arts.
2 questions; first, can you render in a split pane?(both images in the same render, or do you have to render 2 images seperatly using the diferent cameras?) second-can you use the "point at" command to aim the cameras?
Talisman 9:48AM Wed, 14 January 2009
I worked as an photo interpreter for the U.S. Army, and we used stereoscopic viewers, which basically facilitated the cross-eyed view for 3D. In that faux-3D environment, I recall that I only see one, composite image, but a lot of eye-strain and headache goes with the technique. This, too, diminishing over time, however. In the parallel viewing technique, I see one, nice 3D view, and the two satellite 2D views on either side. Once you can stabilize and hold the 3D view, you can easily switch between all three views. I think what is easy for one is maybe not so easy for another, so 3D glasses may be the ticket for those who find it harder to do without assistance.
I know, in my stint of military service, that I could generate any 3D view with just the sight of the appropriately-spaced 2D views but some of my fellow interpreters needed the stereoscopic viewer to create the 3D image. To each their own, I say. If the use of Open GL eyeglasses would help, then maybe that should be pursued. The technique that the author of this tutorial tries to present, successfully, I think, is that, for some people, no aids are needed usiing these views. You cna see 3D in real time, with camera motion. You do not have to render anything to see the views; I can generate a 3D view of the mannikin on this website by just focusing properly on it, thru both cross-eye and parallel techniques. I also recognize that this is not necessarily possible for all people. In the service, some people, who wanted to be photo interpreterrs, simply couldn't see in 3D and so failed the qualifying entry tests. So Kuro's point is worth noting: other tools to view the 3D-ness of this technique may be needed for some people. I think, however, that Eye has done us a valuable service when showing us how to create the camera setup. All that is left to us as users is to find the way of emulating the 3D imagee from this useful perspective. I am curious, however, to see how making a movie using two cameras would actually work; a walking sketch would be ever-so-nice with some of my favorite maodels!
I worked as an photo interpreter for the U.S. Army, and we used stereoscopic viewers, which basically facilitated the cross-eyed view for 3D. In that faux-3D environment, I recall that I only see one, composite image, but a lot of eye-strain and headache goes with the technique. This, too, diminishing over time, however. In the parallel viewing technique, I see one, nice 3D view, and the two satellite 2D views on either side. Once you can stabilize and hold the 3D view, you can easily switch between all three views. I think what is easy for one is maybe not so easy for another, so 3D glasses may be the ticket for those who find it harder to do without assistance.
I know, in my stint of military service, that I could generate any 3D view with just the sight of the appropriately-spaced 2D views but some of my fellow interpreters needed the stereoscopic viewer to create the 3D image. To each their own, I say. If the use of Open GL eyeglasses would help, then maybe that should be pursued. The technique that the author of this tutorial tries to present, successfully, I think, is that, for some people, no aids are needed usiing these views. You cna see 3D in real time, with camera motion. You do not have to render anything to see the views; I can generate a 3D view of the mannikin on this website by just focusing properly on it, thru both cross-eye and parallel techniques. I also recognize that this is not necessarily possible for all people. In the service, some people, who wanted to be photo interpreterrs, simply couldn't see in 3D and so failed the qualifying entry tests. So Kuro's point is worth noting: other tools to view the 3D-ness of this technique may be needed for some people. I think, however, that Eye has done us a valuable service when showing us how to create the camera setup. All that is left to us as users is to find the way of emulating the 3D imagee from this useful perspective. I am curious, however, to see how making a movie using two cameras would actually work; a walking sketch would be ever-so-nice with some of my favorite maodels!
EyeSoare 11:05PM Wed, 27 May 2009
PixelPusher2004, First: my experience with Poser is that it renders a single camera view so each view/each frame will be rendered separately. There are a variety of post-production techniques available after that, depending on the results you seek. Second: yes, each camera has a "point at" command. Based on my limited experience with this awesome technique, I suggest you create a separate invisible object for each camera to point at, rather than pointing both at a single invisible object. As you've done with the Cameras, slave the AuxCamera object to the MainCamera object and move it where you want the focus to be. Move the AuxCamera object to make sure your Camera convergence is on the surface of any visible figure and not within it or in front of/beside it. The flexibility will be invaluable for closeups.
Talisman, Wow! I appreciate your comments. Flattering and illuminating. I really was trying to serve the community by posting this and I'm gratified to know that it's appreciated. In the early 90s, I was fortunate to have worked at Exploratorium in San Francisco and was able to brainstorm with 3d exhibit developers, among others. That led to this: 1+1=4D: Real Time Stereoscopic Video Projection http://www.youtube.com/watch?v=JbUBSxru1is I'm glad that cartoons are finally! going 3D. I believe we haven't begun to see where stereoscopic video is heading but I predict viable home stereoscopic recording and projection systems in 5 years. Anyway, thanks for the thoughtful and informative comments.
PixelPusher2004, First: my experience with Poser is that it renders a single camera view so each view/each frame will be rendered separately. There are a variety of post-production techniques available after that, depending on the results you seek. Second: yes, each camera has a "point at" command. Based on my limited experience with this awesome technique, I suggest you create a separate invisible object for each camera to point at, rather than pointing both at a single invisible object. As you've done with the Cameras, slave the AuxCamera object to the MainCamera object and move it where you want the focus to be. Move the AuxCamera object to make sure your Camera convergence is on the surface of any visible figure and not within it or in front of/beside it. The flexibility will be invaluable for closeups.
Talisman, Wow! I appreciate your comments. Flattering and illuminating. I really was trying to serve the community by posting this and I'm gratified to know that it's appreciated. In the early 90s, I was fortunate to have worked at Exploratorium in San Francisco and was able to brainstorm with 3d exhibit developers, among others. That led to this: 1+1=4D: Real Time Stereoscopic Video Projection http://www.youtube.com/watch?v=JbUBSxru1is I'm glad that cartoons are finally! going 3D. I believe we haven't begun to see where stereoscopic video is heading but I predict viable home stereoscopic recording and projection systems in 5 years. Anyway, thanks for the thoughtful and informative comments.
It's 2013, 5 years later, and I recently purchased an amazing for what it is stereoscopic still/video camera that I can plug into HDMI anything for less than $200. Projectors are still steep but they are available. I've been puttering in my 3D studio for decades devising stereoscopic exhibits and it's my hope to go full circle and work with an old friend in their new home on the bay, to bring some additional depth to the whole exploring curiosity thing. I hope we can all look forward to it.
©2008-2015 Michael Joseph Woody