I primarily use Poser to make side by side stereo images. I finally started working with animation. My fist one had a dancing girl and a stationary main camera. For the second rendering, I just dolly the camera a little to the right and I'm all set.

My new project had zooming, rotating etc. etc. of the main camera. When I go to the first frame for the second rendering and dolly the camera, it only works for the first frame. There MUST be a way to tell the camera to be 6 inches or so to the right for the entire sequence without having to change each frame. Does anyone know how to do that?

Thanks,

Mike

don't have Poser open right now, but far as I know you need to add a 'constrain' to the camera

     Do you mean 6 inches to the right relative to the world, or do you want the camera to dolly so as to follow the doll?

I want the exact same motion as the first pass but 6 inches to the right as if there were 2 cameras linked together shooting in the first place.

I would just render the first camera frames, then add a animation layer to the camera and offset it, then render the second set of frames.

**"My new project had zooming, rotating etc. etc. of the main camera. When I go to the first frame for the second rendering and dolly the camera, it only works for the first frame. There MUST be a way to tell the camera to be 6 inches or so to the right for the entire sequence without having to change each frame. Does anyone know how to do that?"**

Hi you can address this easily in your graph editor.
What is occurring is that you have key frames
affecting your camera position AFTER the First frame where you are setting the position  you would like for the rendering.
Probably left over from the first rendering sequence
you need to drag select and delete those keyframes  so your you camera does not move after the frame where you set it position.
I hope this is clear.

Cheers

shvrdavid

Could you be a little more step by step? I figured out how to add a new animation layer on top of the base layer but how do I offset it and what about the other buttons, "include in playback""replace""add?"

IMO, all wrong answers.

Assuming you used the main camera first, parent the dolly camera to the main, six inches away. The dolly camera will be lock stepped to whatever you did to the first camera. Use the dolly camera for your other "eye".

You can also make a new camera for this purpose if you like - but the idea is the same. When you want one object to stay in specific position relative to another, you parent one to the other.

Quote - IMO, all wrong answers.

Assuming you used the main camera first, parent the dolly camera to the main, six inches away. The dolly camera will be lock stepped to whatever you did to the first camera. Use the dolly camera for your other "eye".

You can also make a new camera for this purpose if you like - but the idea is the same. When you want one object to stay in specific position relative to another, you parent one to the other.

That would be the easiest thing, i think. But the problem here is that Mike wants his camera not only to move around, but also to zoom. I tried to bind the focal length of one camera to the other camera with dependent parameters, but for some reason it does not work (with Poser8); the cameras seem to completely evade the grasp from the dependent parameter's force (for both driving and receiving them). I did not try with the deltaAddDelta thing in the pz3 file though. Also, the dependent parameters might work in Poser2012, if so, it would be a very elegant solution.

MG - I see that and you're right - he wants it to zoom.

The original post does not ask if anybody knows how to lock the zooms, so I didn't react to the zoom issue. It only asks "There MUST be a way to tell the camera to be 6 inches or so to the right for the entire sequence without having to change each frame. Does anyone know how to do that?"

And so parenting is how to do that.

I will experiment in PP2012 and see if the parameter linking is borked there, too.

 

This is interrupting some serious material research, just so you know. *grin*

Instead of screwing with the camera, just rotate the items in your scene.  Render with same camera.

Quote - Instead of screwing with the camera, just rotate the items in your scene.  Render with same camera.

Please demonstrate. This sounds preposterous.

Well well. It seems SM completely forgot to connect the camera parameters to the dependent parameter system. I can make the focal length be a master parameter that controls other things, such as prop scale, but I cannot get any camera parameter to depend on anything else at all. Somebody should file a bug report if they care about this. I don't.

**"IMO, all wrong answers.**

** It only asks "There MUST be a way to tell the camera to be 6 inches or so to the right for the entire sequence without having to change each frame. Does anyone know how to do that?"**

And so parenting is how to do that."
Ummmm.. Why??

Perhaps I am misunderstanding the problem
But Does poser now have the ability to render from two cameras simultaneously?? if not parenting and second animation layer etc. is way over complication of a very simple keyframe animation task
with your GRAPH EDITOR

Read this:"When I go to the first frame for the second rendering and dolly the camera, it only works for the first frame."

The OP did his first animation with a "stationary" main camera(No key framed movement)

the OP Does NOT know how to set additional key frames for his dolly moves as the first is set automatically from its position on frame one.

I just did a quick animation by moving my dolly camera around and properly setting a key frame for each position
and as expected the camera dollied and zoomed as I intended
see what it looks like in my graph editor( attached)
This is basic key frame animation 101.

Please stop confusing him with all of this gobbledygook

Cheers

Oops, I said "rotate" instead of slide everything on the trans x axis by the same tiny amount (whatever six inches would be)

I don't understand what's confusing about that?

 

 

 

Quote - Perhaps I am misunderstanding the problem

Yes you are. The objective is to have a pair of cameras in lock step, but only have to set the animations of the first one - the second is attached and automatic.

Your suggestion is to re-keyframe for the second camera, as if its movements are independent.

Parenting the second camera means that it will exactly do what the OP asked. And parenting is not gobbledygook. If I explained the math of parenting, which is actually required to make clear why William's suggestion is wrong, then it would be goggledygook.

William - when the viewpoint is turned, the displacement is no longer on the x axis. For the second camera to be 6 inches to the right of the first camera at all times, means that you move the second camera with respect to the first camera, not with respect to the world axes. In particular, when the cameras turn 180 degrees, the translation is in the exact opposite direction. When the camera turns 90 degrees, the translation is completely in the Z axis, and not at all in the X axis. When the camera tilts and rotates (as would be the case when shooting from an airplane's point of view), the translation would be non-zero in all three axes.

Yes, I already said...Forget rotating the scene/elements.  Just slide them on x axis. 

William - I responded to that already. The world does not appear to move only on the X axis.

You're not grasping the problem. Visualize this:

An airplane has a camera on each wingtip. The animator wants to record what these two cameras see by animating the airplane through the world. When the plane turns the two cameras are changing position and orientation, and at each moment through the turn, the relative position (i.e. translation of the perspective) is DIFFERENT. Only at one point in the turn is it a pure positive X axis shift, if at all. Everywhere else, it is both X and Z, and possibly even Y. If the plane is banked, one camera is higher than the other - a change in Y position.

Poser cannot render both cameras in one
pass so he will have to render the
second "six inches to the right " version by itself correct?

@Bagginsbill just like you have to see a persons node setup before you can definitively tell why his material has no "specular"

until we see a screen shot of his GRAPH EDITOR of the problem scene file every suggestion we are all offering must be considered sheer speculation.

My GUESS is There is no need for a second "parented" dolly  camera to be "in lock step" with first
in fact I dont find where he states such in his post

this can be achieved with ONE camera

"For the second rendering, I just dolly the camera a little to the right and I'm all set.

My new project had zooming, rotating etc. etc. of the main camera. When I go to the first frame for the second rendering and dolly the camera, it only works for the first frame."

My GUESS is What is happening he is moving his former stationary camera on frame one this is autimaticly setting a key frame for that frame only in the" six inches to the right" position'

he probably has his interpolation method set to linear.

that one new position frame is not going to hold his camera "slightly to the right" for his entire sequence he needs to set at least on addition keyframe at the END of the sequence "holding" that "slightly to the right position.

anyway its friday  Got to run to the Mosque
OP: please post a screen shot of your  graph  editor for a proper diagnosis

Cheers

You keep asking if this requires one pass or two. The way the OP is doing it requires two passes, but I can think of a way to do it in one pass, using mirrors. But that has nothing to do with the question of stereo projection.

Meanwhile,

You're still not understanding it. He does not mean 6 inches to the right IN THE WORLD. He means six inches to the right of its position in the first pass. Look at this illustration. The left-eye camera has a red arrow. The right-eye camera has a green arrow, and stays six inches to the right of the left-eye camera, with respect to the orientation and position of that camera, not the world.

Imagine these are key frames.

In the first key frame, the right-eye camera is +6 in X. In the last key frame, it is -6 in X. In the middle key frame, the translation in X is almost zero, and the translation in Z is almost 6 inches. You cannot manually set up these positions as an adjusted key frame. You must do this via parenting. Any attempt to do it manually will result in eye-angle and distance deviations that will be obvious when viewed on a 3D TV. You will make people throw up.

We don't need to see the key frames of his actual scene. This has nothing to do with adjusting his animation. It's the simulation of stereo vision, using two cameras or one camera rendered twice, with a shift in position. Or one camera with a split mirror setup.

 

Quote - My GUESS is There is no need for a second "parented" dolly  camera to be "in lock step" with first in fact I dont find where he states such in his post

He states it by using the word stereo. If you don't know what that is, then you should take it as a given you're not responding to the question.

Quote - I primarily use Poser to make side by side stereo images.

Real world - one-camera with slide, or two camera setups for 3d stereo photography.

http://www.pokescope.com/cameras/index.html

The OP wants the Poser equivalent.

"I primarily use Poser to make side by side stereo images. I finally started working with animation. My fist one**{ON THE LEFT OF YOUR SCREEN}** had**{POSER DORK WALKING}**and a stationary main camera.

 For the second rendering**{ON THE RIGHT OF YOUR SCREEN}, I just dolly the camera a little to the right{ OF POSER DORK}** and I'm all set.

My new project {ON THE RIGHT OF YOUR SCREEN} had zooming, rotating etc. etc. of the main camera

 

When I go to the first frame for the second rendering and dolly the camera, it only works for the first frame.{ UNLESS I ANIMATED  IT PROPERLY WITH POSERS ANIMATION TOOLS)

There MUST be a way to tell the camera to be 6 inches or so to the right**{WITH AFORMENTIONED ZOOMING& ROTATING}** for the entire**{SECOND}** sequence
{AS SEEN ON THE RIGHT OF YOUR SCREEN}
without having to change each frame. {I ONLY SET THREE  KEYFRAMES}"

>>>video link>>WHAT AM I NOT UNDERSTANDING HERE

Cheers

Wolf, I have no idea what you're saying or asking.

mikeihde wrote

Quote - I want the exact same motion as the first pass but 6 inches to the right as if there were 2 cameras linked together shooting in the first place.

I don't know what else to show you. I rendered two cameras linked together moving through various positions. I showed you real 3D cameras from another website. That is parenting. They are locked together, moving as one through a scene, recording two different viewpoints, one for each eye.

It does not require looking at the animation time line. Regardless of how the animation proceeds, whether panning, turning, tilting, pitching, rising, falling - whatever the motion, the two viewpoints need to stay locked together with respect to each other, i.e. an imaginary pair of eyes in an imaginary head.

 

"For the second rendering, I just dolly the camera a little to the right and I'm all set.

My new project had zooming, rotating etc. etc. of the main camera. When I go to the first frame for the second rendering and dolly the camera, it only works for the first frame.....
I want the exact same motion as the first pass but 6 inches to the right as if there were 2 cameras linked together shooting in the first place."

Well bill I am going to wait for the OP to return and clarify himself.

You are utterly ignoring the first post where he mentions the zooming and rotation.

If indeed he has changed his mind about zooming& rotating then  
Look at my video.

it is rather obvious to anyone I could have easily NOT added any Zooming and rotation in my "second pass and achieved the described objective(Camera slightly to the right of Dork) Exactly with just simple key framing

 

I dont know what else to show you

 

 

Cheers

I'm talking about rotation. I showed rotation in my image. What are you talking about? I've also shown you that translation by a fixed amount and direction is incorrect - it depends on the rotation. I feel you're being intentionally opaque.

And we already were discussing zooming much earlier - if dependent parameters worked that would have taken care of it.

Failing that the zooming has to be dealt with another way - one easy way would be to run a script to copy the zoom settings to the other parameter. Such a script is fairly easy to write.

 

Again Bill we both may be guilty of making assumptions about an OP who apparently
(for now at least ) has abandoned his own thread
and frankly has contradicted himself when he said:

"My fist one had a dancing girl and a stationary main camera."

then he goes on to say:"I want the exact same motion as the first pass but 6 inches to the right as if there were 2 cameras linked together shooting in the first place."

Now Bill I ask you ,what "motion" in the first pass"??

according to him :"My fist one had a dancing girl and a stationary main camera."

the only "motion is a dancing girl not the camera

this alone is enough for me to doubt any of us knows his true objective so I will cease any further speculation about what animated camera move he was trying to create.

Cheers

Guys, thanks for all the ideas. I feel the parenting idea sounds the most promising. The idea, again, is like having a stereo camera with 2 lenses, each about 6 inches apart from each other. My first pass is with the left lens, the second with the right. Then I put them both into final cut and make a side by side 3D movie.

I'll try the parenting idea later tonight and let you kow how it goes.

The image wolf369 posted was just turning the camera from it's original position creating an angle. 3D lenses must always be parallel just like our 2 eyes.

Hi Thanks for the clarification!!

your original question seemed to indicate you did not know how to make you second pass camera remain in its offset position for the length of your second animation
this gave me the impression you were having problems  with keyframing it in its offset position

Stereoscopic  2D footage rendering to produce "3D footage"
( for use with the 3d glasses) is available in Blender and the latest version of Maxon C4D

I would be surprised that even parenting two poser cameras would produce a proper Stereoscopic image but perhaps bagginsbill could script you something.

Cheers

Quote -
My new project had zooming, rotating etc. etc. of the main camera.

Zooming as in changing the focal length? The human eye can't do that, it has a fixed focal length. I would expect that this would look very strange in a stereoscopic animation.

I have not read every post in this thread, but from what I have read, In think everyone is neglecting an important point. People are talking about making two cameras parallel, but I don't think that is how it should be. In real life would not both eyes point inwards towards a point of focus? And thus, should not stereo cameras do the same? I have no special knowledge of the human visual system, or stereoscopic photography, but it just seems to be that parallel cameras would give a totally wrong and unnatural effect. Anyone have any thoughts on this? At the moment I have started to construct a set of stereo cameras that work on the assumption that my above theory is correct. The cameras will be linked entirely by ERC, including zoom. Poser 'Point At' will be used to get the parallax adjustment, which will need to be set manually, as appropriate to the particular scene, and including animation.

Quote - Zooming as in changing the focal length? The human eye can't do that, it has a fixed focal length.

That is true, but it is also true that the fact hasn't stopped movie makers using zoom. In the movies anything is possible, and reality is never allowed to get in the way of a good fantasy.

Stereo cameras have perfectly parallel lenses. True, our eyes may "toe in" as we call it but as a rule, all stereo cameras over the last 100 years or so have had non adjusting parallel lenses. I tried parenting the aux camera to the main camera, copy and pasted the keyframes from the main to the aux then moved the aux camera's key frames 3 blocks to the right and I think that may do it. I have some more experimenting to do.

 

I'm attaching a stereo Poser image. If you cross your eyes, you'll see 3 images. The middle one will be in 3D.

Mike

I figured it out. I parented the aux camera to the main camera then copy/pasted the keyframes from the main to the aux and moved them all 3 frames to the right. That did it. I rendered the main camera and saved then the aux camera and saved, opened them and payed them together and BINGO!

Thanks for all your help!

> Quote - I have not read every post in this thread, but from what I have read, In think everyone is neglecting an important point. People are talking about making two cameras parallel, but I don't think that is how it should be. In real life would not both eyes point inwards towards a point of focus?

Yes, but when viewing images on your computer screen, your eyes already do that by themselves (unless press your nose on the screen while looking at it, of course). The amount of squinting varies depending on the position of the viewer, most noticable in big cinemas, with those polarization 3d movies, where the seating position in relation to the silver screen has a large impact on how you perceive the movie.

Quote - And thus, should not stereo cameras do the same?

That would be useful if you project the image the camera takes directly into the eye, like with a retinal display, but for "normal" images looked apon from an unknown distance it is usually more common just to use the more eye-relaxing parallel approach.

Quote - I have no special knowledge of the human visual system, or stereoscopic photography, but it just seems to be that parallel cameras would give a totally wrong and unnatural effect. Anyone have any thoughts on this?

The most irritating thing about the stereoscopic 3d, in my opinion, is the fact that you cannot focus on things that were filmed out of focus anymore, or vice versa. I.e. 3d movies sometimes are taken equally sharp in foreground and background, whereas in the real world the viewer could decide what should be in focus and what not. The cross-eye thing is hardly noticable for distances more than a few meters. (But for close distances, it can be; sometimes it can be experienced when looking straight on a very regularly patterned surface that covers the whole field of view, like a checkerboard painted wall).

Quote - At the moment I have started to construct a set of stereo cameras that work on the assumption that my above theory is correct. The cameras will be linked entirely by ERC, including zoom. Poser 'Point At' will be used to get the parallax adjustment, which will need to be set manually, as appropriate to the particular scene, and including animation.

Something like in the illustration above, i understand? That would be interesting to know how this could work. Apart from the fact that poser's cameras seemingly do not want to be dependent on any master parameter (see one of the BB's postings above), i have never had any success in using dependent parameters if the master parameter is a rotation generated by a PointAt. If this worked somehow (with some trickery probably, if at all) it would be very useful far beyond this camera application.

Quote - > Quote - Zooming as in changing the focal length? The human eye can't do that, it has a fixed focal length.

That is true, but it is also true that the fact hasn't stopped movie makers using zoom. In the movies anything is possible, and reality is never allowed to get in the way of a good fantasy.

OK, if you don't need the cameras to "toe in", then that makes things a lot simpler. You might like to try the attached "StereoCamsNoToe_v01a.pp2", as an alternative to your method. It contains two cameras, "Camera_A" and "Camera_B". Camera_B is offset to the side of Camera_A by -0.028 PNU (-73.396 mm). Camera_B is locked by default and its rotations, translations, focal length, and Focus Distance are slaved to those of Camera_A via ERC. This means that Camera_B should follow almost everything that you do to Camera_A, maintaining a side by side stereo view through dolly and orbits. The separation of the cameras is implemented via extra xtran channels in both cameras. If you want to change the separation, unlock Camera_B and change the min and max value of 'dollyX2', then re-lock the camera. You can also split the offset giving half to Camera_A and the other half to Camera_B' if you prefer it that way. Note that 'Restore Camera' will not work on these cameras, but 'Restore Element' will. If you are going to use motion blur, you will need to unlock Camera_B, and enter the fStop and shutter values manually, as these have not been slaved. To use the attached zip, you will need to loose the ".TXT" part of the file extension. Finally I would point out that this is experimental, I hope that there are no bugs, but make no guarantees. If you find any problems, or have suggestions for improvements, please let me know.

@millighost,

Quote - Apart from the fact that poser's cameras seemingly do not want to be dependent on any master parameter (see one of the BB's postings above), i have never had any success in using dependent parameters if the master parameter is a rotation generated by a PointAt. If this worked somehow (with some trickery probably, if at all) it would be very useful far beyond this camera application.

I don't know what caused BB's problem. In my stereo cameras I used traditional deltAdd ERC rather than the new keyed type. I also slaved the focal channel of Camera_B directly to the focal channel of Camera_A, without any intervening valueParm channel. It may be that camera parameters do not like being slaved to valueParm channels, but that's just a maybe. What ever, all my ERC seems to work in PP2012 (SR2.1). I don't know if it will work in other versions, but I see no obvious reason why it should not. As to 'Point At', there is no way (that I know of) to to use data from a rotation generated via Point At as input for ERC. You should be able to control the strength of Point At via ERC, but that's all. However I'm unsure of why or how you want to use ERC with Point at. What are you trying to achieve? If you can describe what the goal is, there may be a way. I was working on a version that supports "toe in", I kind of put it on the shelf when mikeihde said he did not need it, but if there is any interest I can dust it off and try to finish it.

Remember those old style 3D movies where you had to wear red and green specticals? Does anyone know the exact colors needed for the camera filters? Also where to find the specticals?

http://www.rainbowsymphonystore.com/3dglasses.html

Also, here's a tutorial on how to make anaglyphs (red and blue 3D)

Quote - I don't know what caused BB's problem. In my stereo cameras I used traditional deltAdd ERC rather than the new keyed type. I also slaved the focal channel of Camera_B directly to the focal channel of Camera_A, without any intervening valueParm channel. It may be that camera parameters do not like being slaved to valueParm channels, but that's just a maybe. What ever, all my ERC seems to work in PP2012 (SR2.1). I don't know if it will work in other versions, but I see no obvious reason why it should not.

Yup, the valueOpDeltaAdd seems to work in 2010.

Quote - As to 'Point At', there is no way (that I know of) to to use data from a rotation generated via Point At as input for ERC. You should be able to control the strength of Point At via ERC, but that's all.

Hm, if you do not know it, it is very likely not possible, too bad :-(

Quote - However I'm unsure of why or how you want to use ERC with Point at. What are you trying to achieve? If you can describe what the goal is, there may be a way. I was working on a version that supports "toe in", I kind of put it on the shelf when mikeihde said he did not need it, but if there is any interest I can dust it off and try to finish it.

I give an example in a seperate thread, because it is off topic and a bit lengthy.

Hello all,

Mikeihde: Yes, parenting the aux camera to the main camera is the simplest way to make a movie. This approach is described somehere and I have used it in my Poser Python script to automate the generation of the right and left view. See the following. 

http://www.renderosity.com/mod/freestuff/details.php?item_id=63140

My script is not always working well but follows carefully the instructions if you use it. Since your are able to see in cross view, I suggest that you use the two pane views available in Poser. With it, it is possible to see directly the stereo effect.

Lesbentley: I like very much your approach of using ERC to control the Camera B. In your message, you suggested to use a toe in approach of the cameras. It is OK but that approach genereates what is know as a Keystone effectd (a trapezoidal deformation that is not similar in the two views). In some cases, this can be disturbing when looking at stereoscopic images. In fact, the focusing (toe in) toward a specific object in done during the stereoscopic post-production. The two views are superposed imitating the natural movement of the two eyes toward a specific object of the scene. In stereoscopic langage, we say that the images are placed at the stereowindow.  

Allstereo, thanks for the clarification on this subject. I think I understand a little better now. I can imagine how the Keystone effect could be disconcerting in many circumstances. It's interesting to realize that our eyes must be experiencing this Keystone effect much of the time, but it would seem that our brains are doing a bit of postwork here, which sounds like no mean computing feat.

Eyes are round. Renders are flat projection, and not representative of how the images form on the retina, which is curved.

allstereo,

Thanks so much for that. I cant wait to give it a try. I looks like the perfect solution to my problems. Thanks again.

Mike

Quote - Eyes are round. Renders are flat projection, and not representative of how the images form on the retina, which is curved.

What you say is true, but I can't quite grasp how it is relevant to the Keystone effect, which was presumably the subject of your comment. It seems to me that focal plain of the eye being roughly a section of a sphere would mean that the hyperfocal distance is represented as a concave section of a sphere. But that's focus. Wouldn't the Keystone effect be related only to the incident angle (line of sight), which is invariant with respect to the shape of the image plain?

pardon me for interrupting such a spirited discussion but, it seems to me that in order to get both cameras oriented correctly both cameras would have to rotate on the same center axis... EG:place both cameras at 0,0,0  move first camera 3" to the left the second 3" to the right ...then move their origins back to 0{if possible} and parent either one to the other...

... forgive me if this doesnt make sense i have a simplistic mind  

:O)

Quote - pardon me for interrupting such a spirited discussion but, it seems to me that in order to get both cameras oriented correctly both cameras would have to rotate on the same center axis...

Yes, it is necessary that the two cameras rotate around the same point. However for the purpose of producing a stereoscopic image, it is not necessary (though it may be desirable) that the axis of rotation be half way between the cameras. The axis can be anywhere, and the image will still be stereoscopic.

In my 'StereoCamsNoToe_v01a' the axis is directly in front of Camera_A, and thus offset slightly to the side of Camera_B. This may not be ideal from the point of view of manipulating the cameras or making calculations, but it's the simplest way to go, and does not compromise stereoscopic nature of the images.

 interresting topic and the discussion thought provoking, deeper thought reveals hidden considerations , such as the way our brain handles visual input,  almost like it reassembles what is presented to it and makes use of whats most logical and disreguards the rest... i put 2 soda cans on my window ledge and looking between them at a fixed point shows the stereo view, closing either eye without moving still focusing on the same point reveals details that i dont notice when both eyes are open unless i conciously bring them into focus ... seems theres more involved than just the mechanics of stereo photography, perhaps the brain has to accept what is presented as 'real enough' in order to 'see' dimentionally... i know i have seen a few 3D movies where i had to almost force myself to accept what was presented as 3dimentional . some techniques are better than others but, i think all have to be conciously accepted  in order to be enjoyed fully...

:O)

Hello all,

  Mikeihde: I also wrote a Python script to generate the left and right images of common stereoscopic images. See the following item:

http://www.renderosity.com/mod/freestuff/details.php?item_id=62715

 

Allstereo

Allstereo, should the stereo base vary with the focal length? If so what is the formula?