Cineversity

Hello,
I’ve been looking for a solution to Stereo Rendering since Team Rendering has been introduced.  We do a lot of Stereo work, and have been using Net Render even though we were on V15.  We have since moved to V16 (hopefully to V17 sometime next year).  We’ve been using net render from V13 to do our network rendering.  Since V16 introduced Reflectance channel, we are concerned that net render won’t be consistent.  I haven’t found an alternative to Stereo rendering via Team Render up to this point.  Is this rig still available somewhere? Is there another option for Network Stereo rendering at this point that you can suggest?
Sincerely,

Mike Goldberg
Post Production Supervisor
Center City Film & Video

Hi Mike Goldberg,

If you encounter any problems currently with the Team Render, please contact the support.
http://www.maxon.net/en/support/support-questions.html

Two things come to mind based on your question:

• Download the demo of R17 and check out the Take System. If that could help.

With the Take System you can set up for the left and the right camera a “Take” and queue this for Team Work Rendering (I’m currently not a sure how much the Demo version allows for a full Team Rendering* It seems that the Render Queue and Team Rendering is not available).

I followed the Stereo idea from the very beginning in C4D with an focus on VFX, so I know it is not just two cameras and done. With that being said, your workflow with R17 might be different than someone else’s. The demo version will allow for an individual evaluation. Especially here, to check the set up you that you might have in mind with the Reflectance channel. Note that you can stay in a kind of “Legacy” mode as well. Low samples in the available algorithm might be not advised for Stereo, but I guess you know that very well.

• You might get in contact with MAXON US directly, as some people have specialized themselves for studio support, certainly with an emphasis on Team Rendering. Have you checked out Rick Barrett’s tutorials about the Team Render and the Take system? I guess it will accelerate any further discussions, just a thought.

The stereo rig - do you have the “native” Stereoscopic rig of the camera in mind or a specialized rig?

All the best

*
The CINEMA 4D demo version’s project files are of the same format as the commercial CINEMA 4D version, which allows files to be exchanged. Rendered images and animations or painted textures contain no stamp or watermark. The render size is limited to 800 x 600 pixels.

Hello Sassi,
How are you?
Thanks for the quick response.  I know we probably won’t be able to upgrade our systems to V17 at this time.  We are under a bit of a crunch, and software is always the first thing that they don’t want to update.  We are stuck with V16 for at least a few more months.  I’ll check out the Take system at some point.  What I’m looking for is a way to render the individual channels of the stereo camera within Team Render.  It ofcourse does not work as expected.  I was going to simply create my own rig by parenting 2 cameras to the stereoscopic camera, and adjusting each as needed.  I’d then render those individual cameras.  When I saw Kai’s tutorial here:
https://www.cineversity.com/vidplaytut/stereo_camera_settings
it looked like he was using an old stereo rig that he had built prior to Maxon introducing the stereo camera in (I believe) V13.  I thought if that rig was still available I’d give that a try rather than building my own.

Sincerely,
Mike

Thanks Mike, I’m fine, I hope you are fine as well.

Let me ask you, where do you see an advantage of Kai’s set up vs the one natively integrated? Since I do not have that set up anymore.

If the native Stereoscopic is set to Mode>Left or Right, Symmetrically defines the position of the rig and single channel set to 1 or 2 the camera of the rig, the rendering should work as with one camera—per channel. Perhaps I assume that the channels stay separated during rendering, and after post production the material will be set for a specific viewing method. Anaglyph or Polarization based, etc.

Is the problem perhaps given with the idea to have it merged during rendering?

With the two images below, you should get two different results, Left and Right, as long as Single Channel is selected. If the settings are different, things might be overwritten and only one channel is stored in the image. The Left right thingy is just for testing. It sets a specific camera in the main position, so this is only thought as testing!

Either way, I think it is a good idea to check with the support why you have no success with it. It should work, even if I’m not exactly clear what kind of result you are after.

My best wishes

Yes, Sassi,
that’s exactly what I was hoping for.  In fact, that’s our normal workflow.  However, Team Render does not support Stereoscopic Rendering.
I did several tests today to try to get Team to work, but nothing worked (when set to a stereo mode, the render did not complete), and when I set it to an individual camera in non stereo mode, it only rendered 1 camera (I actually rendered it twice with Single Channel set to 1 and then to 2, but only got the same render on both).  I have contacted support, and they simply told me that Team Render does not support Stereo Rendering.  I was thinking by using kai’s setup, it wouldn’t interfere with Team. 

Mike

Mike,

Since Team Render is not something I can simulate here with my small set up, running only one OS here anyway, it is difficult to tell.

What I can tell is, that If I set up the channel 1, the picture viewer shows the left, if I THEN render the right Channel which is 2, it renders the right camera. Let me know if you get other results, while following along to the point, in other words, is the Picture viewer and the Team Render presenting you with different results?!

If I get your results correctly, this Left OR Right is always the same image.

I have made a one minute clip, all steps to get a LEFT and RIGHT image/camera (or channel 1 or 2) in the Picture Viewer. The 100cm between the cameras is for dramatization only!
https://www.amazon.com/clouddrive/share/zSopbJEU4hBFmnvd0bhCa4E3HwLulA8O1rBQrwRapJp?ref_=cd_share_link_copy

The set up from Kai, the focal point would be the Target-object (distance in Z from the middle camera) and the Eye distance would be half of the X distance (plus and minus), all of that can be done with a little script. The Target object can be a Plane (Z direction) set to no-render and with a Display Tag to the function you need—if a no parallax indicator is a must have. The data can be taken from any camera via XPresso, as all other parameters.

Cheers

P.S.: here is the file, a little bit updated.

In the camera Mode> Left, Right, or Symmetrically works as well of course, but for the testing I believe it is clearer to leave it alone. This switch is not necessarily needed.

In the Render Settings> Single Channel and Render Settings>Channels [number] 1 or 2 (the channels need to be set to two [amount] to work) (... was that eh missing part?)

See the entries accordingly to the movie
https://www.amazon.com/clouddrive/share/zSopbJEU4hBFmnvd0bhCa4E3HwLulA8O1rBQrwRapJp?ref_=cd_share_link_copy

Note: this is a test and perhaps more a workaround for demo purposes, e.g. the 100cm eye distance, when we get here a difference in the images, then we might just found the way. In the moment I know more what you are after, we specify the way then, if needed

I noticed that I was not clear in my posts above, I have edit them and made certain that the left and right camera is mentioned but not confused with the parameter [Mode]

Hi Again,
So I did some further testing. It seems that for Team Render to work, I need to check the box that says “Non Stereoscopic Image”
I then set the Calculate Stereoscopic Images to Single Channel.  Next I set the Channel to 1 or 2 to render each eye separately.  This seems to get the job done, although I do have to render separately.  Without Team Render, I can render both eyes at the same time.

Mike

Thanks for the update, Mike.

Sounds like a step further.

Just so I understand it clearly: the Picture Viewer shows the correct render [channel 1 or 2], but the Team Render gives you only one “channel”? If so that would be in need of a fixing from my perspective. As I have checked this morning the manual for any limitations of course, this was not mentioned anywhere (Render Settings, Stereo Camera).

That a mixed image, as in the dual color anaglyph needed, is not doable currently in the Team Render—I can [kind of] understand. However, a single camera rendering should be possible. This sounds certainly something the support needs to inspect. It makes no sense that this [again] single camera request is not accepted from TR.

To ask you again to contact the support for clearing this detail (was anaglyph discussed or the output of a single channel?), seems not really cool, I understand. However, if it is a technical problem, then I’m not supposed to handle it, Support is one part, explaining tools and function the other one.

If you have one day a lot of channels, like in Alioscopy, this would be a huge limitation, as this format is not little in rendering work.

I hope that is acceptable.

My best wishes

P.S.: How about double the image width and use the side by side mode. Easily to separate in post.