Stereo3D Toolbox v.2 released

[Pavel Houda]

Congratulations Tim!!!! the feature set is awesome. Can I set my system so that I can have both v1 and v2 co-resident, just in case I have a very long learning curve?

[Tim Dashwood]

You don't have to do anything expect download and install FxFactory 2.1.6.
Version 1.1 is still included in the installer but is declared obsolete by FxFactory. This means that old projects that had v1.1 applied will still find the plugin but you won't see it in your effects menu by default.
If you want to see the old version to continue working with it then go into FxFactory's preferences and uncheck "Hide obsolete plugins."

I recommend that you start working now with version 2's Geometry plugin. It is the primary filter for convergence adjustment and tuning out disparity.

For those new to stereoscopic 3D we also released Stereo3D Toolbox LE, which is the $99 "limited edition" version of the plugin.

We just launched the new website at 9AM today and we are still uploading backend content for user guides and tutorials but you should get a pretty good idea of how it all works from what is already there now.

[Pavel Houda]

Hi Tim,
After installing the FxFactory 2.1.6 I get a warning that two instances of the Dashwood Stereo3D Toolbox are present, then after proceeding, I find my older filter (1.0.4 which I re-named per your previous advice and want to keep) and the LE version of the filter (which apparently expired), but not the version "2". Is that what is suppose to happen? How can I try the v2? If I launch the FxFactory, the window tells me that I have 2.0 and that it is licensed to me, but I can get just the 1.04 to show up in the FCE. Do I have to delete the 1.04? When I try to check on/off the 1.0.4 (or 2.0) , they activate/de-activate together. Please advice what to do. I would preferably have the 1.04 and the 2.0. Thanks.

[Tim Dashwood]

Hi Pavel,

It would be best if you abandoned 1.0.4 altogether. v1.1 superseded it, is backwards compatible, and is still included in the Stereo3D Toolbox FxPack. You mentioned you were having issues before and that may have been a result of beta software I had sent you in the past.

I would suggest that you go into your Macintosh HD/Library/Application Support/Noise Industries folder and move the 1.0.4 FxPack to your desktop. You should only see two FxPacks in that folder with "Dashwood" names. Then relaunch FCP or FCE.

The LE version uses a different license code and as a license holder of Stereo3D Toolbox it won't serve you any practical purpose.

[Pavel Houda]

O.K., I have followed your advice and am abandoning the good old faithful rev. that served me really well. I am at the v 2.0 and all seems well except for some learning curve. Quick look - seems to behave fine. As always, thank you so much for your advice, you are most kind and helpful.

[Pavel Houda]

Hi Tim,
I just remember why I stayed with 1.0.4, rather than proceeding with 1.1. I have a problem with interlaced footage. Unfortunately, my camcorders only capture 60i, and I have no choice there. The problem is that when I maintain the interlaced video throughout, I get terrible output, probably due to scaling. There is highly enhanced interlaced edge effect. If I try to de-interlace the video on inputs, and use left clip on timeline and right in image well, the right side becomes 2x zoomed in, so of course I can never align the video pairs. This never happens when I keep the video interlaced, but then I have the above mentioned artifact. Do you have any suggestions? I am using JES deinterlacer. Thanks. You can see the results of keeping interlaced until the last moment here: http://www.youtube.com/watch?v=6qg2C8tszQs . The final deinterlace is done by YT. If I interlace after output from FCe, the results are the same.

[Tim Dashwood]

Here is our knowledge base article on this topic. Right Side Resized Help

Yes. Interlaced in 3D has many of its own temporal disparity issues however version 2 (and v1.1) of Stereo3D Toolbox handle interlaced footage much better than v1.0.4 ever did. In fact we put a lot of resources into improving interlaced inputs in the plugin, but FCP passes interlaced footage to FxPlug plugins using a different method than Motion and After Effects so it has been a real headache for us as plugin developers. This is one of the primary reasons we started from scratch on the internal programming of version 2.

Please try a standard interlaced workflow with version 2 first (interlaced clips in an interlaced sequence) and then compare to interlaced clips with field-dominance set to "none" and your sequence field dominance set to "none."

In our internal testing the best 3D results are had with field-dominance set to "none." However we created the clip-coupler plugin to aid in the creation of stereo-pairs. I have new tutorials coming within the next week to explore various workflows.

[Pavel Houda]

Thanks Tim. I will try later today or tomorrow. I have no scaling problems with the interlaced footage, with that I have the fuzzy edge problem when I look at the output in QT. It doesn't seem to matter on how I set the input field dominance, but I didn't try to change the sequence dominance. The scaling problem I have after I deinterlace the input clips with the JES deinterlacer before using them in the FCE, in which case the input clip field dominance is "none", again, I didn't check the sequence setting. I tried to deinterlace first to get rid of the jagged edges. I will keep looking into this. I think that the image might get slightly scaled as the two sides are aligned, and than the deinterlacer cannot do a clean job deinterlacing the sequence outputs because the actual scan lines no longer line up with the original field lines.

[Pavel Houda]

Tim, thank you very much for your suggestions. I looked into the field dominance settings, and it indeed is the contributor of the problem. I set up interlaced video with field dominance = none for both the clips and the sequence, and there is the problem: http://www.youtube.com/watch?v=hkC1Ke-OgDo . Setting the same interlaced clips and the sequence to "upper" field dominance, the video is fine: http://www.youtube.com/watch?v=At4CJyEKD_A . When I de-interlaced the clips first, and again, setting the clips as well as the sequence to "none" for field dominance, the video looks fine: http://www.youtube.com/watch?v=PehKMWrBekQ . I also made a quick sequence from the NAB that I took at your booth in the Panasonic area, interlaced, with the field dominance set at "upper", and everything looks good to me. The URL is: YouTube - NAB Booth - Tim Dashwood - Stereo3D Toolbox v 2.0 - 3D HD Stereoscopic . In conclusion, thank you for your continued help, I had really no issues with using the Stereo3D Toolbox v 2.0., once I got the settings correct, however, there are lot of new great features in it that I didn't use yet, that should greatly improve the outputs and make the work faster and easier.

[Pavel Houda]

Hi Tim, could you do a tutorial on how to optimize anaglyphs? Never mind, you already have one. Sorry.

[Guillermo Ibanez]

This might have been asked before but......any chance of a future version for PC after effects? that'd be awesome! thanks!