Avid for Stereo Editing?

[Matt Faw]

Dear all,

So far, I've done all my stereo editing on Vegas, mostly to stay with the native .m2ts files I was using, and to escape the data explosion which came from transcoding.

However, I'm a professional FCP editor, and found Vegas often frustrating, so for this feature-length project, I've started prepping the footage for Final Cut use.

An Avid instructor friend of mine is pushing me to reconsider that plan, and instead, cut on Avid. He says that Avid has some of the best stereo tools on the marketplace (although I don't think that he has cut stereo, himself). And besides, FCP seems to be on its way out, so it's a good chance for me to re-gain the Avid skills I've let deteriorate, to remain commercially viable as an editor.

I'm wondering if others on this board have cut stereo on (the latest) Avid, and if so, what were your experiences like?

[Bruce Allen]

Avid is pretty good for stereo editing. Well, it's very good for editing in general... and OK for stereo.

They've had basic stereo support in there for a while - all of the major stereo features were edited on Avid so it just made sense. But it was meant to work in conjunction with a finishing program. Eg do offline edit in Media Composer, finish in Quantel.

As of Media Composer 6, they are adding more stereo stuff - eg stereo-aware effects, convergence adjustment etc - which is very good. And you now have full-res on both eyes (1920x1080 for both L and R) instead of interlancer / over-under / side-by-side.

The main gotcha is that if you want to monitor full-resolution left and right eye (1920x1080p each), you need Avid video output hardware (which has gone down in price by a lot but still is more expensive than a Blackmagic box). They haven't implemented it on 3rd-party hardware yet - currently only side-by-side or over-under.

Of course, many pro monitors (the ones with the polarized glasses that are easier on the eyes) only have 1/2 vertical resolution, so there should be no penalty.

Personally, though, I am using shutter glasses and a full-res Sony projector so this kinda sucks.

Currently, I use Avid in conjunction with DaVinci Resolve - which supports full 3D of course via Blackmagic UltraStudio 3D. So I have basically an offline/online workflow with the creative 3D edit done in Avid - and Resolve used for final finishing / color / picture QC. Which is kinda needed anyway.

Here's a quick overview:
http://apps.avid.com/Get-trained-mc6...reoscopic.html

Bruce Allen
www.boacinema.com

[Matt Faw]

Thanks Bruce! It's good to hear your experience, and the videos you linked to are very useful. I wish I had seen the terminology vid months ago, when I was still trying to figure out what you were all writing about.

I haven't even tried to figure out my finishing solution yet. I finished my previous short in After Effects, but I've heard a lot about Nuke in this context. I hadn't considered DaVinci Resolve before, but I'll look more deeply into it, especially since they say they have a lite version download for free.

I'm a recovering luddite, so I still have this lingering myth that "DaVinci is something that tech-heads do for me, rather than a tool I can use myself". Time to shed those myths!

Thanks again!

[Matt Faw]

Bruce (and others),

One of the Avid videos brings up an interesting point for me.
What Could You Do with Pro Tools|HD Native?

When setting up the project, Media Composer asks for the eventual size of the screen, the average distance to the audience, etc.

While I understand why this is useful, I'm not sure I can tell what the answer will be. My aim for this is theatrical, but I'm fully aware that statistics are not in my favor, on this point. So television (presumably 3Net) seems like a much likelier target. However, do I want to set everything up for TV, if there's a possibility that I might be able to exhibit theatrically?

Further considerations: at least in the near term, my focus will be on creating a promo, which presumably will be viewed on a variety of TV screen sizes. Plus, I'm sure I'll also use my 8" autostereoscopic screen to show off the promo, and probably send some videos over the web. How in the world can one setting handle all those different applications?

So I'd love to hear recommendations: how do I choose what settings to use, knowing that there might be many different output versions, for different audiences? And how will I judge depth on my puny 23" monitor, if I'm working for an eventual projection on a much bigger screen?

[Bruce Allen]

Matt - sorry, missed this. Must check in here more often!

Yeah, screen size is a tough one. Three points:
1. you can push stuff further back on a small screen, whereas on a cinema screen you need to be really careful that stuff you're pushing back isn't so far that the eyes would have to go the opposite of cross-eyed to converge on it!
2. the 3D effect is definitely dependent on screen size and there is a real danger in designing stuff for a small screen that just doesn't work on a big screen. When I did the titles for the Hannah Montana 3D concert movie, I designed them to look good on my 24" monitor (ha). Then I tried a "more conservative" option just for safety. A quick trip to Fotokem later and seeing it projected there made me realize that the "more conservative" option was not conservative at all. Vince Pace very politely told me: "less is more!".
3. you can compensate slightly by sliding the small screen version away from the viewer in depth a little bit, and sliding the big screen version towards the viewer.

Personally, I have a 3D projector at home for that exact reason - so that I can see what stereo looks like on a 90" screen, which is close enough for me to get an idea about what it'll look like in theatres. Though with experience you can pretty much imagine how it will look - for Thor trailer titles, Smurfs, Rio, etc we were able to get stuff pretty close in the first pass. Still you do want to see it at the right size.

Basically though, you can't go back after you've shot and adjust the IO a lot to match a certain screen size (sure, Ocula, etc can do it but not 100% automatic operation for every shot). So you have to just sort of think of a target market screen size when shooting, sadly. If you target theatrical projection, your 3D ends up being a little conservative on TV and very conservative on a tablet. But those autostereo tablets have their own issues, so maybe being a little conservative isn't too bad.

For post, we used After Effects for all of the above title design projects. If Nuke didn't suck for design work and weren't so damn expensive when paired with Ocula, we'd be on it in a heartbeat though because its stereo support is wonderful and After Effects really doesn't know how to deal with stereo footage or renders well. You can duplicate your comp and drop in the other eye's footage... but when you start doing fun stuff with convergence adjustments and then decide to tweak the design, you have to either keep two comps in sync (harder than it looks, even with scripts)... or just delete the one eye, make the changes, then create the new eye again and repeat convergence adjustments. Painful. It does work though!

Bruce Allen
www.boacinema.com

[Matt Faw]

Thank you very much, Bruce, for such great feedback. I'll have to bookmark your post for a later date, and refer back to it, because I think those points will inform many more decisions down the road. At least in the short term, I'll be focused on creating a promo, which will primarily be seen on TVs, the web and my tablet. Hopefully, the promo will lead to a distribution deal, so I will know what my target screen will be, for all final work.

Of course, as you say, the final screen size needs to be considered even before shooting! Tough, since I've already shot my interviews... At least, they are all shot in front of greenscreen, so there should be some room for playing with convergence later (cross fingers). Both cameras I used (3DA1 + NX3D1) have fixed interaxials, so I'm not sure how much I could've changed my shooting style, in order to compensate for screen size.

At some point, as you say, I'm sure I'll have to invest in / rent / borrow a stereo projection system, if I want to prep the cut for a larger screen. CGI will be tougher, too, because the stereo rig I am using for 3D Studio Max only allows me to previz the stereo in sample frames, and then, only in anaglyphic (which is too forgiving, in my experience, to be a very good alignment tool). It's hard to know how things line up, until after I've rendered the project and brought it into After Effects.

I know what you mean about After Effects, for posting. Once I had everything set up for side-by-side, I realized that created a series of limitations for nesting comps, etc. For example, I couldn't nest a stereo comp inside of another stereo comp, without disastrous results. What I ended up doing was using individual stereo comps to converge each layer, and then combined them on a non-stereo comp, for the final composite image. And that (usually) worked for side-by-side, but I don't know yet how to use that for frame-packing outputs. Stereo really does make the complexity exponential, not just doubled!

Thanks again!