[Frameworks] another quicktime export question

[Dave Tetzlaff]

The issue is not 'what looks good', but what's the best looking thing you can send out to to an ad hoc venue like a festival — where they typically have limited tech resources and clueless people manning the projections — with the highest level of assurance that a) they can actually get the work up on the screen, and b) they won't screw up the display.

There's no standardization out there. You have no idea what the people you're sending stuff to are using for hardware or software.

I find the suggestions for dcp etc. amusing, since the reality is there are plenty of people out there still sending out stuff on SD-DVD because they either can't figure out how to make usable HD files in any format and/or have had bad experiences at screenings with their HD versions and have fallen back to SD-DVD as the least-common-denominator they can trust to get SOMETHING up on the screen at showtime.

(Are significant numbers of screening venues really that bad, you may ask? The answer, alas, is you still can't even rely on your SD-DVD being shown in the correct aspect ratio...) 

You can trust a venue that shows work year round, e.g. Anthology or The Siskel Center, to get HD digital files right. They'll have someone on regular staff who knows what they're doing, and will be able to answer advance questions about formats, tech specs and whatnot. On the other hand, the only once-a-year-type screenings I've ever been to that DIDN'T have a projection screw-up were ones I ran myself. If Lincoln Center can't consistently get it right for Views - and they can't - what can you expect from any festival? (I know there are some fest folks out there who DO have their act together, but I remain convinced they're in the minority.)

As Aaron noted earlier, more venues use Macs than PCs. Thus, you've got a decent shot with anything you can put in a Quicktime container. I suggested ProRes for short works, but of course that's only really an option for people working on Macs, not PCs. So few festivals are going to handle DCP that making such a file is a waste of time unless you have an acceptance from a very important venue you KNOW can handle it, and has the projectors to realize the difference on screen. But even Quicktime isn't universal, and some fool could be running a screening on a PC that's only equipped to handle .AVIs.

Mass market video technology is driven by streaming media, and high-compression capture. To my eye, images from high-end DSLRs like the 5D MIII look mighty sweet, and they're throwing what are basically variants of H.264 onto flash memory. Tell me, oh video sages, if you have an H.264 file in an .mp4 container under what conditions will any common software/hardware combo screw it up and fail to display it properly? 

If you can name a more RELIABLE format to send out into the unwashed and unknown, please make your argument.

H.264 does not have to suck. It is highly scalable. If your H.264 files look crappy, you're using a bad software encoder or you've got the settings wrong. Given the existence of a mature, high-quality, freeware, cross-platform encoding program like Handbrake, the only excuse for ugly H.264 files is ignorance. Encoding H.264 isn't a just-hit-the-default-button thing. There's a learning curve, and some of the settings you need to control in Handbrake aren't intuitive or part of the main interface So you have to figure out what's where in some of the more obscure menus. There's actually not that many things you have to set, or that many variables you'd want to change from one project to another. It just takes some time to figure out where those things are, and what values you want to use on a regular basis for different purposes. The render times for high-quality output are long, too, as a lot of data is getting crunched figuring out how to create accurate P-frames and B-frames. (You have to use a two-pass method to get good results, in my experience anyway.)

But the various time investments have a benefit for anyone sending out experimental work in digital form. You wind up with a 'master' file that's compressed enough to distribute easily, good enough to look as good as most projectors out there are going to allow, and that most venues will be able to play back without f-ups. So you don't have to keep making new output files in all kinds of different formats that might or might not work when they reach their destination. Of course, you still have a big master output file on your computer in whatever format is native to your edit timeline -- ProRes or whatever they use on PC. And if you do want/need to send some other format to some other venue, you can always generate that from your best-quality master. 

I don't always share Aaron's views on tech matters, but I do feel some of his pain at the lack of understanding of digital tech fundamentals among moving image artists who are now working in that realm (happily or reluctantly...) A comparable lack of knowledge in photochemical film would be an embarrassment to anyone making personal work with those tools. Maybe the difference is that it would actually be disabling to photochemical work if you didn't understand how movie cameras work, the differences between film stocks, how to talk to lab techs, etc. etc. But digital stuff is so much more complicated at base that it has to be engineered with some kind simplifying interface in order for ANYONE to be able to use it, and the designers toss in enough simple pathways that newbies can get by. So the newbies do JUST get by, and they never realize what they're missing, or the problems they're causing themselves, by not understanding the simpler stuff on the 'Advanced' menu. 

E.g. I'd guess that the majority of makers and screening programmers on this list don't understand the difference between a container and a codec. It's actually pretty simple. And you do need to know it. But the industry's not going to tell you how important it is, nor or they going to explain it to you. But that's how it's always been for independent artists, like forever. They need to know their tools, and because they're making art, not product, they need to know the tools in ways that require a certain level of figuring-out, for which no hand holding is available.