[Mark Suszko]

Just a Google away…

https://www.digitalrebellion.com/webapps/videocalc

[Mark Suszko]

Considering how sloppy some local cable companies can be, and the relative age of their play-out automation gear, my guess would be their play-out system needs one second of black “handles” front and back, to allow for dissolve transitions between other programming segments.

[Mark Suszko]

Maybe a dumb question, but why not turn your B&W masks into green and white ones, then you can chromakey one part and luma key the other.

Also, have you tried different blending modes with these combos?

[Mark Suszko]

In FCPX you might handle this by making compound clips that just contain the relevant parts. Stacks of compound clips allow independent manipulation.

in Motion, I’d want to look at how you use grouping to do the same thing. Grouping or nesting can be very powerful in letting you make adjustments in one place that don’t interfere with others..

[Mark Suszko]

Well, when you’re filling out the info in the “share” panel, there are blanks below the title that that auto-fill with the name of your project, etc., and if that’s what he’s talking about showing up during playbacks, maybe going into those blanks and back-spacing over the data there to wipe it out will solve it? I could be thinking of something else, but I hope this works.

[Mark Suszko]

Sign me up; this bother me too. It’s like FCPX is trying to be helpful and anticipate that adding new media must mean I’m ready to start importing it, but that’s almost never the case with me. I’m P2-card based, and I always first create a folder in my RAID where I’m going to offload the P2 car prior to my edits. I’m not even halfway done labeling that folder when helpful FCPX pops up the ingest dialog box, like an insistent 2-year-old. Happens when introducing almost any new memory device to the computer… “down, boy!” I wanna just tell it to cool it’s jets and wait for a command.

[Mark Suszko]

This may be something you could hire out to The Mechanical Turk to get it done faster,by throwing many people at it at once, but it won’t be cheap.

Microsoft once demoed something a few years back: it was an A.I.-driven “automatic home movie maker”. It would take a long file of raw video and automatically make cuts based on length of shot, how much things in the shot changed, if new angles appeared. It wasn’t very artful at doing this, kinda like the old saw about the dancing bear. I never saw it come to market; just the demo.

The audio changes are less daunting and more automate-able than automatically sensing where to put ins and outs on a continuous string-out of raw footage. QT 7 or QT Pro can modify files where they sit to delete audio, I think… Trying to think of ways to speed this along, I imagine some kind of hybrid system, where you play the string-out in real-time thru a switcher to another digitizing station. It still doesn’t really work.

Mpeg streamclip could take a long single file of these cuts, and let you skim in JKL mode to place ins and outs, then you’d hit “export” and render out single short segments from that master, like slices off a salami. You can render without audio. So you get a folder of silent, individual clips out of that. I’m guessing someone good at using FFmpeg could maybe program a batch operation to marry each silent clip to the same length audio track, and output to a folder. None of that is anything near real-time, though.

[Mark Suszko]

Naw. I know just enough to build one, though.

[Mark Suszko]

Sounds like a setting mistake somewhere, but I couldn’t say where. Something I’ve had to do to “airbrush” a screen flaw is, duplicate the layer, stack it above, crop it down to an area of clean white space above or below the offending line, feather the crop, move it down or up or whatever to cover the target.

A more aggressive approach might be to screen-capture the frame, take it into Photoshop or GIMP, sample the clean area and then apply it universally to create a blank full page, bring that back into FCPX and make a composite of all your needed masking against that. Or re-compose the text in Photoshop and just key the hand back in over it.

[Mark Suszko]

*Today*, it’s a gimmick…

But I just imagined a scenario where you asked every person in a football stadium to whip out their phone and video the halftime show at the same moment, then use a supercomputer to synch every individual stream, map them to a sphere and create a virtual 3-D VR view that could be rotated, zoomed, and focused at will…