NP!
Currently, I have to create pixel accurate framing charts for any of our high end jobs we do at Sim. As you know, the cameras framing charts are never pixel accurate. They are a good reference, but that’s it. We do more TV than features here, so it’s often less about many framing options for each camera, but plenty of cameras to deal with. So for one of the new Netflix shows I’m on right now as an example, which has actually been pretty light on various cameras, I’ve created a chart for the:
- AlexaMini_OpenGate_3424x2202
- Phantom4K
- RED5KHD
- SonyVenice4K
I create these diagrams, people create their framing presets knowing my chart is pixel accurate, then they apply said framing preset on the camera recorded chart, and we see how it lines up. Its never bang on, because again…camera charts never are. But its a great reference!
So to me, the goal here would be to stop having humans manually choosing how much to zoom in, pan/tilt, crop, matte, etc… Have the software read the framing preset, apply it. And then you could even save that as a preset in your software.
Because not every job has people like you and I making these pixel accurate charts, dailies may frame one way, VFX then re-creates the framing trying to match the QT Reference. But maybe dailies f-ed up the framing and now VFX is just matching it, but no one hears about it because they are both wrong. haha. Anyhow;
I think we’re on the same page for the goals 
But what I’m curious to hear is why the example I gave would not allow software to make the crop/zoom, etc to match the frame from on set? Where I wrote extraction frame, really what this meant is; this is the area that is intended to be active. If this active area happens to be 2.39, and you throw it into a 1.78 window, the software would know to matte it.
My thought had been that in camera, you set up your frame lines, the same way you currently do. These are each then an extraction name set of parameters from the saved XML. But in camera you could choose to use any of these 4 optional frame lines. I have never actually created these frame lines though and am not sure how they currently correlate with what gets saved into an XML. So if you’re saying this way is not do-able, copy that. I don’t think I fully get it yet though.
Maybe the disconnect is that I don’t understand the difference between your point about the framing on set, vs the crop? Is this as an example people framing on set for 1.78, but the crop would bring it down into a 2.39 as a different frame? If so, wouldn’t the 1.78 by one frame line you set in camera, and then the 2.39 be another. Then we have both?
I assume I’m missing something, but thought I’d also offer up the idea of chatting over the phone, versus continuing in this thread?
P.S. I totally hear your point about lots of vendors. We have 28 right now on Watchmen…Its nuts!
Cheers