Here’s one I run into fairly often. Maybe someone out there has a clever solve?
I’ve got a camera track for my shot that is Arri Open Gate (4608x3164). The track is great for fixes, compositing and cleanup. Once all that business is done, I render all that down and drop the clip on to my HD timeline. My client loves this shot. Wants to marry this shot. Likes the shot so much he carries a heart-shaped locket with a picture of this shot it. I digress.
Now I want to place a graphic in space in the shot now. So.. I could go back in to the source resolution and look at it in context to see how my graphic will look and feel in HD… but maybe there’s a better way? Has anyone tried to crack this nut? What I’d love to do is take my 4.6k track and somehow translate it to HD so that I don’t need to render my tiny little graphic into such a large canvas. Who among you has the answer to this age old question?
Render out a UV map as part of your original comp either as sep sequence or as a channel in a multichannel write. Turn your tiny graphics into a BFX and use the UV map with a resize to comp size, a 2d transform for position/scale and an Stmap node to get the movement. Low cost computationally and works on the timeline.
I might be missing something, but I don’t think that math works. I’ve definitely done the tiny graphic route many times. It works for sure, I just think that there might be a better way…
I think it was off because I did a simple ratio of horizontal sizes. Cameras tend to be calculated on vertical in flame. Also, I think the difference in aspect ratios between the original and HD don’t help. I did a fast test going from the original to 1920x1318 (matching aspect ratio), and put a 34.13% reduction on the camera and, at first glance it worked. I’ve done this in the past with simple cameras without analysis.
Much appreciated Tim! I think if you maintain the aspect ratio you can get it to work without really changing much. Things will likely line up pretty well. From all of my head scratching, I don’t think there’s really a way of scaling the camera to adjust for the aspect ratio change especially if you have a moving camera. If you scale the camera, you are scaling the camera move as well and that won’t really work out so well. If you have a static camera then grab yourself a red jacket that says “Ferrari” on the back because you are already off to the races.
I can keep doing it the way I’ve been doing it and likely the way that others do it too (or Christopher’s UV map method which I’ve done a variation of as well). It’s not the end of the world by any means.
Just something I’ve been thinking about for a while…. A kid can dream (*looks off longingly)
Seriously appreciate the thoughts and suggestions!
Action also does funny things depending on how you copy a camera from one action to another. If you copy the action, then change raster size, it changes the default distance in the camera to match the default FOV based on raster height. If you copy/paste the camera into an action of different raster size, it maintains its original data. I have a feeling it might force the FOV into what it thinks is proper depending on raster height. I remember looking into it a few years back but never followed through since I do very little camera tracking.