I'm doing some de-rig work on an animation, shot at around 4K on DSLRs, but delivered to me at something more like 2K, and eventually being delivered at HD. Occasionally the editor will decide to crop in to an image, and when that happens I get given the 4K frames and I do the work on the 4K image. At this point the editor has often removed frames, etc to speed up a particular action or whatever and so I also recreate the timings. But that only happens after the finecut is signed off, so we know there won't be further changes. Sometimes it happens to something that needed a little work at the original 2K size, and I just redo the work. No big problem.
This has happened on a couple of previous shots that had more involved work but only using masks, and I was able to pre-comp the source layer - leaving attributes behind - and overlay the 4K image in the pre-comp and using the collapse transformations function I could do the resize and it would use the 4K so the final sequence is not softened unnecesarily. I checked this by overlaying the 2K with the same transformations and there was a noticeable improvement going to the 4K pre-comp layer. Great!
But I've been sent a shot that took about a day and a half originally, because of issues with light, length of shot, amount of work needed, etc. I hoped I could repeat the technique described previously, but unfortunately I have layers with clone paint effects and layers that I have used time remapping on to make freeze frames to mask over areas and these are not playing ball. Everything looks fine until I nest my final comp and perform the neccesary resize. After the resize the clone paint strokes don't resize and move, same goes for the masked freeze frames. It's very upsetting! And I would really rather not go through all the work on the shot again from scratch, as at 4K I think it will take a very long time and it really seems like it shouldn't be neccesary.
I've tried doing the resize in a few different ways, and it seems to happen whatever. Any workaround anyone might have would be great!