Adam Berk
Forum Replies Created
-
Because of this, I have based our workflow on using pre-conforms and pre-conforms only. We do spots, no long form stuff, so it’s not a big deal for us to just output a new pre-conform when vfx are completed and replace the image sequence right under resolve’s nose. I find it quite ridiculous that doing ANYTHING other than a basemem or delete node can cause one to lose grades. I also find it ridiculous that there seems to be no way to lock down a conform to make sure resolve doesn’t decide on its own to start linking shots to different or newly added media.
-
If anyone’s wondering why I needed to do such a thing….
I was manually color tracing from another timeline and the animation came in a bit off. Four frames off to be exact. This sync error was caused by the fact that I was using pre-conforms based on EDL’s that came out of the scene detect module (no source timecodes), and the shot I was tracing to didn’t have exactly the same frames or length as the shot I was tracing from. Seems like something that could be a fairly common scenario in rush grading pickup shots, no? It certainly can be for me.
-
I come from smoke/flame, and there I’m used to a switch that lets one view and edit animation channels either as keyframe data (keyframes with curves between them) or tracks (with which you can slip, compress or expand entire animation channels). I’m not sure, but I would assume that it works the same or very similar in Lustre as well.
Another simple option would simply be to allow us to multi-select the marks! I can’t believe that I can only select one mark at a time (shift and command do nothing).
-
+1 on this for sure. On superless projects that come in as online pre-conforms, it totally makes sense to be able to do the layback right in resolve. Kind of silly to tie up our smoke or flame just for a layback on something that comes out of resolve as final picture. Most especially when deliverables are files/digital only (no tapes). At least for us, layback used to be something that was done wherever our client was mixing, but these days we’re almost always the ones taking care of it.
-
A proper CMX3600 EDL should have reel names and TC for each shot. Do this in FCP by selecting each shot used in your sequence in the bin and adding reel names and TC to those shots. Your sequence should be conformed to a single layer. If multiple layers are required you will need to export an EDL for each layer.
Make SURE your reel names are comprised of letters and numbers only, and that they are 8 characters or less in length. Make sure you have no two shots that share TC on the same reel. Also, get rid of your audio on the prepped sequence and instead just export your rough mix as a wav or aiff and send it along with the edl. When exporting the EDL(s) turn audio mapping to none.
Be sure to start your program at an even hour and to include a 2-pop frame (in the picture) and sync pop (in the audio) exactly 2second before the even hour.
You should include a reference picture quicktime with the edl and roughmix audio files.
Another thing to make sure of, watch out for any single shot that has cuts in it for no reason. If you’ve got a shot in your sequence that has a cut right in the middle, take out the cut. There is no reason for it to be there. I dont understand why editors do this. So irritating.
I can tell you that there are very few things during the course of a day at a vfx/grading shop that come anywhere close to as annoying as badly prepped edl’s from the offline company. Your finishing artists and colorists will be much kinder to you if you can master the art of prepping an edl 🙂
-
Can probably just issue a shake terminal command to file-out the FCP edited PR4444 file as DPX as well. Then chop it up with the EDL once it’s in flame. No need to involve the Resolve at all in that case.
It will be interesting to see if FCP X will fix any of these age old quality issues that have plagued FCP and quicktime since the beginning.
Anyways. Thanks so much again to everyone for their responses and input on the weekend.
-
Sorry if I’ve confused you.
My goal is to pull selects from the PR4444 media without decompressing, then decompress only those selects. Simple as that. We are not stuck using any one tool, though it is much easier to give this task to an assistant to perform in FCP in the machine room than it is to have to involve multiple systems in rooms that are booked (resolve, flame, etc). The goal here is to have the uncompressed selects sitting on the SAN monday morning for me to start ingesting into flame.
-
What is UC?
-
To avoid generating a massive amount of unnecessarily duplicated uncompressed media.
-
This is certainly another option, though the rounding errors produced by the float conversion with Nvidia gfx boards in Color are not too desirable. We have no systems with ATI graphics adapters, without which one cannot render in less than float/more than 8bits.