NOTE.. I wrote this a while ago but for some reason never posted it. I Just ‘rediscovered’ it and I think it’s worth throwing up here 🙂
So… the other day we got a gig which we planned to cut in Premiere. It was a short piece, a couple minutes long with about 8 takes, 2 cameras rolling on each. We got the sources, R3D 4k files, 1080 422 HQ ProRes transcodes, and multichannel WAV location audio.
As it was a 1080 delivery, the plan was to use the transcodes. Easy right? It was a quick turnaround gig, first cut due the middle of the next day.
Right off the bat there were a couple issues. We had a little difficulty synching the audio but got that sorted out. Next issue… We made multicams but, even though there are only 2 angles and we used the ProRes files, they were not playing smoothly at all. It may be a network bottleneck, but I made duplicates in FCP X and they played fine.
Also, the audio was a mess, difficult to tell who was on what track, and all of it was noisy and/or over modulated. Bad audio isn’t Premiere’s fault of course, but Premiere does use tracks, and Track Tetris with unlabelled multichannel clips sucks.
To save some headaches, I decided to sync the files in FCP X while they watched the footage and then export multichannel QT’s that they could work with. I brought in the files. The WAV’s were of course all nicely labelled with iXML channel names. Everyone was excited about that, until I told them that that would all go away when I exported the QT’s to use in Premiere.
The syncing took all of 5 minutes. I exported the files, the folks on Premiere now had files they could work with, and the editing got started. As I had the original synched files and had made multicams in FCP X, I started my own (very) rough cut. End of Day 1
I finished my cut, and the guys had finished theirs. We watched them through. Mine kinda sucked. I wasn’t “officially” working on this, but a sucky cut is a sucky cut. lol So… we now had a good rough cut to polish up. But sorting through and working with all the audio was still not easy in Premiere. Again, nothing unusual when working with tracks, but it was really slowing them down and the clock was ticking.
Because of the ease of working with multichannel audio in X, we decided to import the cut from Premiere and do the final audio editing in FCP.
WHY WORK LIKE THIS…
I got the xml from Pr, ran it through Send to X and imported it. The audio was messy and of course all of it was detached from the video and labelled A1, A2 etc. Effing tracks! Even though I had the original synched files with Roles in X, the sequence now linked to the QT’s I had made. No problem.
I batch set the Roles on the imported multichannel QT’s to match the original sources. That took 2 minutes. Then I went through the cut, matching back and replacing each clip. Since there’s no tracks/targeting needed in X, it was literally 2 keystrokes per clip, using the KB to select each clip. Another 3 minutes.
WHEN YOU CAN WORK LIKE THIS…
Now the fun part. Using the audio Inspector to clip-skim the subroles, enable/disable channels (components), expand clips in the TL to edit the components, and get the cleanest audio cut that I could. Fast and easy, as usual 😉 Added some bits we liked from my version to the cut, made a few tweaks, and met the deadline.
The next day we went through a few rounds of changes and eventually the cut locked. Since we’re finishing picture in house, it was time to do some simple grading. The Color tools in FCP made this a piece of cake.
Now, of course this could have been done in Premiere or any other “Pro” NLE. But, because of the time constraints, without FCP we would’ve likely had a nightmare on our hands. And it certainly wouldn’t have been any fun. And if you’re not having fun, why are you editing?
BUT WAIT, THERE’S MORE…
Actually, it definitely would have been a nightmare, for two reasons that popped up right before delivery. The first was a biggie. All the source files, the Premiere sequence, and the FCP X Project were 23.98 ProRes. But the delivery needed to be 59.94 DNX HD. Whee!
I have Media Composer, so the codec wasn’t a problem, but I was… concerned about the timebase change. I shouldn’t have been. I created a new 59.94 Project, copied the 23.98 timeline, and pasted it in. Perfect. FCP X did the conversion and every frame was exactly where it should be. Made a DNX share preset and exported. Done.
The last fun thing was during the mix. I had disabled various channels as I was editing, and created the AAF via X2Pro. The mixer called in the AM and wondered if I could send an AAF with all channels enabled on all clips.
Now, in Premiere or really any legacy NLE, I would’ve had to match back to each clip and overcut all the channels, carefully targeting tracks to offset overlapping clips etc. I have done this often in the olden days. It’s awful, and can take forever…
In FCP X? Easy. KB shortcut to select each clip in turn, enable all it’s channels in the Inspector, and export AAF. It took maybe 5 minutes.
So, once again, FCP X comes though. Now, to be fair, the original version was cut in in Pr, and it could have stayed there, it’s a perfectly good NLE. But it woulda sucked, and passing it to FCP X made everything easier.
So why does everyone still gravitate to Premiere or Avid? Here’s my theory. FCP X is easy to learn, but for seasoned editors it requires some “unlearning” and that takes some time. It’s also ‘missing’ a few features that pro’s are used to having. Not deal breakers to me, but Premiere/MC work like they always have so…
Welcome to Hollywood. Hey, I’m trying!