I edited the movie, but I would like to emphasize that I did not edit this trailer. This is the so-called “red band” trailer, so put on your headphones if you’re at work.
In a silly–but wholly appropriate–stunt The Weinstein Company is releasing a movie I edited, Cheech & Chong’s Hey Watch This on 4/20. I’ve heard a rumor that it might come out in a few theaters as well. There’s no trailer yet, but you can pre-order the Blu-Ray for hardly any money on Amazon. I can assure you that it is a lot of fun.
If you want to see some very similar material (almost exactly the same material in fact) in animated form, check out the trailer for the new animated Cheech & Chong movie which apparently does for Cheech & Chong’s old comedy albums what HBO is doing for Ricky Gervais’ old podcasts.
Another fun little project. I edited this for Tom Papa’s new website.
I edited this promo for The Marriage Ref a couple weeks ago. It was shot entirely on the Canon 5D Mark II. It was a really fun experience, and Mr. Seinfeld came in a couple times to work with us. It’s great to see something I edited released so quickly but the quality of the encoded video is pretty low. It’s my understanding that this full version of the story will not be shown on TV.
I was cleaning out my closet yesterday and I found a VHS tape of the original 31-minute version of Camera Noise. It’s currently at a lean 29 minutes, and I was curious if there were any gems in there that I cut just to get the length down. It turns out I was definitely right to cut out everything I did, and of course the whole thing is almost unbearably long. But here’s a fun little bit from what’s still the best scene I ever shot.
At my current gig, I sometimes need to use videos generated by Snapz Pro. We recently ran into a problem where previously-working FCP timelines wouldn’t render and would generate the mysterious message “Codec not found. You may be using a compression type without the corresponding hardware card.” I had already found that capturing videos at 30fps in Snapz generally led to FCP reporting the frame rate as 10fps and resulting in generally unreliable editing. I could never be sure the in-point I chose in the viewer would actually be the in-point in the timeline. It’s important to run Snapz videos through Cinema Tools to conform them to 29.97 or whatever frame rate you’re working at. But the new problem seems to be related to resolution. With Snapz I invariably capture some bizarre resolution with the plan of re-framing it in FCP. These days there has to be a good reason for me to work in a codec other than ProRes, so I’ve been saving the Snapz videos as ProRes. Unfortunately, It seems that although Snapz can save the videos to ProRes, if it’s a weird enough resolution, FCP and Compressor are unable to generate new video. Compressor says “Processing service request error: Codec not installed.” I didn’t have this problem in the summer, so I suspect it’s related to a QuickTime update. Between this summer and now, the computers we’re working with were upgraded to QuickTime 7.6.4. The solution I’ve settled on is saving the Snapz videos in the Animation codec. The Animation codec doesn’t seem to have any resolution restrictions, and it seems well suited to screen captures anyway.
Traditionally, film editing has 3 distinct phases. You cut the picture. Then you cut & mix the sound. Then you “finish” by cutting the negative or doing an online edit. These days a lot of that work is combined. Sometimes your edit is actually at full resolution, so an online isn’t strictly necessary. Sometimes your sound editing can be done with the same software you edited picture in, and by the same person.
But other times you need the special expertise of a sound editor, and the special tools that only specialized audio software like Pro Tools can offer. So today I’m going to talk very specifically about the steps you should go through to prepare a film for hand-off to a sound editor.
- Talk to your sound editor. Ask him what he wants. Don’t be afraid to ask stupid questions. Maybe there’s even a checklist he can provide you. Talk before, during, and after preparation of these materials.
- Lock your goddamn picture! I can’t stress this enough. You must have enough time in your schedule to finish editing the picture before you give it to the sound department. Peter Jackson can afford to do endless conforms, including editing the picture during the mix, but if you’re reading this I can assure you that you can’t. It might seem so simple to just make a few changes, but those little changes that are so easy to make in your picture editing software ripple out into multiple hours of work for everyone down the line. Consider how many days of sound editing you’ve budgeted for and ask yourself if you’d rather have your sound editor spend that time (which is already too little, you know) working on the sound or working on implementing the picture changes you’ve made. Also sound editors don’t like conforms. You’re probably not paying these people enough as it is. Keep them happy.
- Break up your movie into reels. I’ll be honest with you. This one is dying out. There are a lot of times where this just isn’t necessary, but I’m going to talk about it anyway. Some of you might not remember this, but movies used to be shot on giant strands of plastic, and shown in theaters that way too. It turned out that putting a whole movie on one giant spool made it hard to move around, so they were broken down into 2000 foot reels (about 22 min). While your job no longer includes renting a truck in order to deliver all your editorial materials, you can still benefit from the wisdom of the ancients. If for some reason your film is shown on film, you’ll be setup for it from the beginning, but there are actually some good technical reasons to work in reels. If you happen to do color correction in Apple’s Color program, it has trouble with projects that have too many cuts. It will also help you get around the pesky OMF file size limits. And I find that it helps psychologically to be able to say during the mix “Reel 1 is finished” rather than “We’re 20 minutes into this 120 minute movie and we’re already in overtime.”
A few tips for breaking down into reels: Keep it under 22 minutes. Don’t try to squeeze it right up to 22 either. Nothing wrong with going under 20. Often the first reel has to be shorter (1600 feet, 17 min 46 sec) to accommodate things like trailers. Try to avoid having black at the head or tail of the reel (except for the head of the first or tail of the last) because projectionists might cut off the black bits if they’re not paying close attention. I once got an email from a projectionist at Sundance who just wanted to make sure we did it on purpose. Other projectionists won’t have your email address. The best place to end a reel is at the end of a scene that has quiet audio. If loud sounds carry over between the reels, there might be a problem. Probably a bit of a pop or click. Certainly don’t let any music cross over the change. But these days pretty much anything goes, since the reel change during projection will be frame accurate.
Each reel starts at a different timecode. There are two different ways to do it. The more “filmy” convention is to start the first reel at 01:00:00:00, the second reel at 02:00:00:00, etc. The more videoish way is to start the first reel at 00:59:52:00, which puts the First Frame of Action after the 8-second countdown (see below) at 01:00:00:00. Reel 2 would start at 01:59:52:00, etc. Either way is fine as long as everyone is on the same page.
- Countdown with 2-beep at the head of each reel. Final Cut Pro actually comes with a great countdown. It’s on the install disk in the Extras/Head Leaders for Cinema Tools folder. You can use the already-generated ones or open up the project and adjust it to your particular frame rate and resolution. I also recommend turning off the stupid flicker. The countdown includes 2 seconds of black after the “2.” Depending on the timecode style you’re using that means your First Frame of Action is at 01:00:08:00 or 01:00:00:00. The key thing here is that there is a one-frame beep at the “2.” Put this beep on every single audio track in your timeline. That way when your sound editor imports all of the files you generate, he or she can tell right away that they are in sync because the beep happens on the same frame as the 2. If that doesn’t happen, you know you have a problem. Of course timecode makes this less necessary than it used to be, but it’s a nice simple way that a human can tell things are working the way they should.
It’s an optional step for most workflows, but I also like to put a beep at the end of the reel. In the old days you would actually use a hole-punch on the frame exactly 2 seconds after the Last Frame of Action, along with a one-frame beep on every track of audio. I put in a virtual hole punch by making a one-frame white circle. This lets you know if any tracks have drifted out of sync for some reason. It’s unlikely to happen these days, but it certainly doesn’t hurt.
- Audio reference files. The QT you output will include a stereo reference track of the audio work you’ve done so far. Sometimes the sound editor will want the mix broken up into separate dialogue, effects, and music files. Just turn off the tracks you don’t want, and export the ones you do. Preparing these files can also help you organize your tracks better than the mess you’ve made during picture editing.
- All original audio files. Usually the sound editor will want these. If you shot double system you already have them ready to go, but if your audio came in with the video files, you’ll have an issue. The simplest way to generate this stuff is to select all your clips in a bin and do an audio-only batch export.
- Export OMFs. This is pretty easy. FCP & Avid both have simple methods of generating OMFs. You’ll have the option of setting handle lengths. This is the amount of audio media you want to include in the OMF before and after each cut. It gives you flexibility in the sound editing to extend audio to cover up seams, or to find bits of room tone to fill in holes, or any number of little tricks that will make life easier. Make the handles big. A minute is good. More if you can stand it. One thing that’s going to limit you is the archaic OMF standard which restricts the file size to 2GB; a number so large that people in the 90s couldn’t even count that high. In FCP you can only export OMFs with embedded audio, so if you have 24-bit audio, big handles, and a lot of tracks, you’re probably going to bump into this one, even if you’re only working in 20 minute chunks. The easy way around this is to simply turn off a number of the tracks (using the green dot next to each track) until you get the file size below 2GB. Rename the exported OMF so that it indicates which tracks are in the file, then turn off those tracks and turn on the ones you had turned off and repeat. You might have to do more than two OMFs. I recently had to export three or four OMFs per reel for a movie I was editing. If you’re working on Avid you have more options. You can actually export an OMF that only references the media, so the file size limitation doesn’t really come into play. Make sure you check with your sound editors before you do that though. They might want embedded audio. Also, in Avid if you’re working with certain media types you can only export AAFs. It’s pretty much the same deal, but without any file size limitations. Once again, check with your sound editor to make sure you’re generating the right kind of file.
- Generate QuickTime reference files. You may be working in 9K with 7:7:7 92-bit log color, but your sound editor won’t be impressed by that. Your sound editor is getting by fine on a G4 and a 500GB hard drive. As always, deliver what is requested, but what is usually requested is an NTSC DV QuickTime file. You can probably get away with 23.98 if you’re cutting with that framerate. There are a few things you can do to make things easier for everyone. Add a visual timecode track with the timeline’s timecode. Don’t make it too huge, and put it in a letterboxed area if you have the option. In FCP you should get Andy’s Timecode Generator, which lets you add a generator to a video track rather than applying it as a filter to a nested sequence. It’s definitely easier that way. Avid has that functionality built-in. If your footage doesn’t already have source timecode burned in, you might want to apply a Timecode Reader filter in FCP to all your clips. Double-click it, set the size and location, select all, and then drag it onto the clips. This can be useful if your video TC matches your audio TC and you’re looking for a particular piece of audio. It’s not always necessary.
Export in the format requested, and include audio in the file.
- Audio EDLs. You may or may not need to make EDLs. Sometimes the sound department needs to replace the junky low-quality audio you were working with, although as always, this sort of offline/online workflow is less common now than it was a few years ago. It depends on your workflow. As always, ask, ask, ask. FCP can export EDLs. With Avid, use EDL Manager.
Ok, now you’re done. Put all this stuff on an external hard drive and get it to your sound editor. And remember that if you make any changes to the picture you’re going to have to do most of this all over again.
I recently upgraded my editing monitor, since the 13″ Sylvania I bought for $50 in 2002 really wasn’t cutting it anymore. I’d been planning it for years, but since I edit outside of my office so often, I was putting it off until a job came along to will help me pay for it. That never really happened, but I did it anyway. I moved my 37″ 9UK Panasonic professional plasma into my office and mounted it on the wall so it’s not right up in my face and clients can watch it comfortably from the couch. I bought a new 46″ G10 Panasonic consumer plasma for the living room. I only really wanted 42″, but 46″ was barely any more money! It’s huge! It’s also 1080p, and actually has HDMI inputs, so I see a real boost in picture quality for Blu Ray discs. There’s a general consensus around the Internets that the 48hz mode for 24p input causes too much flicker, but I think it looks pretty good on most images. Bright graphics definitely flicker like a PAL CRT though. I’ve only watched a blu ray discs with it turned on, so we’ll see how I feel about it after more testing.
Anyway, the real point of this is that in the course of my research I stumbled across the latest HDMI marketing gimmick known as Deep Color, which uses 12-bit-per-channel color (36-bit total, or 68.7 billion colors). Now I’m certainly a fan of high quality images, but I had never even heard of 12-bit video. I’ve mastered movies in 10-bit color whenever possible, which I thought was great. I’ve worked with plenty of 8-bit source material, and I’ve seen its limitations, but it still looks really good. There are always going to be people who want more though. And I guess 12-bit color is going to deliver it for us.
But what does this mean for content producers? I certainly don’t anticipate finishing a film in 12-bit color any time soon. A cursory search shows no widely available tape formats that can hold 12-bit color. HDCAM SR only goes to 10-bit. The Red One shoots 12-bit color, as do some other digital cameras, so theoretically a fully tapeless workflow could accommodate 12-bits through the whole process. But then how do you deliver it to consumers? I can’t imagine there are any cable/satellite signals even broadcasting 10-bit. The broadcasting trend seems to be going for lower data rates, not higher. Blu Ray in its current incarnation can’t pull it off, but some future form of digitally distributed media could. However, like cable and satellite, the trend is towards more compression and lower bit rates. 12-bit color files will be HUGE. I can see where this could be a good thing, but it seems a little like 240hz LCD screens: a higher number that’s just used to justify increased prices.
This is a collection of YouTube videos that highlight projects I’ve worked on in the past few years, as an editor, trailer editor, assistant editor, post supervisor, title designer, or even boom operator. Enjoy.