Or is it a new camera? The new RedHydrogen One phone is intriguing… And apparently real, not just vaporware. Holographic capabilities, insane computing power, and more expensive than my first car by far. Will it lead to a revolution? Interesting times, at least.
It boils down to objectivity, to getting away from all the baggage of remembering what happened on set. He makes a compelling case, and his advice is the conventional wisdom. It reminds me of WIlliam Faulkner’s advice on editing yourself: “In writing, you must kill all your darlings.” He meant that you need to let go of your attachment to any one little bit, to focus on the needs of the story.
And yet… There are quite a few noted filmmakers who edit their own work. The Cohen brothers spring first to mind. But also Akira Kurosawa and Gus van Sant. And quite a few documentary filmmakers edit their own work, although perhaps as much because of a lack of budget as from creative considerations. Here is a short list of filmmakers who edit:
I think most of the filmmakers I admire – Doc and Narrative alike – are ADHD. That is, have attention-deficit hyperactivity disorder. Enough so that it should maybe be called attention-deficit hyperactivity superpower instead…
It postulates just that – that ADHD may have some incredible positive sides, and that our educational industrial complex ought to accommodate. It makes sense to me. Many of my idols had great difficulty fitting in to a 9-5 educational model: Jerry Garcia, Joe Lex, Albert Einstein, Alan Steinbach, Walt Disney, Alfred Hitchcock, Ansel Adams, Whoopi Goldberg, Paris Hilton…
OK, not Paris. Whats up with her?
Anyhow, it makes sense, right? Attention flitting back and forth… The ability to hyper-focus on something you are really interested in. Very diverse interests… It fits filmmaking perfectly. Think about what a filmmaker needs to include in her repertoire:
Idea development from the earliest, ill-formed stages.
I tried it again. Final Cut Pro X, Apple’s oops of a slap in the face to professional editors and post production workers – especially those who depend on collaborative workflows. One of the basic phases of post production is audio mixing. As a former audio engineer, this is my comfort zone. In normal NLEs (non-linear editors), you can export the audio tracks, then bring them in to a DAW (a Digital Audio Workstation) for mixing and sweetening. In narrative film work, you might also re-record some dialogue (automatic dialogue replacement, or ADR), or record Foley (like footsteps and door slams). This workflow would be something like: Read More…
I was supposed to be editing… I had been fighting gremlins for weeks – some of my own making, many that appear to be bugsidiosyncrasies features in Adobe Premiere Pro CC. You know, Adobe – the company who got themselves hacked, meaning my credit card number stolen, so that I had to update every other utility and such that had that number. Maybe those folks are onto something, when they advise that your editing machine not be connected to the internet. But, honest, I was only looking at documentary film related news!
I just saw the list, via the Below the Line newsletter, of the short-form documentary films that will be vying for an Oscar this year. This is the short list, later to be winnowed to five or so before the Academy of Motion Picture Arts and Sciences votes. Hopefully, a DVD will be produced that gathers them all for our viewing pleasure. The whole slate looks intriguing, as usual. Here are a couple that caught my eye…
I was most intrigued by the trailer for Jujitsu-ing Reality, a short doc directed and edited by Chetin Chabuk and lensed by Chad Mann. It tells the story of Scott Lew, a screenwriter (most recently Sexy Evil Genius) who continues to write, using adaptive technologies and help from others, while suffering from advanced ALS (Lou Gehrig’s Disease). I don’t think it looks at what is possible for less privileged people facing such disabilities, but it sounds inspiring.
Another intriguing entry is Cave Digger by Jeffrey Karoff and filmed by DP Anghel Decca, edited by Erin Nordstrom The film is about Ra Paulette and his ‘art caves’ dug into the sandstone cliffs of Northern New Mexico. An iconoclast, he is embarked on a decade-long project. How does he pay for it?
None of these films seem very daring, let alone experimental. But it is the Academy, after all…
So we did this little shoot. Three cameras, plus multitrack audio. Slated most of the takes, like I said…
Here is the workflow we evolved:
Shoot seven takes total of three different pieces by the aerialist team, Charity and Jon.
Broadcast wav in multitrack Pro Tools sessions for audio
Canon AVCHD files from two Canon Vixia cameras
Canon MXF files from the XF 300.
Manually edit any clips that spanned two takes (i.e., where we left the cameras rolling between takes) in order to create usable chunks for Plural Eyes. Otherwise, with multiple takes using the same backing tracks, it all got messed up. I used MPEG Streamclip, but it couldn’t edit the native Canon MXF files. So I had to first use Sorenson Squeeze to transcode to ProRes – I could as well have used Apple Compressor or Adobe Media Encoder, I suppose.
Manually align one or two media clips that Plural Eyes couldn’t handle (most synched fine).
Export from Plural Eyes to Premiere Pro flavored XML.
Ingest a total of seven sequences back into Premiere Pro.
Rearrange tracks to standard order of cameras and audio tracks…
Select all, then “Edit in Audition” to fly the audio tracks out to Adobe Audition. I couldn’t easily export to ProTools – Premiere couldn’t’ extract the audio from some of the camera files to export separately, I think. Also, Premiere didn’t send audio track names to Audition – they seemed to go ‘poof.’
Save Multitrack audio mix for each take (sequence) from Audition.
Back in Premiere, the next step is to Nest these synchronized sequences for multicam editing.
Before Nesting, though, add color correction to make the cameras match (and cut together nicely).
Unfortunately, with this process you need to mix audio down to just two tracks – not exactly conducive to the normal workflow of edit -> lock picture -> only then send out audio for sound mixing. Hence the extraction of audio tracks first…
Also, the Nest command doesn’t behave as I expect. It opens a dialog, asking you to pick a name for the nested sequence. But that sequence is the original, but with blank tracks removed. Then it actually nests the sequence you originally had open… Sheesh!
We recently did a small multicam shoot for some circus artist friends, and were reminded not to cut corners in prep… Specifically, don’t rely on software to sync your clips for you! It is worth taking the time to carefully and correctly slate every take. Editing, and the entire post process, will be so much smoother if you do.
I have shot multi-cam plenty of times. My last big project was a school musical: three shows, three cameras per show, plus 14 channels of audio direct to disk with ProTools. There were issues, there are always issues… But we were careful to slate every single take, so it moved pretty smoothly.
The biggest hassle with that production was moving audio from Avid Pro Tools into Avid Media Composer! Yes, I know, same company… The history is pretty divergent, Avid having swallowed ProTools’ creator, DigiDesign, some years ago. It is a pretty recent development that you can even run Pro Tools and Media Composer on the same computer. But editing on Avid, then exporting the audio to ProTools, then finally finishing with the mix from ProTools is a pretty standard workflow. What hung me up at first was trying to go from Pro Tools to Media Composer using AAF files (the OMF successor) that just pointed at the actual audio files on disc. It is supposed to work, but I couldn’t make it… Once I gave up and embedded the audio in the multitrack AAF file, it went smoothly.
Hey, hard disk space is cheap.
Media Slate really is the premier slating app…
With this shoot, we were using a similar setup. Fewer mics, but also lighting the performers in a challenging space. So when we were ready to role, and the updated version of Movie*Slate wasn’t clapping at first, we steamrolled on ahead. I figured that software, either Adobe Premiere Pro (where we planned to edit) or Red Giant’s Plural Eyes, or maybe even iMovie Pro X (also known as Final Cut Pro X), would handily sync up the various clips and files. This was true – but at the cost of hours of mucking around that we would have avoided if we had a good slate for every take.
If you don’t know, the slate, classically using a clapper board, is actually a vital part of film production. Not an affectation at all. The slate uniquely identifies each and every bit of film, video, or audio in a production, and provides a precise sync mark for lining things up. It doesn’t matter as much if you are shooting with a single camera. But with three separate video clips and three or four audio clips for each take…
Plural Eyes can sync clips by matching audio
All of the software I mentioned can now do what Plural Eyes pioneered: sync all these files up by matching the audio. Automatically. In theory… I have had fairly good luck with Premiere Pro and Final Cut X. But they are both black boxes – they either get it in sync, or fail. No tweaking. Plural eyes is a bit more deep and robust, and this project needed that extra help. In principle, I should have been able to ingest everything into Premiere Pro, then export a sequence to Plural Eyes, finally re-importing the result. If only…
First, I ran into a version incompatibility: Plural Eyes doesn’t seem able to read Premiere Pro CC projects, only older versions. I fussed around with that for a while before I gave up. But I could export a Final Cut Pro 7 XML file from Premiere, then bring that into Plural Eyes. Problem solved….
Then, I ran into some bugs in Premiere Pro. One of the basic features in a modern non-linear editing program (not just Media Composer or Premiere Pro – also iMovie and Windows Movie Maker and such) is making subclips. For example, we had let our #3 camera run continuously – it was up in the balcony, after all. So we ingested that entire big clip, then used markers to cut it up into subclips. Except Premiere Pro had a bad case of gas over this! Instead of exporting the subclips, it kept sending the entire, original clip. And Plural Eyes was confused by this big wad of material.
The timeline screenshot
What’s a filmmaker to do? Well, first I wasted a bunch more time. Eventually, though, I fired up MPEG Streamclip, the fantastic video Swiss Army Knife by Squared 5. This gem let me manually create and save subclips, dicing up the original video files, which I then sent to Plural Eyes. In the end, all but one of the takes were perfectly synchronized by Plural Eyes – one clip in one take I had to do by hand. That is tricky – one of the aerialists is a violinist, and lining up a legato instrument is much harder than something more percussive (like drums or even piano or guitar). And, it turns out, Plural Eyes isn’t quite as facile as Pro Tools at zooming in on waveforms.
Famous documentary filmmaker Ronn Pitts has died. A mentor to many, he also filmed many historic events in his town, Chicago. He is routinely credited with desegregating the Chicago filmmaking and news communities. Some remembrances from Kartemquin films are here:
3) A new camera from Arri aimed at documentary filmmakers
Arri have announced the Amira, a $25,000 (without a lens) on-the-shoulder camera with many similarities to the beloved Alexa. Shooting 2k images to various flavors of ProRes, it will capture up to 14 stops of dynamic range (with log C as well as rec 709 color spaces). The price may keep all but the best funded away… More here:
Or maybe more that National Geographic has gone “Reality TV” in their ethics. In any case, it seems that this Fox broadcast piece on the death of a US soldier was broadcast without much thought or sympathy:
I have nearly completed a rather large project – a DVD of Video Killed the Fairytale Star, Sunnyside Environmental School’s Spring musical. I ended up only editing one complete performance, rather than cutting together bits and pieces. But that is a two hour show, with a three camera shoot and 14 channels of field audio in Pro Tools. Plus, there is the Wednesday music-only performance (only one camera, and only 5 additional audio channels). And highlights from rehearsals…
The round trip between Avid Symphony (Media Composer is more or less identical, except Symphony has some added color correction functionality), and Avid Pro Tools continues to be a pain. Both programs say they can export and import AAF files that just link to existing media. But that mainly fails completely from Pro Tools field audio into Symphony. And now, going from picture lock in Symphony back to Pro Tools is also a pain.
I have tried variations on the “Send to Protools” workflow, and there are problems. Problems whether I tried to link to existing media or copy media. In particular, including video doesn’t work at all. Maybe I should have tried before I relinked to the original AMA media files – the lower-res proxy files, transcoded to Avid DNxHD 36, it might have gone more smoothly. And maybe Media Composer 7 will work better – AMA is supposed to be further buffed.
Fine. So I have to separately export a video file. And you can’t export a QT reference with AMA, so it takes a while (45 minutes per hour long segment). Then I’ll have to repeat that, using mastering quality, once the audio is mixed and laid back in. Sigh…
Then, the Send to Pro Tools was working, except for the two audio channels associated with my mixed down former multi-cam video (e.g., A1 and A2 don’t get transferred, although A3-A16 do just fine). WTF?
So, finally, I did an Export (to Pro Tools), consolidating the audio into OMF aiff-c files. The only problem now is that the markers didn’t make it to Pro Tools. So if I go back and – one by one – copy the markers onto one of the audio channels I am exporting, then they will come across. Maybe I should put markers on audio tracks only to begin with, or the timecode track? I tried just exporting the two channels, A1 and A2, that have the camera audio from two of the three cameras on the shoot. But for some reason, the Avid then does some really slow bit where it is “Copying media files…” that didn’t seem to happen when I exported all 16 audio tracks at once. I guess it is the AMA media again? Because I already had to import (instead of AMA linking) the field audio from Pro Tools to begin with – so only those A1 and A2 tracks are AMA linked.