"What took so long?" I get that question a lot. Short answer: the film was more complicated than anyone expected, and people were doing it on their free time, so it took the time it took. And while that answer is true, it’s just the “elevator pitch” version of the story. I’m not here to write scuttlebutt or point fingers. But I thought it was worth sharing my observations on both how the show got bogged down and how it got to the finish line so that maybe other fan filmmakers could learn from it. * * * * * PART 1: Getting From There to Here My understanding is that when the “The Tressaurian Intersection” was first assembled there were obvious defects that the key stakeholders all knew would have to be addressed. Part of this was production realities (e.g. due to time pressures on set the coverage of some scenes wasn’t that good) and some scenes didn’t really work when cut together. This pointed to the need to do some pickup shooting. Somewhere in there the decision was made not to sit on the whole show until all these issues could be addressed, but to start releasing the show as individual acts, like episodes. Now, post-production on an all-volunteer shoot is prone to slippage. People can’t treat it like a full time job, so, despite the best intentions, stuff happens to cause delays. As such, each subsequent segment took longer to complete than was anticipated. Also, there were strong personalities involved, and individuals who didn’t always agree on what was the best solution to the challenges the show presented. This is part and parcel of passionate people in creative pursuits. By the time Act 3 was in postproduction this added up to substantial delays. The fact that Act 3 contained a briefing room scene which just didn’t work to anyone’s liking (despite pickup shots) meant that it got edited and re-edited and re-re-edited ad nauseam, until an acceptable compromise could be reached and released. At that point, several things contributed to the show grinding to a virtual halt: Robbing from Peter to pay Paul Because the show was edited as a series of acts rather than as a whole, when changes were made to make each individual act work, said changes had a domino effect into subsequent acts. Act 4 Didn’t Work I mean that. I know several of the key players felt they had an acceptable cut in late 2006, but when I reviewed it I knew that wasn’t so. It might’ve been perfectly passable as an average fanfilm ending but it failed to live up to the promise of the rest of the show. Part of this was the fault of the shifting act breaks messing up the arc of the act, but another was simply that the cut wasn’t very good, and I think on some gut level everyone knew it even as their brains said “good enough”. Fatigue & Real Life It had been a long hard slog to get to Act 4, and when there wasn’t unanimous love for it and no obvious solution to its problems, parties were worn out and had other real-word concerns to tend to. It’s difficult to maintain momentum on a project of this scale, and it’s perhaps inevitable that it ran out of steam when the pieces didn’t fall together. The FX Mess Act 4 required a lot of visual effects—arguably more than any other act—many of which would be amongst the most complex in the show, and many of which had not been designed or had to be redesigned because the Act 3 ending changed the nature of the shots needed for Act 4 over what had been scripted. All this added up to a project which effectively ground to a complete halt. Even when it got moving again, it was slow going. All of the above had to be dealt with, and some of it promised not to be easy. Which brings the story to Act 4 and the long voyage to bring it home. Next up: Paul Picks Peter's Pocket
That's pure Dedication, Maurice! Thanks for sharing! Very excited about the next parts. So (not to be taken offensively, please!) YOU were the person who was responsible for the delay, because you saw the cut not being good enough for being released. Right? Greetings Per
PART 2: Paul Picks Peter’s Pocket There are some longstanding misconceptions about the production of “The Tressaurian Intersection” that are worth dismissing once and for all. The show was not shot and released one act at a time. Principle photography was done in one two-week period in July 2004, followed by a one-day location shoot for the Teaser’s brief planet sequence, and some pickup shoots. While the show was completed and released one act at a time there DID exist a complete rough cut before any of the segments were made public. In fact, it was this rough cut which convinced the key stakeholders that they hadn’t gotten everything they wanted and thus additional shots were needed and some scenes should be changed. This is no failing on anyone’s part. It’s nearly impossible to know if everything is going to work until you assemble it, so pickup shoots (meaning you “pick up” shots you didn’t get before) are par for the course. The various pickups included Harris floating in weirdspace (unused), additional close-ups of a few characters, revising and reshooting the Act 3 briefing room dialogs, staging a new stunt for the Act 4 lizard escape (unused), and adding material to the climactic action in the transporter room. However, this all lead to the fateful decision to not delay the entire show and instead complete it sequentially, act by act, like a series of mini episodes. When you’re cutting a complete film you periodically look at it as a whole to make sure it all holds together and that adjustments made early in the cut don’t have unintended consequences later. This is much harder to do when each of five acts is fine-tuned, polished and released individually. It makes it nearly impossible to look at the whole film as a piece, so you end up having to make adjustments, segment by segment, to deal with unintended consequences one act might create for subsequent ones. For example, Act 3 was scripted to end with Garrovick calling Harris from the Briefing Room and telling her, "I’ve got it — we’re going to destroy that thing out there," but as released it ended much later, after the shuttle launch and the Exeter starting her run towards the BFD (Big F***ing Device). (This change may have been motivated in-part by the fact that the problematic Briefing Room scene had been pruned back from its original form and the act end just seemed weak given the compromised scene.) Not such a big change, right? Here's what that decision wrought : Act 4 was shortened by 2.5 pages, (about two minutes of screen time) or about 20% of its scripted length. The bulk of the dialog for Act 4 was in those pages, which resulted in an act that felt unbalanced because there was no build into the action. The Act would have to start mid-action. The shuttle distraction sequence was scripted to take place prior to the ship beginning its run, but Act 3 ended with the ship starting the run before the "distraction" to lure the Tholians aside, so the order of events wasn’t logical. The shuttle launch in Act 3 was tweaked to create a bit of misdirection regarding who was aboard it, and imply Cutty's the pilot by not showing him after it launches. But since the episode wasn't shot with this in mind, there was the problem of how to get him back on the bridge. Act 4's whole setting was changed by the last shot in Act 3, which threw the Exeter into “weirdspace”. As originally scripted, the ship stayed in normal space and flew into a “CENTRAL POWER TUNNEL” of the device, but that was out the window. So, instead of flying through normal space as planned, all the action would have to happen within weirdspace (or we’d have to jolt the ship back into normal space, which seemed like a cheat). This meant: —No normal star fields as backgrounds during the run —A lot of weirdspace shots would need to be designed and executed, requiring... —Designing a "dimensional"/fly-through-able weirdspace (the glimpses we’d gotten of it previously were relatively stationary, not racing through it). What would that even look like? This was a terrific problem. And—as Act 2 established that the Exeter crew couldn't detect the Kongo when it had disappeared into the “weirdspace” dimension—shifting the Exeter into weirdspace before the distraction created the problem of how our heroes inside weirdspace could trigger the distraction back out in normal space, let alone even know about what was happening out there. And that's just what that change to Act 3's ending did. So, to take from Act 4 to give to Act 3 solved one problem by incurring others. It was robbing Peter to pay Paul. Next up: The Act That Wasn’t
You really know, how to explain things and tease: "The Act That Wasn't". It's a wonderful read and it makes one wonder, how anybody wouldn't give up over the years. It reads like "unsalvagable" - but it has been done nontheless. And it doesn't show any inconsistencies. Congrats once again.
Thanks. It helps that once upon a time I wrote a beginners column for a computer magazine and also wrote manuals for software, so I have a feel for how to explain and illustrate things. As I said in the intro, I think the process here may be useful to others making such films, if only because they'll be aware of the kinds of pitfalls that can trap the unwary.
Don't start the next segment without me! I just need to go pop the popcorn. Everyone likes butter, right? (Thanks for providing these insights, Maurice. Fascinating.)
Awesome and informative so far. Thank you for doing this. The biggest surprise for me so far is: Wow! I remember in one of the trailers that the effects evidently of the climax were substantially different, like flying along an orange grid or something, IIRC, but after Act 3 came out I'd always assumed that those were just placeholder effects for the ship flying through weirdspace.
The Power Tunnel was in the original Next Episode trailer. It also features a work-in-progress sky replacement for Corinth IV and a Tressaurian attack cruiser which is upside-down from its final orientation and which fires phasers from a different spot than in the finished episode.
^ Cool. And now it is evident that the Next Episode trailer contained shipboard footage from the final act.
The Act 4 trailer also contains some placeholder VFX (including an unused "BFD" design) and a version of the flying redshirt stunt before the wall was stabilized by NEO f/x. Oh, and Scott Cummins doing a misdirection loop of Garrovick.
PART 3: The Act That Wasn’t Sorry for the delay here. I had two out-of-town trips which delayed me on the follow-up. Shortly after the release of Act Three of "The Tressaurian Intersection" there was a concerted push to try to slam into the 4th act. By late March 2008 there was a rough cut of the act that some key stakeholders thought was good enough to proceed with, but which didn't receive a unanimous thumbs up. Furthermore, a proposed deadline to get the film done in time for an invitation to screen it was rejected, which led to some, shall we say, heated disagreement, and what with real life intruding, it was here that the project ran out of antimatter and went adrift. It was somewhere after this stall that I finally followed up with Jimm Johnson on his interest in my helping edit the show (I had not previously because I was not editing in Final Cut before 2008). I sent him a hard drive and he sent back all the drive images for the show. The first thing I did was re-review the last Act Four cut. Now, I'm going to be blunt here. The edit in question did not work. It just didn't. Oh, it could have been completed and most fans of the show and fanfilm watchers would have thought it was just fine, but it certainly was not exploiting the full potential of the material. I identified several key concerns, in the following order of priority: One-note pacing As related previously, by shifting around the sequence of events in Act Three the result was that Act Four started in the middle of action and just barreled through towards the ending without a break. There was no buildup. No arc. Lack of human moments Typified by the moment when B'fuselek screams and falls down, possibly dead, and no one reacts to his plight. Poor action sequences e.g. the lizard escape and transporter fight Drawn out denouement From the moment the action ends to the fadeout there was just too much talking and too much explaining going on. Too much V.O. (voiceover) There was a lengthy, on-the-nose Captain's log. And, after the waaaaaay too long log in Act Three, this was just one voiceover too far. Too many O.C. (off camera) lines In an attempt to "fix" problems, a lot of "wild" lines had been added. So many, in fact, that it drew attention to itself because you were hearing all these lines from people you rarely saw from the front. Lack of cause and effect in action Yes, weirdspace is dangerous and bumpy, but the ship would lurch violently all the time and systems would fail just because. I think on some gut level everyone knew the act had problems, even those who felt this edit was workable. But I suspect that after putting so much effort into trying to make it work people had lost perspective. They'd become so focused on patching the tire that they forgot to make sure the wheel was spinning on the axle. Oh yeah, and then there was the VFX mess... Next up: Pressing the Reset Button
PART 4: The Disorganization Man Given that one of the issues I’d identified with the 2008 cut of Act 4 was that there had been so many fixes attempted that the thing had become an unfocused mess, it seemed the best thing to do was to not start by trying to fix the act, but just to start over. So, once I got the hard drive from Jimm I was ready to dive in and see what I could do with Act 4. The first things to do were: Review the last cut of Act 4 Re-read the script to see what the original intent had been Review all the Act 4 live action footage and assess what was useful and what was missing Review the completed visual effects and all the effect elements/components But this assessment was hampered by this one teeny organizational issue. And by teeny I mean huge. The data for the show came on five logical drive partitions. Over the years, as the show was being worked on, each time the project filled up a drive, another drive was added. Since the show was being edited in discrete chunks instead of as one big project, files ended up all over the place, in folders with names like "Project Overflow" or “Misc”. Some footage was on a drive titled Mac HD, another on Drive001, pickup shots in other locations. Bits of ADR (dubbed dialog) were in all sorts of places. Finding anything was a challenge. Frequently I was forced to open project files for the other Acts and use “Reveal in Finder” to figure out where a clip had come from. Heck, I was stumbling upon bits I’d missed previously right up to the last month before release. And then there were the files themselves... Beheaded. In importing the footage, someone chopped off all the heads and tails of the takes, which included the slates (clapper boards) and anything that happened after “cut” before the camera was shut off. Cutting off the heads was a big no-no! Invariably some files were misnamed and without the slates it was needlessly harder than it should have been to figure out what such files were. My guess is that during transcoding someone decided to cut the heads and tails off the takes to save drive space. Ouch. What's in a Name? A lot,it turns out. There was no universally applied file-naming convention. Most video clip files might have names like Harris Wide_41_3 but some others would be Sc. 8 Harris 001. A few didn’t even feature scene numbers or would abbreviate the character names. Everything should have been named in a consistent fashion—for instance—leading with the scene and take number, like 58C-T1_Cutty_CU.mov. Having a convention like that makes it easy to locate elements, as you can sort directories by name to see all the pieces of a given scene clustered together, or search by character name to find all the shots of them. Likewise, the visual effects elements should have had version numbers in their filenames, but didn't, which made it difficult to figure out what was the final version of what. Because of this haphazard situation, I literally had to comb through thousands of files to find every little potentially useable piece. This was an incredible time chewer. I was discovering shots and elements right up to the week we finished. In addition to the aforementioned concerns, there are a few other very important reasons not to chop off the heads and tails of takes: Technical Without the slate and its clap the most base-level backup system for syncing sound and picture is missing. Yes, most modern editing software does this pretty well and automatically, but there's no reason to throw the option out the window in the event all else fails. Storytelling/Editorial You never never never throw away ANYTHING. In a Battlestar Galactica commentary Ronald D. Moore once related how he felt the edit of a particular episode required a moment of the President with an expression they’d never shot, and with no other practical alternatives he scrubbed through all the “off-slate” material—which means the stuff before “Action” and after “Cut”—and in that footage he found a moment where actor Mary McDonnell was preparing herself for the take and there was the expression he needed. Likewise, you sometimes find extra bits of useable business in off-slate moments or in bloopers, such as an actor entering or exiting a shot, laughing or showing anger or some other expression or doing some useful or interesting bit of business that you can use. VFX Finally, I’ve found clean plates helpful to the VFX guys in off-slate moments before an actor steps in, etc. Ideally I should have just started over, reviewed everything, renamed like mad, and organized the project logically, but that would’ve created other headaches, including backwards compatibility concerns with the edits of the previous acts. In the end I left the existing files where they were and created a single root level folder into which I put everything new I created, and everything new would have standard naming systems. (I learned a valuable lesson from this, though. For Polaris I created a spreadsheet for tracking all the audio and video data, and that allowed me to rename files during transcoding but still know what the source files were in case anyone wanted to go back to the original raw data.) There was a silver lining, however. Despite all these issues, I wasn’t forced to start completely from scratch. The 2008 edit of Act 4 wasn’t up to snuff, but it contained the takes that the director and producers liked best. This meant that I wasn’t forced to undertake a shot log review and selection process to select takes. That big job was largely done. So, I had the pieces. Now I had to put it together. Next: Every Film is Written Three Times EDITED APRIL 17, 2016 to include what had previously been in an addendum reply (since deleted).
FASCINATING. Thank you so much for taking the time to do this, Maurice... I'm sure a lot of budding fan-film producers, directors and editors are getting a lot out of this.
Thanks. I was beginning to wonder if anyone was reading this or cared. I'm writing this to be helpful to other fan filmmakers, not to hear myself type.
I haven't commented on these but allow me to join Professor Moriarty in praising these updates/expounding on the production and post-production of "Tressaurian." These are great histories and I am quite enjoying reading through them Maurice, as I'm sure many others are as well.
I've been finding it fascinating. Even though I don't do fan films, it's a great example of taking over any stalled project.
I think it's a very valuable thing you're doing Maurice. The part about trying to sort stuff out when there was no convention for labeling material especially resonated with me. After years of doing audio projects and trying, mostly without success, to get people to use a standard labeling system for my audio work I can vouch for how time-consuming it can be when you have to come in behind someone else's work. I've personally settled on a system where each act is given it's own folder where the lines for that scene go. Then I create a sub-folder for draughts of each mix. If re-cuts are required they can be put into their own sub-folder to keep them from being mistaken for the originals are getting lost in the main folder. The more work you do on the front end to come up with a sensible, logical labeling system, the less time you will waste looking for bits and pieces that may seem unimportant now, but can be vitally important later. It doesn't have to be industry standard (but if you want to do professional work that's what you should strive for) just something that works for you. It's the same whether you're working on audio or video. The difference is just a matter of degree.
As others have said, this is very useful. I've been guilty in the past of haphazard file naming and storage, and while I've been striving not to do that in recent times, it's always helpful to hear about others' experiences. And the 'warts and all' story is great reading.