The Kanashimi post-production workflow is very much based on the traditional pre-digital film-centric way of doing things. For example, we shot with a separate digital sound recorder and used an old-fashioned slate and clapper board to sync the sound later. It was not even a digital slate with time code that goes "beep". It was just a regular old dry-erase slate with the black-and-white hinged sticks at the top. Stone-age!
But I like that. Simple == reliable. A non-digital slate does not need batteries or cables. It won't crash because of a bug in the firmware. Lining up the sticks and the "snap" in Adobe Premiere works exactly the same way it did on a Steenbeck or a Moviola (conceptually, at least). Also, it feels more like I'm making a film.
Other aspects of the work flow are film-like, too. For example, we've transcoded the good takes to Apple ProRes proxy, a full-resolution, low bit-rate format that is very resource friendly for a smoother, faster cutting experience. This is the equivalent of printing the takes you like with standard printing lights in a more-or-less disposable form for cutting, taping, ripping apart, marking up with grease pencils, and so on. For the grading and effects, the plan is to transcode just those parts we need to the ProRes 4444 format (which is much more disk-hungry, but is nearly lossless in quality). This is something like making fine-grained dupe negatives or interpositives for detailed opticals and color work. For heavy-duty stuff (and there will be some), we'll go for lossless PNG image sequences, which hold up through an indefinite number of generations. This is sort of like blowing up the negative to 70mm low-contrast fine-grain stock so the multipass optical effects don't degrade the quality as quickly.
In the olden days, the goal of all this film manipulation was to preserve the negative so that the highest-quality prints could be struck from it for distribution to theaters. In the digital age, we could theoretically work from a perfect copy of the "negative" throughout the whole process. Some do. But we just don't have the money for huge disk arrays and extremely beefy video cards. So we're borrowing the techniques of filmmakers past and using digital equivalents to keep our resources manageable for more modest equipment.
I suspect many readers of this blog who are also independent or hobbyist filmmakers will find themselves in the same boat, trying to decide whether the limited budget they have is better spent on food, locations, actors, and logistics, or on renting or purchasing a more robust editing system that allows them to work faster with fewer resource constraints. We've tried to find the "sweet spot", but - frankly - these are untested waters for us, er - me. As you can probably tell, despite my working in computer technology, my filmmaking chops (such as they are) were cut on more, shall we say, analog technology. More often video than film, too, for economic reasons. In fact, there are analog video equivalents for this workflow schema, too - the use of small-format dubs with the time code "burned in" for offline editing, for example, followed by the output of an EDL (Edit Decision List) for final assembly on a more expensive online system with the precious master tapes at stake (instead of the camera negative).
So I'm taking the somewhat unusual step of actually "finishing" the first four minutes of the film before moving on to the rest. Soup to nuts - fine cutting, effects, grading, and sound. Even the score if I can get David to start working on it. Why? So I can shake out the process and make improvements to streamline it before going all-in. It might even make a good promo piece for YouTube.
I'll let you know how it goes - what problems crop up, and whatever solutions we find. If you have suggestions or comments please leave them. I look forward to reading what you have to say.
Stay tuned!
Monday, March 25, 2013
Monday, March 18, 2013
Post-production begins in earnest
Many of the most respected names in the movie business believe that editing is where the film actually gets "made". Filming is just the prelude; FX, sound design, and music are the finishing touches. Editing is where you decide just what, exactly, the audience will be seeing and hearing (in terms of dialogue, anyway) from moment to moment. Which angle, which take, how long it will be on screen, whether you show the door opening or just have the actor begin in the room, whether the teardrop falls from her face or just slides down her cheek, whether he actually sweeps everything off the table in a fit of anger or just contemplates it in a silent rage, and so on.
A film is made of thousands of these little choices, each of which - taken on its own - is relatively insignificant, but which add up to the whole. It is the equivalent in creative writing of figuring out which words to use, which to leave out, how long to let sentences run, how thoroughly to describe a setting, etc. Each of this little decisions adds up to something that either succeeds or fails collectively. It's not that the writing, casting, shooting, set design, costumes, locations, and performances are unimportant. Goodness no! All of them (especially the writing!) are critical elements. But consider this:
Suppose you took the raw footage of any great film of any era, and went out of your way to pick the worst takes, the most mangled lines, and the technical mistakes, then used them to cut together your version of the film. Don't worry about pace - cut scenes off willy-nilly, or let them run as long you can. Don't think about the dramatic impact of each choice. Just slap the thing together and be done with it. How do you suppose it would come out? I have a feeling even Casablanca, Amadeus, or 2001: A Space Odyssey would turn out pretty poorly.
Great editing can make the most of a mediocre film - even "save" it in some cases. But poor editing can utterly destroy the best written, most compellingly acted and photographed effort. There are exceptions to every rule, of course. Consider Hitchcock's Rope, a play-like film paced entirely in 20-minute takes with no "editing" in the conventional sense.
Let's look at the typical post-production work flow as I learned it (which was through self-study and actually doing it) to illustrate the how's and why's of many of these choices. Some readers (especially younger ones, or those schooled primarily in video) may take issue with them, and not without reason. The world of film production is changing quickly, and even in traditional "old school" film production (from which I draw these post-production steps) the lines are often less distinct than I present them here. With that said, a film in post-production generally progresses like this:
ROUGH CUT: the rough cut is put together with very little awareness of pacing. The idea is not to create a "complete" film but to see if what you have is working dramatically. It emphasizes the performances, not the montage. The goal is to see where things are weak or strong "out of the box"; where the acting and dialogue can carry things without much help, and where they need the assistance of other departments (additional photography, music, special effects, tricky editing) to shore them up. Most rough cuts are unwatchable to the average film-goer. But they are valuable in evaluating your material and making crucial decisions about finishing.
I did not let anyone but the cast and crew watch the rough cut of Kanashimi, because it's frankly pretty crummy, in places. The parts that really work are the ones where the acting and writing are strong enough to "punch through". That's the point of a rough cut, though: to be able to evaluate the material without the distraction of artsy editing or glitzy effects to distract you from the writing and acting. You do pick your best takes, and you do try to put together the best moments from them. But - by design - they get no other help. They stand or fall on their own merits.
WORK PRINT: this is the part where you spend the most time, and this is the phase Kanashimi is entering. The term comes from the days when most films were actually made on film and temporary "work" prints were made from the camera negative with little or no color-correction so that the editor can "play with" the film with impunity. He or she can cut, re-cut, cut again, add stock sound effects and temp music; screen it, evaluate it, and keep playing with it generally until it starts to look like something. An advanced version of the work print is what most people think of as a "rough cut", but that's more a marketing thing to drum-up word of mouth, than a formal definition. Work prints are often shown as "rough cuts" in major cities at surprise screenings both to get feedback from audiences and to get word-of-mouth going for the film. An actual rough-cut (as I define it here) would NEVER be shown like this, unless the studio marketing department is suicidal.
FX-laden films (science fiction, fantasy, and - increasingly - every other genre) will sometimes have low-resolution, quickly-rendered "placeholder" shots, or animatics, or just shots of the storyboards cut in to inform viewers of what is supposed to go where, since these elements are usually just beginning serious work.
FINE CUT: (as opposed to "final" cut). This is really a very worked-over workprint with most or all of the completed effects shots, opticals (dissolves and such, which are mostly done digitally, now), and sound effects put in. But NOT the music, although there is probably a temp track! By this point, color correction has been done (though usually not final grading). The film is very close to what general audiences will actually see at this point. This is where you really let people (the stakeholders, formal test audiences) look it over. Adjustments are typically small from here on out. A lot of effort has been put into polishing and bringing out the little moments that make the film work. Or fail to, as the case may be.
FINAL CUT: when the fine cut is "locked", and no more changes are anticipated, the "final cut" has arrived. The composer and sound designer get hold of it to compose the score and perform the sound mix with all the effects, and any re-recorded dialogue. The reason these things happen so late is that the sound and music must be in sync with the picture, and cutting scenes longer or shorter will require re-working them. You don't want to pay a 100-piece orchestra to record a new version of the background score unless you absolutely HAVE to!
ANSWER PRINT: again, a term from the days when films were exclusively shot and edited on film. It used to be the actual camera negative ("original negative", or "o-neg") would get cut together (or "conformed") to match the final cut, then it would be printed by the lab, along with the completed sound track, for screening by the makers to ensure that everything is fine and dandy. If so, then the process of making prints for theatrical distribution would happen next. If not, then more adjustments would be made (usually to the sound mix or the color-correction) and a new answer print would be struck, until everyone was happy.
Digital technology has has made these boundaries quite slippery. It's now common to never even touch the camera negative (if the film was even shot on, well, film), even to make a work print. It's transferred directly to hard disks for FX work, color grading, editing, and everything else. A printing negative is made from the digital sources, or (more and more, these days) copies are made on digital media for electronic projection. Or even just beamed to the theaters over encrypted digital satellite links.
Nonetheless, each of these stages have some equivalency in the new world of digital film production, and so I shall use them as a reference point in Kanmashimi's post-production.
A film is made of thousands of these little choices, each of which - taken on its own - is relatively insignificant, but which add up to the whole. It is the equivalent in creative writing of figuring out which words to use, which to leave out, how long to let sentences run, how thoroughly to describe a setting, etc. Each of this little decisions adds up to something that either succeeds or fails collectively. It's not that the writing, casting, shooting, set design, costumes, locations, and performances are unimportant. Goodness no! All of them (especially the writing!) are critical elements. But consider this:
Suppose you took the raw footage of any great film of any era, and went out of your way to pick the worst takes, the most mangled lines, and the technical mistakes, then used them to cut together your version of the film. Don't worry about pace - cut scenes off willy-nilly, or let them run as long you can. Don't think about the dramatic impact of each choice. Just slap the thing together and be done with it. How do you suppose it would come out? I have a feeling even Casablanca, Amadeus, or 2001: A Space Odyssey would turn out pretty poorly.
Great editing can make the most of a mediocre film - even "save" it in some cases. But poor editing can utterly destroy the best written, most compellingly acted and photographed effort. There are exceptions to every rule, of course. Consider Hitchcock's Rope, a play-like film paced entirely in 20-minute takes with no "editing" in the conventional sense.
Let's look at the typical post-production work flow as I learned it (which was through self-study and actually doing it) to illustrate the how's and why's of many of these choices. Some readers (especially younger ones, or those schooled primarily in video) may take issue with them, and not without reason. The world of film production is changing quickly, and even in traditional "old school" film production (from which I draw these post-production steps) the lines are often less distinct than I present them here. With that said, a film in post-production generally progresses like this:
ROUGH CUT: the rough cut is put together with very little awareness of pacing. The idea is not to create a "complete" film but to see if what you have is working dramatically. It emphasizes the performances, not the montage. The goal is to see where things are weak or strong "out of the box"; where the acting and dialogue can carry things without much help, and where they need the assistance of other departments (additional photography, music, special effects, tricky editing) to shore them up. Most rough cuts are unwatchable to the average film-goer. But they are valuable in evaluating your material and making crucial decisions about finishing.
I did not let anyone but the cast and crew watch the rough cut of Kanashimi, because it's frankly pretty crummy, in places. The parts that really work are the ones where the acting and writing are strong enough to "punch through". That's the point of a rough cut, though: to be able to evaluate the material without the distraction of artsy editing or glitzy effects to distract you from the writing and acting. You do pick your best takes, and you do try to put together the best moments from them. But - by design - they get no other help. They stand or fall on their own merits.
WORK PRINT: this is the part where you spend the most time, and this is the phase Kanashimi is entering. The term comes from the days when most films were actually made on film and temporary "work" prints were made from the camera negative with little or no color-correction so that the editor can "play with" the film with impunity. He or she can cut, re-cut, cut again, add stock sound effects and temp music; screen it, evaluate it, and keep playing with it generally until it starts to look like something. An advanced version of the work print is what most people think of as a "rough cut", but that's more a marketing thing to drum-up word of mouth, than a formal definition. Work prints are often shown as "rough cuts" in major cities at surprise screenings both to get feedback from audiences and to get word-of-mouth going for the film. An actual rough-cut (as I define it here) would NEVER be shown like this, unless the studio marketing department is suicidal.
FX-laden films (science fiction, fantasy, and - increasingly - every other genre) will sometimes have low-resolution, quickly-rendered "placeholder" shots, or animatics, or just shots of the storyboards cut in to inform viewers of what is supposed to go where, since these elements are usually just beginning serious work.
FINE CUT: (as opposed to "final" cut). This is really a very worked-over workprint with most or all of the completed effects shots, opticals (dissolves and such, which are mostly done digitally, now), and sound effects put in. But NOT the music, although there is probably a temp track! By this point, color correction has been done (though usually not final grading). The film is very close to what general audiences will actually see at this point. This is where you really let people (the stakeholders, formal test audiences) look it over. Adjustments are typically small from here on out. A lot of effort has been put into polishing and bringing out the little moments that make the film work. Or fail to, as the case may be.
FINAL CUT: when the fine cut is "locked", and no more changes are anticipated, the "final cut" has arrived. The composer and sound designer get hold of it to compose the score and perform the sound mix with all the effects, and any re-recorded dialogue. The reason these things happen so late is that the sound and music must be in sync with the picture, and cutting scenes longer or shorter will require re-working them. You don't want to pay a 100-piece orchestra to record a new version of the background score unless you absolutely HAVE to!
ANSWER PRINT: again, a term from the days when films were exclusively shot and edited on film. It used to be the actual camera negative ("original negative", or "o-neg") would get cut together (or "conformed") to match the final cut, then it would be printed by the lab, along with the completed sound track, for screening by the makers to ensure that everything is fine and dandy. If so, then the process of making prints for theatrical distribution would happen next. If not, then more adjustments would be made (usually to the sound mix or the color-correction) and a new answer print would be struck, until everyone was happy.
Digital technology has has made these boundaries quite slippery. It's now common to never even touch the camera negative (if the film was even shot on, well, film), even to make a work print. It's transferred directly to hard disks for FX work, color grading, editing, and everything else. A printing negative is made from the digital sources, or (more and more, these days) copies are made on digital media for electronic projection. Or even just beamed to the theaters over encrypted digital satellite links.
Nonetheless, each of these stages have some equivalency in the new world of digital film production, and so I shall use them as a reference point in Kanmashimi's post-production.
Tuesday, March 12, 2013
Principal photography (finally!) wraps
Ah, what naivete we have going into these things! For reasons that seemed (and were) quite practical at the time, we'd scheduled Principal Photography over two weeks in January. For a 30-minute short film, this is not an unreasonable schedule. It is, in fact, pretty generous compared to your average hour-long TV drama.
But everything takes longer than you think. I've been a professional software developer for about 14 years, and the same is true in that profession. As alluded to in previous posts, even the most seasoned software developers and filmmakers sometimes have a hard time knowing how long something will take to do, given a minimum standard of quality to meet, plus all the thousands of variables over which you have little or no actual control.
So we had to shoot a couple of weekends worth of pickups, inserts, and what would, on a bigger production, be considered "second unit" photography. We have no second unit, but idea is the same: anything that the actors are not in, which is used to, say, establish the location or provide connecting material is usually done separately - at the same time, by another crew - the "second unit" - or (as in our case) by the "main unit" after the pressure of having actors on the set waiting to work has been relieved. In many cases, it was just me and Erik Mayne (http://www.mediafxvideo.com) driving around town, shooting things that looked interesting and which might fit in dramatically. Some of it was carefully planned for use in the film; most was just stuff that looked like it could be useful someplace. I encouraged Erik to just grab stuff that he thought was cool looking. A couple of those shots he grabbed are DEFINITELY going in. They are beautiful.
We also had to mix in work with the principal actors to get some missing material. Anyone watching the rough cut from a few weeks ago, who had not also read the script, would have been very confused because some key moments were still missing. I'm happy to report that we have them now, and they look fantastic. Supporting actor Darryl Small played a key part behind the camera, as an honorary crew member, because he happened to live close to one of the locations and could get us a car we needed to make the shots work.
Special thanks to Holly Rose, whose outstanding reading (of a shortened version) of my fake Japanese legend (posted in this very blog a few weeks back) turned out to be the key to making the revamped opening of the film work. Lisa Sherman's (http://www.mediafxvideo.com) cinematography is at its finest in this scene - a simple tracking shot of a mother reading her son a story. The lighting and composition are spot-on.
Leading man Thurman Kellogg also stepped up to the plate to deliver the goods for some reshoots against green screen of perhaps the most emotionally draining scene in the film for everyone. The reshoots were not because of anything he or the crew did wrong - it was all me, baby. I just didn't get everything I needed the first time. He and Erik graciously committed their time and talent to make it work. They went 100%, full throttle. I cannot WAIT to get it cut together with the background plates keyed in.
Making a film - any kind of film - is very hard work. Next time you see a movie, even a bad one, think about the fact that a lot of blood, sweat, and tears went into making it. Hats off to anyone who can see the process all the way through.
And a standing ovation for those who manage the trick of making a wonderful film while they're at it. It's too early to say if Kanashimi will be as good as we all hope, but if it is not, it won't be for lack of effort or dedication!
But everything takes longer than you think. I've been a professional software developer for about 14 years, and the same is true in that profession. As alluded to in previous posts, even the most seasoned software developers and filmmakers sometimes have a hard time knowing how long something will take to do, given a minimum standard of quality to meet, plus all the thousands of variables over which you have little or no actual control.
So we had to shoot a couple of weekends worth of pickups, inserts, and what would, on a bigger production, be considered "second unit" photography. We have no second unit, but idea is the same: anything that the actors are not in, which is used to, say, establish the location or provide connecting material is usually done separately - at the same time, by another crew - the "second unit" - or (as in our case) by the "main unit" after the pressure of having actors on the set waiting to work has been relieved. In many cases, it was just me and Erik Mayne (http://www.mediafxvideo.com) driving around town, shooting things that looked interesting and which might fit in dramatically. Some of it was carefully planned for use in the film; most was just stuff that looked like it could be useful someplace. I encouraged Erik to just grab stuff that he thought was cool looking. A couple of those shots he grabbed are DEFINITELY going in. They are beautiful.
We also had to mix in work with the principal actors to get some missing material. Anyone watching the rough cut from a few weeks ago, who had not also read the script, would have been very confused because some key moments were still missing. I'm happy to report that we have them now, and they look fantastic. Supporting actor Darryl Small played a key part behind the camera, as an honorary crew member, because he happened to live close to one of the locations and could get us a car we needed to make the shots work.
Special thanks to Holly Rose, whose outstanding reading (of a shortened version) of my fake Japanese legend (posted in this very blog a few weeks back) turned out to be the key to making the revamped opening of the film work. Lisa Sherman's (http://www.mediafxvideo.com) cinematography is at its finest in this scene - a simple tracking shot of a mother reading her son a story. The lighting and composition are spot-on.
Leading man Thurman Kellogg also stepped up to the plate to deliver the goods for some reshoots against green screen of perhaps the most emotionally draining scene in the film for everyone. The reshoots were not because of anything he or the crew did wrong - it was all me, baby. I just didn't get everything I needed the first time. He and Erik graciously committed their time and talent to make it work. They went 100%, full throttle. I cannot WAIT to get it cut together with the background plates keyed in.
Making a film - any kind of film - is very hard work. Next time you see a movie, even a bad one, think about the fact that a lot of blood, sweat, and tears went into making it. Hats off to anyone who can see the process all the way through.
And a standing ovation for those who manage the trick of making a wonderful film while they're at it. It's too early to say if Kanashimi will be as good as we all hope, but if it is not, it won't be for lack of effort or dedication!
Subscribe to:
Posts (Atom)