You can pretty much do anywhing with AE and if you're out to just prove a point you can even edit a 20 minute feature film only with AE. The question at hand is...should you? Given that editing is quite fast paced and you want realtime performance to judge and change your edits to your taste and then change it all for your director, then change it one more time for the producer etc. will be quite hard with a RAM preview. And when we come to that point where we are talking about audio I'm curious if you would even reach the one minute mark until you switch to a NLE. But if you're really out to prove a point here, be sure to provide a link for some feedback
Can you drive a nail with a screwdriver? Can you chop down a tree with a pocket knife?
The average length of a shot in a film now days is about 3 seconds. 20 minutes / 3 seconds is 400. Now you've got 400 layers in you After Effects project. That, in and of it's self, is not such an impossible project, but say you decide that you want to change the timing of shot 200. This means you've got 200 layers to move. Also not an undoable task, but say that once you have moved changed the length of this shot you decide that the scene that uses shot 100 to shot 150 needs to be before the scene that starts at shot 72...
This is just the start of the polishing process of editing a film. I've been doing this for a very long time. More than 40 years. I have several personal friends that are feature film editors. I've taken classes and workshops from academy award winning editors, visual effects professionals, and I've worked on a bunch of features and produced hundreds of documentaries, commercials, made for TV movies and industrial films. I've never seen a project longer than about 30 seconds that wasn't re-cut at least 4 or 5 times. Polishing an edit in After Effects is a more difficult and tedious process than reorganizing the family photo album by just dumping 50 years worth of photographs on the floor and starting over.
Now let's consider sound. I recently cut an hour long made for TV movie called Stagecoach Santa. It is a fairly simple piece with 4 main characters and a small supporting cast. Each of the main characters had their own separate audio track. There were 3 audio tracks just for background sound presence. The music tracks consisted of 4 different tracks for each song and 4 more tracks for the transition from one piece of music to another. Then there were 2 tracks for sound effects and 2 more just to fix problems. Let's count them up. That's 19 audio tracks. Again, that's not impossible to handle in AE but adjusting and mixing would be extremely tedious and inefficient.
Now let's talk making picture, sound, and story judgements. On a hotrod system with a bunch of ram you can get maybe 30 seconds of preview at 50% on an HD project. That's usually more than sufficient time to judge whether or not a shot or a sequence is working, but it's a frustrating way to judge a 2 minute scene.
Ok, now you have the first cut of your film completed in AE. You render it out, take a look, let it rest in your mind for a couple of days, then you go back and look at it again and decide that you need to change some things in scene 20, 23, 56, 75, 84, and 105. There all just little changes.
You dive into your minimum 420 layer project and start making changes a layer, an in point, an out point, a keyframe at a time.
Compare this to a well organized workflow in an NLE where you can do rolling edits, grab entire sections of a project and just drag them around in the timeline while maintaining the integrity of all of your other sequences, and watch, in most cases, in real time without any wait for a preview, any part of anything you are working on. If I were to take a rough guess I'd say that a 20 minute project cut primarily in After Effects would take 4 months of work while the same project cut in any NLE would take 4 weeks.
Does that help. No matter what anyone says, telling a story with pictures is hard and tedious work. It always takes longer than you thought it would. A really fast news cutter is usually given an hour and a half to produce a minute thirty news story with straight cuts and 2 audio tracks. Spot news stories take less time, but special reports take 3 or 4 times that long to produce. Cutting a 20 minute film in AE is certainly possible. So is driving a nail with a screwdriver. So is cutting down a tree with a pocket knife. But I'd rather have a hammer. I'd rather have an axe. Heck, as long as I'm wishing for better tools, I'd rather have a chainsaw.
I would use AE to edit a 30 sec. spot. I wouldnt not use AE to edit a 30 minture film. Use the right tool for the job.
How long is my piece of string?
Thanks to all for providing some incredibly helpful and brilliant answers to my incomplete set of facts. Very illuminating.
The key point that clinches so much your analysis is that one cannot do "ripple" edits in an AE timeline, which means that any change in scene selection that deviates from the storyboard injects spiraling levels of complexity into what are routine workflows in an NLE. If AE only had a ripple edit capability on its timeline, that would greatly help in working on larger projects. I'd love to see such a capability in AE.
That said, my current timeline in AE is doing very well and providing some nice graphics change of paces and it is already 90 seconds long and I only have 21 layers so far. Compared to the boring universe of documentaries out there with virtually zero special effects, even having 18 layers gives me a better looking, sexier product.
Why would I attempt such a project?
If my 20 minute film consists of 11 sequences of 1.7 minutes each (a rounded figure) and each sequence incorporates just one different effect in AE (one instance of a composite, one instance of a cube with a video on each face rotating in 3d space, one of an anthropomorphic set of arms setting out ideas to be covered in the next portion of the film, one recurring transition of a wipe with a 3D cube orbiting to change into the next scene, etc.) I will have completed a very nice study which I can point to afterwards and say I am truly a well-rounded beginner, ready to learn more.
Because of this extremely nasty challenge of not even having a simple, rudimentary ripple edit function, I am forced to consider something like 3 or 4 segments of 6 or 7 minutes each and I am thinking of pulling the plug on any segment if it grows to more than 60 layers. It should be noted that, as a deep student of Japanese cinema (I have watched 15 different Yasujiro Ozu films twice and 18 different Kurosawa films three times for starters) my average shot lengths are much longer than 3 seconds. There is nothing wrong with longer shots if you know about composition, framing and storytelling. The Rope, by Hitchcock, was filmed in one continuous shot.
I'm much more of the classical school of film, from Kubrick and Bunuel and Ozu and Flaherty and Stewart McAllister. Just because MTV accelerated the editing process exponentially does not mean that I need to blindly follow the dogma that faster is always better.
As an exercise in becoming a well rounded beginner in After Effects with four months to devote full time to this project, it seems plausible that I can make a film with just four audio tracks in three seven minute segments consisting of no more than 60 layers each that is both beautiful and entertaining, assuming I have the shots and story to back it up, which I do.
Does that seem more feasible? Really a drag about no ripple edits, but at least I have the "slide" edits which are useful.
any further thoughts? thanks for weighing in so helpfully on such an incomplete set of facts. Matt Dubuque, 100 Trees
1 person found this helpful
My final thought. I love using PrP and AE together. For example, I will do a rough cut of a music video in PrP then using Dynamic Link I will send clips to AE for special effects. When complete I will export the AE comp as a Project Linked movie and replace it on the PrP timeline. I can then use Edit Original to reopen that AE comp any time I want. This is an example of a project I did using that technique. I think it employs the best tool for the job concept and I love the work flow.
Best viewed in You Tube at 1080P and Full Screen.
1 person found this helpful
Matt Dubuque said:
Compared to the boring universe of documentaries out there with virtually zero special effects, even having 18 layers gives me a better looking, sexier product.
I don't know what documentaries you've been watching but the last on that I worked on had sequences with 200 + AE layers showing the relationships between time and events. Most docs that I see are full of interesting well produced effects shots that help tell the story. Only the poorly conceived and badly produced are boring. I haven't seen a documentary without some kind of effects shot in more than 20 years.
The point I was making was that AE is a pocket knife, no it's a scalpel, used to sculpt your shots or sequences into something that best communicates a single plot point of your story. Combining those 'points' into a cohesive story requires a NLE for anything longer than a sequence. In point of fact most films I work on are longer than 30 minutes and most have some effects shots in at least half the scenes. I try and make most invisible to the viewer. I've lost the audience if they ever say "wow, wasn't that a great effects shot." Most of those 'sexy' shots and sequences were created in AE.
Simply having 18 layers doesn't make a better story, and if it's just eye candy then, and I do not mean to insult you, then the shot, the sequence, the story is not only boring, it's ineffective.
In the future you may be able to stay inside your NLE to build your multi layered shots. I would welcome more seamless integration between AE and PPro, but take it from someone that's cut about 300 films over 30 minutes long, the most efficient way to tell a story with picture and sound is to work on the sequences, cut them into scenes, then assemble the scenes into a movie. I can't remember a long form project that was a single sequence (timeline) in my NLE. Most have dozens, some hundreds of separate PPro or FCP sequences in the final master timeline.
1 person found this helpful
After Effects is simply not designed to do what an NLE like Premiere Pro can do. If you would like to submit a feature request asking for more advanced editing capabilities then by all means do so, but the most common workflow is to complete an edit in an NLE and then take it into After Effects to finesse shots and add any effects that your NLE is incapable of handling.
I'm also a huge fan of Kubrick, Bunuel, Hitchcock, Ozu and Kurasawa. And, I'm an even a bigger fan of other long take masters like Michael Cimino, Andrei Tarkovskiy, Andrei Konchalovskiy, Bela Tarr, Theo Angelopoulos, Sergei Bondarchuck and even contempary filmmakers like Ilya Krzhanovskiy and Terrence Malick. But, remember that all these directors are working in a style where the camera movement fits with the pacing, mood, blocking and editing of the story. And since they are all working in live action, everything is meticulously planned, directed and carefully placed so there's as little waste in the production and final edit of the film as possible.
Not only that, but most of these films, especially the older ones, were undoubtedly edited on flatbed manual film editing machines like Steenbecks or Moviolas. Now I've edited several very short films on flatbeds, and I love the tactile feel of it, but there's no real comparison to the speed of working in an NLE. Most movies shot on film these days that go back out to film are telecined, edited digitally and the final edit points are transferred back to the final film edit to create release prints for cinemas. Here's an example of a workflow chart I just pulled off of Google:
Now, unless your film is entirely animated and your shots are so short you'll have very little room to play with your edit after it's shot or scanned, I would highly recommend doing your final edit in an NLE. Then take it into After Effects and finesse your footage as much as possible. I imagine some of these old school directors might've enjoyed having the freedom of working in a digital NLE had they had the resources. However, it's important to take into account that the digital age has completely changed the way a lot of filmmakers are thinking about timing and editting their material. Who's to say that these filmmakers might not be making radically different work had they been working digitally. Although it can be argued that David Lynch's "Mulholland Drive" and "Inland Empire" are very similar films dealing with very similar themes, there's still a noticable difference between "Inland Empire" which was shot on a cheap video camera and "Mulholland Drive," which was shot on 35mm. I personally like both films, but there's a huge body of filmmakers who wholeheartedly disagree and think Lynch's work was much better when he shot film. Lynch himself loves the freedom of working digitally and has completely given up on film. It just depends on your aesthetic and what you do with the medium you're working with.
Having said that, working on long takes in After Effects is something I have quite a bit of experience with. During my first year at CalArts in the Experimental Animation Graduate program I made a film called The War Profiteers, which is an almost 7-minute-long digital 2D puppet film animated entirely in After Effects in one continuous camera movement almost like a haunted house amusement park ride. I made this film mainly because I wanted to explore what animation could offer in a single take setting. In hind site I might have had an easier time had I built the entire film in a 3D application like Maya, but I wanted to boost my After Effects skills and use the tools I felt most comfortable with. This film took about a year to complete from start to finish and by the end I was working in huge 3D comps with tons of textures and lights. Each individual character was made up of over a hundred layers and the final render of the film took over a week of invading several computer labs and sometimes rendering on 12 or more 6-core MacPro towers. I split up the render time, by dividing the timeline into chuncks and rendering out pieces of the film. However, during the first render the camera must've gotten accidentally moved and the pieces no longer lined up. I had to spend another week rendering out the entire film again. I'm definitely glad I made this film, but I don't think I would attempt something like this outside of a school or without a team and a studio. It was an extremely difficult and arduous process to get the entire thing done, and in the future I'll be wary of working with extremely long takes in After Effects.
1 person found this helpful
Not only that, but most of these films, especially the older ones, were undoubtedly edited on flatbed manual film editing machines like Steenbecks or Moviolas.
Until the mid 70's films were cut in 10 minute maximum reels on machines like this. Flat Beds were not in common use until the mid 70's. They were wonderful.
This is Dave Stone cutting sound against work print on a Moviola with 1sound head. I must have been lucky, mine had 2. Here's the interview.
Note the size of the reel. 35MM film runs at 98' per minute. 1000' feet is about 10 minutes. Check out the article and pay attention to asset management. It's still the same today. Manage assets, put the pieces together into the best sequence you can build, then start cutting the sequences together to tell the best story.
Thanks so much fellas.
Each and every one of these posts, without exception, is very beautiful and helpful for me. Each and every one of you has accomplished some truly remarkable things in your career and your posts reflect that undeniable fact, each expressed by you in your own very personal style. What a remarkable aggregation of professionals I have arrived within! Perhaps its something about After Effects that attracts the best and the brightest, given the complex tradeoffs each of us are compelled to make?
Because you all felt comfortable enough to share a bit about your work and your achievements, I would like to share a bit about myself.
While holding down a stressful full time job, I also studied law full time and passed the most difficult California Bar exam in history (with the highest failure rate) while my friends at Stanford Law School were dropping like flies from the brutal 3 day test from the fantastic complexity of the various fact patterns presented. They revamped the Bar Exam after that particular session because the failure rate was off the charts. And I passed that sucker with flying colors, leaving early on three occasions while the best and the brightest suffered. I knew what the hell I was doing and my results show I nailed the exam.
To accomplish such a thing while working an asinine 40 hour a week job, while everyone else at Stanford Law was a full time student, most of them with fat wallet, I had to exhibit meticulous hierarchical levels of organization while exhibiting an extreme ability to synthesize enormous data sets of completely disparate data at breathtaking speed.
Let me assure you this is a transferable skill, the ability to be impeccably organized on a most profound level. I am now mastering the ability to learn how to learn how to learn and I'm ready to go.
We all stipulate and agree that if I don't have a damn story and don't have a clue how to tell it, that this whole adventure is stupid.
As I said, I've got the shots (for example, for the overhwelming percentage of my shots, I have only shot footage the day after it rains because otherwise it isn't absolutely pristine and it looks frankly pathetic to me in an Ansel Adams kind of way) and I've got the story. You've either got the cards in your hand of poker, or you don't.
And the film school I devised for myself over the last 4 years has not been trivial. I have based it both on how I approached the study of law, as well as a careful study of the methodology used by James Longley, a young legend among his peers in documentary film who has only made three documentaries, but somehow managed to obtain an Oscar nomination for two of them, promptly after his graduation from film school. He was recently awarded a Macarthur Genius award for $500,000. I've been admitted to a private forum, based on my work, chock full of Oscar and Emmy award winners not open to the general public.
In short, I'm a bright guy, a quick learner and I'm truly fortunate to have it likely that people as supremely talented as yourselves may actually read this.
You have provided me so much in just this one thread, and I thank you sincerely for it.
I need to study all of your work.
I do have a question.
LASvideo, your video inspired me and I was wondering how it would look if rendered out so as to be displayed in blu-ray format on a 20 foot screen.
Do you feel your image quality from your present workflow in that context would be competitive with documentaries currently winning Oscars? My question is about the clarity and relative freedom from artifacts of your image in that display mode, compared to the median of the seven most popular documentaries of the last three years. Image quality on a medium sized screen.
Thanks again fellas,
I'm off for 36 hours of profound study.
Thanks Matt,. Im glad you enjoyed it. That video was a labor of love.And truth be told, it really was just a personal project for me to learn new chops and explore different styles.
With a good projector and Blu-ray player I think it would look fine. You were watching an H.264 file highly compressed for display on You Tube. The original master converted to Blu-ray whould have more detail and much less noise.
That's an awesome photo Rick. I wish I had a chance to work on one of those older Moviolas.
Thanks so much lasvideo, real nice job on that piece. I enjoy some of your other work as well.... cheers, matt