Sevastopol rough cut

video: Tae Guk Gi
music: Heaven Shall Burn - "Sevastopol"
link: mediafire [4.4 MB]

This is a rough cut of the first 21 seconds of a video I've been not working on for the past three months, since it was too hot to do anything. I only had the first vob cut -- there are like 6 more left, including nearly all of the content that will go in the bulk of the video -- but after such a long layoff I had to do something to assure myself that I even knew how to do this shit. Hence, this cut.

The final video is probably not going to look a lot like this, composition-wise, but the trigger behavior is probably pretty close. I may play around with the color, but in the final analysis I don't know what to do with blastbeats, especially on such a violent source and directly aggressive music, except trigger them, and it mostly works. If this was going into an actual cut, I'd've fixed some points up, but even as it is, it works ok, which is what I was trying to convince myself of.

Note: there are 8 triggered sections in these 20 seconds; compare to 28 triggers in 150 or so seconds on SH112. Granted, some of those triggered sections run on for five and ten seconds at a time, but the point remains.

Fucking Ideas, How Do They Work?

As part of the continuing delusion that I make AMVs any more, I've completed the first shakeout of my to-do list in almost three years, with some probably uninteresting results.

If everything on the current list gets made and nothing that isn't, I'll finish up with SH130. This is probably going to be about accurate, because though there are four ideas on the list that I have no real idea of the viability of, there are a lot of long cuts on the list that difficulty and plain old inertia will push the finishing of all those ideas out for at least five years in the most likely case, and new ideas are probably going to come in inside that span.

Slightly more interesting is the vintage of the ideas concerned; this breaks out as follows:
2000 - 1
2001 -0
2002 - 1
2003 - 0
2004 - 2
2005 - 1
2006 - 1
2007 - 0
2008 - 1
2009 - 1
2010 - 7

I've been keeping a to-do list since 2000, before I even started making these things, and while most of the current ideas are actually current, there are a few that have been carried along from the very start. This has major implications for the anticipated delay of finishing up: ten of the fifteen ideas on the drawing board, and six of the eight that have been around for more than a year, involve cutting at least 20 episodes or episode-equivalents. Several are a lot longer than that.












This is a graph of current ideas of year X vintage versus videos produced that year. Most of what I've done is based on relatively new ideas; SH110 in 2008 was a 2007 idea, SH111 was a 2008 in 2009, SH112 and 113 were 2009 ideas in '09, SH114 had been around as an idea since '06, and SH115 was a '10 idea and production. If a video idea is doable quickly, it tends to get done quickly; the hard ideas stick around as newer projects get promoted past them, and the crap ideas, like those four that I haven't actually done any checking on may turn out to be, die off and fall out of the list unproduced. This is the hard part of crushing out the ideas that are left and winding up the studio: by taking the path of least resistance, I end up saving the hardest part for last.

The other way of looking at that, though, is that these ideas are, like SH070, not coming into being before their time: that now, at the end of the road, I either have the skills to pull them off correctly or the composure to toss them out as unworkable. At the very least, the future is looking bright musically: those fifteen ideas break out into ten metal, two hardcore, one punk, and only two songs outside that bailiwick. The more I enjoy working with this stuff, the more likely it is that it all gets done.

SH094 - Thunder Road

video: Galaxy Express 999 movies
music: Metallica - "Turn The Page" (cover)
link: depositfiles [58.9 MB]
editor: Magix 2.0+ deLuxe
production date: February 2005

This was the first video that I hit Magix's internal 'wall' on. I don't know what this is, even to this day, but it is some combination of length and complexity of video that overloads the editor, a "you shall not pass", that whenever it hits, leaves you no alternative but to go back, blow out the last clip or couple clips placed, render what ya got, thunk out the large raw onto the timeline, uncheck the interlacing options, and pick up where you left off, a wearisome 10- to 30-minute process that strikes about once a minute to once every three minutes in terms of edited video. At the time I didn't know what it was, nor how to handle it; over the three wall hits in this video, I learned, and learned how to anticipate it. Much like the stupid kludge that I had to get through to get my older environment to work, it became something to deal with, not something to change environments over. It's been joined in that caption by interlacing on speed change and the environment's inability to do two filters at the same time, but this was the first and most aggravating.

SH079 - Into The Lungs Of Hell

video: Kujaku-oh
music: Gamma Ray - "Fire Below" (edit)
link: depositfiles [34.5 MB]
editor: Dazzle MovieStar
production date: July 2003

This video was done as a test of zero-cost-of-access AMVing, the idea being to make a video that doesn't require any prior knowledge of the sources going in to appreciate. This is a giant problem in AMVs; on the one hand, people talk down mindless action and insist that story is king, and on the other they go build semantic content into their videos, often without realizing it, that requires prior knowledge of the source in order to parse out the "story." I fucking hate that shit, and ultimately, it's counterproductive.

I got into anime in substantial measure due to AMVs (the rest was an absolutely nontrivial amount of luck in which episode of Marmalade Boy Jo-chan had up on a now-certainly-extinct website in the second week of October 2000), and that would not have been possible if the videos I was watching required prior knowledge. If you haven't watched a show and sit down to watch a video from it and find yourself asking "who the fuck are these people? in what world do these transitions make sense? why should I care?", then the editor has failed, badly, by failing at ZCoA. I'm not going to pretend that my works are in any way perfect in this regard, but I at least try to make what I do independent of the source material, and this one, using a source that I'm pretty sure zero people in the modern AMV audience are remotely familiar with, was the acid test of that. It's not a complicated video, but it's functional and should be intelligible regardless of how much or how little the viewer takes into it.


Of course, I did another experiment seven years later that essentially invalidates this one -- the point of Causality is that due to narrative closure, story is an emergent property of anything cut out of a scripted source just as much as perceptual closure makes synch an emergent property of any juxtaposition of music and video. Even when the editor requires, consciously or unconsciously, that information from the source be taken in, the fact of scripted material being used will result in the naive viewer imposing their own understanding of narrative on the video they're watching. When we assume intent, we try to figure out that intent, at least to create something that seems internally consistent from our vantage point. It may be weird and incorrect, but it'll be there. All ZCoA does is attempt to reduce the error bars on that internal narrative.

SH054 - me[tamor]phitic

video: RahXephon
music: Dark Tranquillity - "Cathode Ray Sunshine"
link: depositfiles [42.2 MB]
editor: Dazzle MovieStar
production date: September 2002

Though it involved 30+ hours of production time, this was another relatively "snap" idea, also coming out of MITAC, indirectly; I hadn't been very impressed with the show seeing it as club, but listening through the Damage Done record for the first time, this song immediately jumped out as needing to be videographed with this source. A lot of the hits in this video, especially the ones that actually work, were done mentally while listening to the CD, in the process of forming the idea. This is actually how most ideas get put together; the exceptional case, like the Polterchrist song I had kicking around in my head while watching Good Witch of the West that hasn't actually become a video yet, goes in the other direction. This is how I get at declaring myself music-first; every idea turns into a combination somewhere, but most SH videos take the form of presenting X music using Y anime, rather than using music Y to say something about anime X.

SH040 - blind to see

video: various
music: Iron Maiden - "Can I Play With Madness?"
link: depositfiles [35.3 MB]
editor: Dazzle MovieStar
production date: June 2002

There's a fair amount of guff in the .org entry for this video about it being broken, and coming out in weird versions, etc., so this is probably the right place to discuss why.

The reason is that MovieStar, while a decent linear editor, was piss-poor at exporting stuff. This is the reason for the very small and very weird frame sizes on the original (non-remastered) versions of SH videos before SH093; well, that and that I didn't entirely know wtf to do with a MPEG2 output. What I should have done was process the final HQ build of the video like a DVD stream (that I didn't have any experience working with yet, but what the hell): drop into DGIndex (or whatever the equivalent was back then), save project, import AviSynth script into VDub, build standards-compliant full-screen two-pass XVID encode like a civilized person.

What I actually did was quite different, though close enough to make any overly neurotic video quality types want to bash me to hell and back with a brick: I exported from MovieStar as a "high quality" .asf (whatever that meant), then brought that into a legacy version of Virtual Dub that had ASF support, which was removed before I started making AMVs even. (In this video, the ASF didn't and wasn't able to export right, cutting the production chain off at the knees.) This was the source for a 3000 Kbps DivX distribution copy with, necessarily, damaged audio, but my speakers were damaged as well, so my ability to hear that the audio was fucked was kind of impaired.

This is the sort of thing that keeps guide authors awake at night: a theoretically smart process that doesn't even know how brain-dead it's being, or how few changes would be required to make it truly intelligent. I obviously had the high-quality MPEG2 builds; that's what all these remasters were done out of. I also had and was able to use AviSynth, at least after I started cutting stuff in Virtual Dub in the fall of 2002; I just didn't put it together for whatever reason. Stuff like this keeps me on my toes process wise: always conscious of needing to be aware of what I may be missing, and what I'm not seeing that I ought to.

Nani Naze Causality

The What and Why of Causality (SH115)

---

By now, I feel that enough space has been given for people to form their own opinions about this video, so I can weigh in without my comments being taken as authoritative about what it is, or what it's trying to do. This was also intentional: this is a self-organizing video, and I wanted for it to at least try, initially, to be self-documenting.

The what, first: this is a video that was to a large degree edited by random rolls (hence the credit to The Machine That Won The War). What I did was roll on a table of all the non-credit cuts in Summer Wars, and place the resulting cut into the timeline until I felt the music changed enough to justify another cut, or until I ran out of frames. Effects are by design rather than random, and at the end of the song I had a separate pool of extremely long cuts to be used as scaffolding that I rolled 1d20 on to place, except for the last two, which were by intent. There are zero frames in this not from the original source; I (mostly) put away my usual tools of negative space, positive space, and trigger abuse in order to concentrate on excessive close cutting to compose an order out of the chaos presented by the literally random clip selection.

The why: This was a confluence of several themes that turned into a video idea. I wanted to do a literally random video (picking cuts with rand(1) or something similar and letting the chips fall where they may) for a while. Since the source collection would be by linearly decomposing the input into its component cuts, it would have to be from movie source, for linearity (because I'm lazy) and to keep the total size of the source pool down. Also, since any cut could come in at any time, I had to avoid the slack or barren animation that you occasionally get in TV anime.

I was looking at The Girl Who Leapt Through Time as a potential source, but that was getting an overused rap on the .org, so I was leaning away from it; I needed a source that was visually striking to work with the assembly method, but I didn't want something where every cut has been seen in a hundred videos before, because that would tip the random-recombination element too strongly. It was in this time that I encountered Summer Wars, having heard a lot of good stuff about it when it was in theaters, and was immediately sold. I ended up not even buying the other movie.

The musical choice was another strong driver; I wanted to do an Arctopus video, but didn't want to kill myself in the process. Random selection takes a lot of the pain out of finding the exact cut, in a pool of nearly 1400, to match to the quarter-second of one of Colin Marston's fills. Additionally, despite being completely unrelated, there was a huge fucking kickup about someone who did a video to the whole of BTBAM's Colors record (which, really, is not good enough to get such treatment, especially in contrast to their earlier albums) and wouldn't acknowledge its weaknesses. I get embarrassed and defensive about such things, and it's not honest to claim that there wasn't at least a little motivation, as long as I was on this project, to show that you can do ambitious projects with long runtimes, absurdly technical music, and challenging concepts and not have it come out like crap.

As things developed, the synergy of the source with the concept drove the project further. The central idea behind the video is mass parallelism, and the implications that has for causality (hence the title). If everything is happening at once -- as manifested by the blizzard of random-selected intercutting -- what prior event is what consequence connected to? Is it possible to pick that out? This is kind of a big theme in the first half of the movie; a case of mistaken identity caused by a scattergun social-engineering attack, with its aim parallelized cracking. The ideas of parallelism, distributed resources and the communication models that bind them together, and the complementarity of social and technological approaches to the challenges of the modern world are, of course, everywhere in the movie, and it's difficult to think of a video that would bring them together into some kind of coherent whole that did NOT come out looking like this one. (Well, fucking duh, this was the Summer Wars video that I did, and if I was capable of looking at the material in a radically different way, I wouldn't've come up with the idea.) This might be why, at the halfway point in the production when I checked, there were exactly zero full Summer Wars videos out.

The random-select concept made the video easier but this is not the same as easy. Not by a fucking long shot. I still had to do all the actual intercutting to make things work, in addition to the rolling, and I also had to spend nearly 15 hours cutting up the movie in the first place, then moving it to an environment where I could keep 100 minutes of HD source on-disk rather than on remote storage. While the base idea of decomposing limited source into its component elements and recombining those in time with the music echoes back to an earlier day, actually doing so was a significant challenge even for me as an experienced editor.

The only regret that I have was that I didn't do this all in AviSynth, or that there wasn't a transparent way to do that. Why? Because I could have distributed a master script calling a synthesis of 500 or so smaller scripts (plus of course the index, omg bootleg), it would be a lot smaller in footprint, and I could claim to have out the first interpreted rather than compiled AMV. As a once-and-maybe-future Lisp head, this would have been neat, but in the larger scope of things is really not that important, and nobody would have the processing power required to interpret a video like this with any kind of smooth playback at the time it was released.

SH030 - BLOOD REIGN

video: Mystery of the Necronomicon
music: Hypocrisy - "Carved Up"
link: depositfiles [38.9 MB]
editor: Dazzle MovieStar
production date: January 2002

By this time, it had been three months and eight videos since I'd done something FUCKING METAL. This was and remains too long for active operations. It's also a testament to what working conditions were like back then; I could get back to school, cap two hours of video, cut it down to strict brutality, avoiding the porn parts, then pound out a draft and a finished version and be done in about 12 hours worked, before classes started again. Nowadays, I seldom have 12 hours over two days to devote to AMVage, let alone the initiative to do so, but part of that is also that standards change. Cutting takes longer in VDub than it did in MovieStar, and the ability to fine-tune cut length and apply effects forces that work to actually be done, to make sure that the video's coming out optimal. Back in MovieStar, if I missed a hit, I missed a fucking hit, or just edited the sequence differently; not too fucking much that could be done about it.

SH028 - and so it ends

video: Rurouni Kenshin OAV
music: Blind Guardian - "And The Story Ends"
link: zippyshare [68.3 MB]
editor: Dazzle MovieStar
production date: January 2002

This sort of video is what keeps me editing; something that somehow catches lightning in a bottle and matches up to the standard of the music. I can't remember what the production queue looked like exactly in January of '02, but I do know that this came forward because of source issues; I had the Kenshin OAVs on hand, but not much else, and could get access to a copy of Imaginations..., even though my own was still up at school. (This and SH029 were cut up over winter break when the demo finished earlier than anticipated.)

This one also shows that you don't have to cut the everliving fuck out of a source in order to make a decent video out of it. In addition to the demo 2 charts in the last post, I also dug out the complete project file for this video, which WAS pictured below on a 6"x9" sheet of note paper before ImageShack died.


[pretend there's some 15-year-old scribbled timecodes here]

Seriously, that's it. Less than 80 cuts for just about 6 minutes of music. No effects beyond crossfades. You could get away with that in 2002, and if your source and concept are good enough, you can probably get away with that even today.

Charts like this, with explicit end marks for all cuts that went into the final video, exist for all non-demo vids through SH038. This dictates, basically, what videos I'm allowed to think about remaking and which I'm going to be too lazy to ever consider challenging.

demo 2 and the Musical Chairs Demo guff

This was alluded to in the post for SH025, and since I was pawing through my project notes last night anyway as part of setting up the rebuild of SH007, I dug out the following pieces of paper to show the evolution of the demo.




This shows the initial plan of what the demo was supposed to be. The videos referenced, for those who can't make out my internal shorthand, are in order:
BAS06
SH025/26; the script from SH025 with the song from SH026, which was originally considered with "Coerced Coexistence", also off the Colony record
SH018
SH026, at least the script anyway, using "Shades of Revolution" from the Kenshin OVA soundtrack
either SH023 or SH020
SH027; well, the script, using a song that would eventually show up in SH048
either SH021 or SH023
BAS11

Things started changing almost as soon as the ink was dry. I had plotted out, at least initially, when I started writing the scripts, a demo that appealed to me, a demographic that at Bowdoin was already in the damn anime club. The first step was to dial the metalness back; this resulted in the more accessible In Flames song getting picked for the new-build video, and SH021 getting cut in favor of another new video that would become SH024. The next step was to make things snappy as well as diverse; SH023 got in over SH020 in that slot, and I had to do something about the seven-minute song assigned to the SH026 script. That something turned out to be slotting the song from the SH025 script in, and throwing a new song in behind the SH025 script.

The last change, for which I'm still not 100% on the reason, was to set the song for SH027. It's probably that The Dreamside were coming off as too kvlt, and I wanted something more accessible, but still fitting the emo parts of the script. Also, "Mirror Moon" would really dominate the script out, which was counter to, um, the entire purpose of making a demo video. One way or another, things got worked over into the form that demo 2 looks like now, as shown in this shot of the credits script:




The slightly different ink color shows that the two substitutions and one addition were done significantly after the script was originally written; this version is what shows up in BAS11.

SH021 - indirect

video: Akira
music: In Flames - "Pinball Map"
link: depositfiles [42 MB]
editor: Dazzle MovieStar
production date: November 2001

This was the first video that I used rough-drafting on, a process that would endure for the next 71 videos, bar one that I did in Virtual Dub alone, and demos 2 through 4 for which that would have been a fucking waste of time. What this involved was, when cutting up the capped MPEG2 source, making two clips, one at the normal, original resolution, and one in crappy MPEG1 with an otherwise-identical name ending in _d (if I recall correctly) for "draft". When making the video, I would use only the draft clips, for better live synch, and then once the video was drafted to satisfaction, go back and rebuild it out of the high-quality source, matching the mark points as noted on the sheet of paper that was serving as the project file.

I could get away with this, of course, because MovieStar was basically a linear editing environment. Fucking around with clips, limited only to shortening them anyway, was a deadly bad idea for stability reasons, and similarly, the only effects were transitions and the only transition that didn't look like garbage was the crossfade. This limits what you can do, but makes it a lot easier for a learning editor to get themselves established in this kind of process.

Paper project files and the drafting process had their limitations, though; on SH063 I left about seven cuts out of the paper notes when drafting, which made things harder in the final assembly process than they needed to be, and on SH090 I accidentally destroyed one of the HQ cuts while making the drafting version. Of course, this was the luckiest video for it to happen on, because the source was an upscaled 320x244 ASF so I wasn't really losing any detail, and the video wasn't ever going to be generally distributed anyway.

Shin Hatsubai 2009 .org stats, now with pretty graphs

One of the neat things about having a stupid amount of videos on the .org, as a stats nerd, is that you get to have your own statistically significant data set to draw conclusions from without being an admin. Here are some numbers (actually, graphs, because this is MAI data and NO U CANT HAZ, make your own 83 local videos) from 2009 and some probably bogus conclusions about them.



This is a graph showing SH downloads for the past year by volume; that is, how many actual times somebody pulled the video from the .org servers. This is a fairly normal heavy-tailed distribution, with a moderately broad shoulder. Does the .org look like this, as a whole? I'm not sure, because 83 is really not a big enough sample to talk concretely about a sample space three orders of magnitude larger, but I'm willing to bet it looks something like this.



This here is a graph of SH downloads over the same period by weight (in MB); that is, the actual impact on the .org server of transferring the downloads graphed above. The shoulder is smaller -- if taller -- here and the tail falls away at a much faster pace. Does the .org look like this? Since I don't have the whole dataset, it's not a firm conclusion, but I'd hazard a guess that it probably does. The number of high-demand videos is very small, and the decay rate off that looked, the last time I looked over the bracketing of the star scale by week, seems to mostly match with this kind of slope.

Now here's where things get really interesting (well, at least as far as graphs based on statistics about AMVs can get, anyway).



This shows volume as ordered by weight. The sinusoidal ebb and flow between short ("fat" videos that reach a higher (further leftwards on the graph) transfer on fewer downloads) and taller ("thin" videos that need more downloads to get where they are weight-wise) bars in the tail is interesting, and indicates that .org users, at least the ones who download SH videos, are not selecting on filesize.



This one here is weight ordered by volume. This is the first time I've seen a sawtooth wave in an AMV-related graph, but all this is doing is pulling the information in the vol-by-weight graph out a little more dramatically. The sawteeth in the tail show transitions between roughly equivalent levels of transfer; the spike that starts is the fattest video with ~x downloads, and the last one before the next spike is the thinnest.

What does all this mean? Basically, that the motivation to compress your videos correctly is social rather than egotistical. A-m-v.org users are not selecting on either heavy or light videos, but, dun dun dun, presumably on some function of content and popularity. You won't get more people to download your video if it is the smallest filesize in the list of their options; the only motivation to compress properly is to minimize your bandwidth footprint. This is a social motivation, not one for the individual -- if you need a kick, though, donate, and the usage of the local server then does become more directly your own issue.

buried by time and dust

INSO15 isn't done yet -- though it will appear in this space after a) getting Youtube'd; b) I PM the link to the band on myspace and get some reaction other than wtfdood this is garbage; and c) both the streaming and HQ links show up on RTTP. In the meantime, though, I happened to run across the following guide to digital video editing:

Digital Video: A Guide for the Fearless

The guides that I had access to when I started editing were about as primitive; I'll see if I've still got any of them around and may append them later. What's interesting is that not only has this guide survived for 13 years online, but large portions of it are still valid even in the age of BluRay, Blender, and After Effects. Some, though, like the stuff on Quicktime compression, are pure lulz for their datedness, and the section on doing your own embeds will probably have some saddos moaning about starry wisdom lost; way of the world.

The Tube Is Good And MEPs Are Bad

As noted, I make too many AMVs. Or, more accurately, I have made too many AMVs, such that I still average over 1 VPM over the total length of my career despite being nearly retired for large parts of 2006 and nearly all of '07, '08, and '09. The normal reaction is to look at that back catalog and say, wow, there's probably a lot of crap in there. Yes. There definitely is, almost certainly more than even I'm willing to admit. But that's kind of the point.

I have an overdue reverence for Robert Rodriguez, mostly because he has a knack for making cool movies, but also because he spends his commentary tracks not talking about his artistic vision, but about production tricks and how he's made his DIY ideals into an actual production process. The commentary track to El Mariachi is a bible for anyone who thinks they don't have the resources to do what they want to creatively, and towards the end of the opening credits, Rodriguez comments that he had a great advantage by making movies on home video before going to film school, let alone actually filming anything. His quote: "I believe every filmmaker has about thirty bad movies in them, so it's better to shoot on video and get them all out of the way, so that you don't shoot them on film and waste all your money."

The point is not, oh, go make 30 crappy videos (or 30 crappy 80-minute video movies) and magically become better, but that practice is a very real part of becoming better at any craft. You get better at making movies by, surprise, making movies, and you get better at making AMVs by making AMVs. You need to have the technical basics as well, and you need to have an idea of what's worked in the past, too, but Rodriguez still went to film school despite all those video movies he made in high school; they may only take you so far, but the guides and watching previous videos shouldn't be just skipped. If you want to make better AMVs, you need to make AMVs.

This is where the Tube comes in, obviously. You can dump whatever the hell you want on it, and if THEY think it sucks, no harm, no foul, delete and learn from it. The next part should also be obvious: this is why MEPs are bad.

MEPs are bad because they take nearly as long to produce as AMVs, but are not AMVs and do not teach you how to make AMVs. What MEPs teach you is how to make 30-60 seconds of sparkly crap. If you want to know why the well is running dry on the .org, look no further than the MEP forum, and how many thoroughly needless projects are underway. If you want to know why your work sucks now, look at your prod history and see how many MEPs are up at the top. MEPs are nice for building community and making something that everyone can feel good about for 30 seconds before going on to the next one. They suck balls, though, at making better AMV editors.

Your frame of reference as an editor is not infinite. You have to train yourself to focus and concentrate on long projects, especially to see the video as a coherent whole where the end follows up logically from the middle from the beginning, and where the beginning sets up how the video flows in the middle and how it concludes. In a MEP track, everything is compressed, and it becomes more difficult to maintain the focus on building a full video -- especially when it's so easy to drop the project and say "screw it, I'll just do a track for Project Timewaster instead".

It also doesn't help that MEPs are easier to come up with than AMVs. For a MEP, usually someone else has picked the music, and maybe the anime depending on the project, and all you have to do is come up with some bare-bones editing concept. You don't need to watch anime and listen to music extensively, which is required for doing regular AMVs. Rather simply put, MEPs mess up your focus frame and let your source concept synthesis abilities atrophy. They don't make you better, and may actually make you worse. Stop joining MEPs, and stop starting new ones.

Yes, I've done them as well, but in my archives, two of the three that I've been involved with fit this pattern. I got to pick all my own tracks for all of them, but AMV Hell is the king of focus busters. The Graveyard handed out general concepts in advance and was, for reasons outside of my control, a huge waste of time in the end. Conet is the only exception: godix invited participants that he knew would meet standards, told them to select-from on a large database of tracks, and told people to come back with something cool. The only pick points might have been the shorter track length and the competitive compulsion among everyone but doki to break the video in a visually distinctive way....but then again, you look at the editors involved, and most of us have a history of that anyways, hence being asked to do this.

As I work my way back into the hobby, it's something to be aware of as much as for anyone else who wants to develop: stay away from short videos and stay away from MEPs. They're nowt helping.

math, graphs, and process crap part II

More process stuff from the re-assembling of my stats sheets that look to have been lost when Keystone went out of service.




This is a graph showing the average number of seconds per video element (spc, seconds per clip) for the SH videos up through SH110. There isn't as much interesting about it mathematically as the one on process time above, but let's see if there's anything else that can be gotten from it. Open the graph up full size if you want to follow along; this one's more squintable, but it's still easier to look at 110 data points when the bars have more than a pixel each.

The first look shows that there's a lot of variation in the sample space, but that spc generally tracks better with time than with video style. The second is that in that regard, there is a pretty uniform falloff in spc from about SH034 to SH054, and another, less notable, drop in average in the early 90s. The second one is easy to explain; I switched editing environments and was more able to employ shorter cuts, since I wasn't restricted to using only clips longer than 15 frames/0.5 sec, or using exactly the clip as cut, modulo any crossfading with the next one in sequence. The first drop might be attributable to re-joining the .org, around SH043, and picking up more pressure to cut closer and make internal synch secondary, but the trendline does go back a couple months even before that. The .org may have contributed to this trend and pushed it forward, but I was already doing the things that resulted in a dropping average spc even before I got my delusions of competence reset.

The video that this trend starts on may provide a clue as to why it started. SH034 sucks. It sucks hard, and I became aware of it doing so soon after the video finished. In reaction to a video that was a pretty good idea coming out like absolute garbage, I did a deep and thorough re-examination of everything about my process to determine where I went wrong, and what I'd need to do to improve and avoid this kind of situation in the future. One potential solution path was to change editing environments, but the trial for Adobe Premiere fucked up my capture board and kind of put itself permanently off-limits. I still have a Verbot against using Adobe editing products as a result of this. The other solution path -- since, as noted in the video entry, many of the problems came from a lack of source and cocommittant overuse of filler -- was to:
a) be more aggressive in gathering source, and conscientious in tracking source volume
b) be more proactive in editing long songs
c) be more willing to cancel videos in progress, especially for source availability reasons

All these things contributed to a falling average spc. Part b) probably the least, but a song that's 30 seconds shorter because a repeated chorus got cut out has 30 fewer seconds that need to be papered over with 4-second filler cuts. Part c) kind of disappears in the results, but there was a Linkin Park video after SH035 also using Yumede and a Nevermore video after SH061 also using WHR that never saw the light of day because they were burning through the few remaining good cuts at a rate too fast to be sustainable.

Part a) was really the most significant, though, and though I was still working from subtitled sources for nearly everything after SH034, the focus really became on getting all of the frames that I would need, and nothing else, in each clip taken rather than "let's just grab every continuous extent without subtitles on it that MovieStar will accept". Of course, I was still taking everything, but in that flensing process, naturally grabbing more smaller cuts that I might have neglected before. The more one-second interrupted-action cuts available, the fewer six-second pans need to end up in the final product. These changes necessarily took a while to accumulate, but over that six-month, 20-video stretch, a lot of effort and introspection was put into continuously improving the process, which paid dividends further down the line; dropping the average spc into the minimum possible range for AMV with MovieStar was pretty much a side effect, but it makes for a nice trailing indicator. Your videos won't get better if you just deliberately reduce your spc, but as you improve, your spc will trend down towards the minimum compatible with your style and your editing gear.

Stuff like this shows the reward of keeping stats on process; you never know when you're going to have a statistically-significant sample space on your hands, and at that point it becomes interesting to look at the numbers and see what you didn't notice happening while you were making all those videos. These are the only tools I have that are of real use; only total number of video elements is tracked in addition to these, and that's just (spc x however long the song is), which is a bad indicator for what kinds of projects are under survey and useless for everything else. If there's some other tool that you think would suit your process development to track, invent it and track it.

math, graphs, and process crap

This is process stuff that maybe ought to be in my .org journal, but I can use pictures here, and not there.





This is a graph of the production time officially noted for every SH video so far. Let's see what's of interest. (open full size if you actually want to follow along, don't just squint at the preview.)

The first and most striking thing is that there is a sine wave rather clearly apparent in the results, with a period/wavelength of approximately 9 videos. This is not perfectly consistent, but it's remarkable that a pattern like this is detectable at all. Even from someone (perhaps one of the few) who keeps stats on their process, it's really weird to see an emergent mathematical pattern in the data.

The second is that most SH videos finish within 20 hours of official start of work, and nearly everything is done within 40 hours of work. This helps explain the frantic pace of completion in the fall of 2002 and winter/spring of 2003, when I was job-hunting and GRE-prepping about 25 hours a week and spending about 25 hours on AMVage. The third is that there are quite a few videos, usually bunched together, running well over 60 hours, but this behavior disappears entirely after SH053. What happened after SH052 that resulted in this sort of spike disappearing?

Looking at where the spikes are, and what videos are represented, there is a simple formula at stake: videos that spike out were produced from digisubs, largely or entirely, and were all done over at least 10 episodes. That the spikes exist, and that those on the demos aren't even higher, is an artifact of changes in the clipping process and how records were kept.

SH videos have always been assembled out of clips rather than from regions of episodes selected in the editing environment. However, the process by which these clips were created has changed several times, and until SH093 when DVD rips came in at long last, some of the stages were dependent on whether I'd capped the video or was using someone else's encode. Up until SH052, the digisub process went like this:
- render source file to MPEG2 (usually; some stuff for demo 1 was done to MPEG1 because that was what I was using at the time) with TMPEG
- clip in MovieStar, producing full-quality MPEG2 clips and low-quality MPEG1 clips for drafting (videos from SH020 to SH092 were drafted with MPEG1 clips, due to problems synching SH019, then reassembled from the appropriate MPEG2 source for the final build)
After SH052, the process went like this:
- clip in Virtual Dub, processing the .avi to 640x480
- render clips to MPEG2 with TMPEG
- produce drafting versions in MovieStar

From the journal at the time of the change:
"I'm trying out a new clip-collection process on the current video as kind of an experiment. With my current hardware, I need to put in 17-18 hours of processing time per ep before I can even start thinking about getting clips. I'm trying now to make clips in VDub, run 'em through TMPEG to make the HQs, then make the LQ draft versions in my main editor. This is all attempting to save time, on the proposition that even if I get 3 minutes of usable source from an ep, that's 19 or so minutes that I'm not able to use in the video and shouldn't waste time working with."

This held up, and the precipitous drop after SH052 is notable; only two later videos -- one involving a lot of building from manga cuts, and the other thanks solely to some encoding mistakes -- even go above 50 hours, while before this it was virtually guaranteed that a "serious" video would require more than 60 hours. The sole reason for this was the extended render time on Keystone, which has been noted here, below, as a 500MHz K6-II with less than 200MB RAM. As noted, each episode represented between 10 and 18 hours of effectively dead time running the render, which all went into the production stats, since this was effectively the same step as capturing the source, which was what the process tracking was initially built on. When this is taken out by clipping in VDub, the resulting stats are a hell of a lot closer to the actual attention time given to the video; speed was improved along with tracking accuracy.

The demos (demo 2 and 3), incidentally, are where they are, despite each using 20-30 episodes' worth of source, because common production time was reckoned proportionately on each demo for the videos included. Each video has about four hours of unique production work, something that you can also pick up by looking back at demo 1; the rest of those 70-95 hours are its share of the rendering and clipping that produced the source pool. If that production time was counted in full on each video, the demos would look like each video was the end product of 250-350 hours of work, which is even less accurate as far as effort goes.

Thankfully, all this is in the past. The limitations of drafting, NTSC-TV framerate, and cuts no shorter than 15 frames are no longer a part of how I make AMVs. A good school in many ways, but one I'm glad to be out of.