-
has taken a turn for the worse
2009-09-19 18:45:22
SH Remasters Project: The Absurdly Long Upload Struggle
--
local: 45/69
indirect: 23/39
A productive day for indirects, even though the two locals put in for Friday are still pending; at this point, all of the indirects connected to a .org catalog entry are up. The remaining ones will get linked here eventually; I need to do a bit of dev to accommodate them first. Maybe not by tomorrow, but things will go forward at their own pace. More than half done; not bad for something that should have wrapped back in March or so.
Preview oddities: the sitewide project to render every video on local into a previewable version is either stopped or, alternately, chewing through the gigantic bulge at the top of the Poisson curve in the star scale. Having a large number of mediocre videos, I can pinpoint that it is somewhere between 3.41 and 3.39; everything I have averaging 3.41 or higher (there's a 3.4 in there as well, but I think it's dropped recently from a higher score) has a preview, and everything at 3.39 or below that hasn't been deleted/reupped since the start of the preview push doesn't. Despite the fact that everything older than SH110 has a large number of stars attesting the average, the current leaders in preview numbers all grade out under 3. People are watching these videos despite large amounts of evidence that they won't like them; maybe this is why, because they half trust their peers and half don't, and want to make sure that they aren't wasting the download.
Maa ii; at least they're getting some Skyclad before they decide the vid's not worth the candle.
--Kai out
-
zeitgeist treachery
2009-09-18 13:45:06
SH Remasters Project: The Absurdly Long Upload Struggle
--
local: 45/69
indirect: 14/39
Apparently there was a change in staff about midday yesterday, as the responsiveness on deletion requests has gone way down. Alternately, whoever was handing it may also have gotten justifiably fed up with the volume of video deletion requests coming from this end, and decided to just handle it once a day, which is understandable. Even under such constraints, though, I should still be able to keep going at current pace and finish before the end of the month (ish). Over the weekend, the plan is to push as many of the indirect out as possible, so that that caption, at least, is over and done with....and to fix that problem stretch in SH111 that is putting a mental block on the video.
Another long-term project (having nothing at all to do with AMVs) is wrapping this weekend, and a mostly-parallelizable one is in the process of starting. As noted earlier this year, I have too goddamned many CDs, and will be reducing this number, if not significantly, at least somewhat, over the next six months to a year. I started in on this last night and quickly became depressed at the magnitude of the task at hand (20-25 vertical feet of discs processed, another 16-20 remaining) and at how few discs, relatively speaking, I cared little enough about to put in the 'get rid of' pile. Said pile is probably going to have about 400 records in it at the end, leaving about 1200 that I have to find someplace to put if I'm going to make a serious effort at processing out my VHS collection. Both of these things peripherally touch AMVage: reducing out CDs means that I'm (probably) never going to a video with a song off any of them, and converting VHS->DVD (which I can't do yet) will make more and vastly more obscure sources available. Given how the schedule is, and how little time I have to edit now, and how much in the way of resources these projects are going to take, though, I'm not sure that this is going to lead to any real new videos.
It may be a multi-year project, but one way or another, I am going to get all this damned media out of my damned closet.
--Kai out
-
tonight i channel | the sorcerer
2009-09-17 11:48:57
SH Remasters Project: The Absurdly Long Upload Struggle
--
local: 41/69
indirect: 14/39
This is an impressive leap forward in the space of day. Things will hopefully click along at a similar pace through the rest of the week, letting all the remasters make it up by the end of the month. The coming weekend may actually see some video progress as well.
An interesting thing to note about doing all these damn reuploads is that doing so is making my historical bandwidth-used figures fluctuate. Going into this stage of the project, I was at about 901 GB all-time (sure, laugh, six years of ~60 videos on the Donut and not in the terabyte club -- that's what comes of middling skills and unpopular sources), but at various stages of the process, it's gone down as low as 891 and been up as high as 903. This variation comes from the fact that the figure is not coming from any log but from simple calculations on the weight of the videos. The .org tracks how many times a video has ever been downloaded, and what its current filesize is, and figures that x * y is good enough for most normal people's purposes. In my case, it probably will be as well, due to the pure crushing weight of large sample sizes. Over the 69 videos that are getting added/replaced (SH110 is in the stats, because it was new when this project started), it's likely that the weight gain of proper audio tracks will be balanced by savings in video compression, even if some individual videos are greatly larger or smaller than their previous versions.
This doesn't change the fact that it's weird to see yourself suddenly "having used" two DVDs worth of bandwidth less than you used to. This isn't what Primordial means by "rewriting your own history".....
--Kai out
-
break the laws of maat
2009-09-16 12:03:01
SH Remasters Project: The Absurdly Long Upload Struggle
--
local: 22/69
indirect: 14/39
At the very least, I'm going to finish uploading the remasters. Seven months of inaction driven by, among other things, moving and the festival season, make the idea of finishing SH111 a little more dicey, but that may actually see some movement as well. Presently, I need to go back in what's done so far and fix a fairly major videography error driven by mistakes in translation (note: read lyrics in the future first); if I can a) make myself do that and b) make the resulting fixed section not suck, this will be a good sign for future progress on the video, which may yet finish before the end of the year. The rest of the schedule, obviously, is going to slip into 2010 and later, but either will not get cancelled (in the case that this one finishes) or will hang about forever in limbo (in the case that there just isn't a SH111, and things stop where they are).
--Kai out
-
wild sage grows
2009-02-19 12:01:40
SH111:
- Source rip: complete
- Music: complete
- Precleaning: complete
- Storyboard/planning: complete
- Clipping: complete
- Edit: 1:35/5:01
- Postproc: none
- Export: none
You know it's a tough video when 20 seconds edited in an hour is significant progress. I've finished the first chorus and am across into the third verse, which is optimistically planned to be done on Friday. What really gets my goat about this is that I effectively have to edit every second of this video twice: once to do the actual videography, and then again to be sure that the effects tracks are in the right place and behaving as intended. It's definitely a learning experience, but the lessons are actually being learned, and the "dailies"/"nightly build" (depending on whether you come at this more from the film or the programming side) gets better as well as longer after each successive editing session.
This is a habit that I've never really done before, since I have a tendency to edit at 1 mph or better and, lately, do short videos, but in this case it was necessary due to the drastically slower pace and shorter edit availability per day, and it's been a great help. I still don't do much in the way of true non-linearity (e.g., jumping around as opposed to editing front-to-back with multiple tracks), which helps the utility of this, but those who do might just do multiple dailies for the sections they're working on. The ability to sit back on the other side of the room (to keep from obsessing about the brick that Magix hits the video with when doing MPEG2, even progressive) and watch the video go as intended helps to take you out of the editor's mindest where you're seeing a 368x240 preview window and three seconds at a time, and into the viewer's shoes, where all those three-second bursts have to hang together, one after the next, and not look like ass on full screen. I've added some effecting because of this, and removed some other effects based on similar observations; I'm up in the air about making some other sections less complicated, but that'll be a re-view of the current build before I start proper work on Friday.
Another fact probably of interest only to me: the largest volume of edits in any SH video to date is 290 (in SH102, which combines a lot of close-cutting and overlays with a six-minute runtime, for those interested). Despite taking out some effecting after building last night's daily, the first minute and a half of this one runs well over 200, so that record is going to be shattered flat to pieces long before the five-minute runtime of this one concludes. This is a concern not only for internal statistics purposes, but because prior videos have shown that Magix becomes unresponsive after a certain nebulous limit of complexity, currently suspected to be based on some intersection of total clips and total runtime, is reached, and if the pattern holds, I'm going to have to do the prebuild workaround for this issue twice. No consequence to anyone of note; relevant only to my inability to document the storyboard with pen and paper, my frustration level in rebuilding the project multiple times, and likely, even further delays in this thing actually getting finished.
--Kai out
JOURNAL: Kai Stromler (Kai Stromler)
Current server time: Dec 25, 2025 15:53:59