anubisx00 wrote:
All I need to know is which way to encode it. *EDIT* About the 29.97 fps, to get that I need to insert pulldown, right? Vegas gives me the choices of 1)29.970 (NTSC), 2)23.976 (inserting 2-3 pulldown), 3)23.976 (inserting 2-3-3-2 pulldown). So which one? Not 1 I would assume. Another question if I may. Should the audio be set at 44,100 Hz or 48,000 Hz? The music was from a CD to begin with. One last thing, what if Vegas' DV file is proprietary in some stupid way, it is in the .avi container. If it is borked, what are free DV encoding possibilities?
NTSC is precisely the desired format. 29.97 (or 29.970, the extra significant figure is pretty meaningless in digital video, though not entirely negligable in the analogue variety).
Standard DV and DVD-complient MPEG-2 both require 48KHz audio, to the best of my knowledge. If you can't get the 41.1KHz CD track to nicely resample and dither into 48KHz, I can take care of that. It's just an extra demultiplex/remuliplex operation - a little time consumung, but nothing difficult.
Whether or not Vegas does something strange with the
container it exports DV in, it is highly unlikely that the DV stream itself is anything non-standard. Part of the beauty of DV is that it's simple, clear, and professional enough a standard that no one half-asses it and calls their implementation "good enough" in the manner some MPEG4 implementations were churned out.
If it uses AVI as a container for DV, that's unusual, but no problem. Once again, I can just demux the video stream out of the AVI. Really, the process of producing "distribution" encodings requires a demux process anyway, the only difference being whether or not it has to be done "by hand".
As a side note, DV is technically supposed to be in it's own container (25 Mbps DV and 48 KHz PCM audio in a .dv file), but Apple saw to it a while back that Quicktime is the container at least half of the professional video editing community is now using for their DV files, by making it the required container for DV editing in Quicktime and Final Cut.
----
If you have problems with interlacing, don't worry. I can take care of that too - especially now that I've actually figured out how each of the deinterlacing techniques works (on a detailed level) and the conditions under which one or the other needs to be applied.
----
I've finally gotten around to installing a FTP server on my desktop (which I re-imaged a while back after making a real mess of it during my first attempts at compiling this project). I'll pass out an IP once I'm home for vacation and have the thing plugged in somewhere (which should happen by the end of the weekend).
----
I'm on vacation now, so I can actually work on this for the first time in a couple of months. I've already gotten quite a bit done since 2:30 (when my last final let out), and now have my system fully prepared for running all the transcoding scripts and narrative renders (those need to be re-done in full DVD resolution - and I might as well do all of Wendy's sections with the three-frame lip-synch Miyazawa I ran for the last couple of narratives (or maybe just a couple of tests, I forget these things over such long periods of time), as opposed to the less smooth two-frame version.
I've also, aside from installing a bunch of software with my newfound knowledge of Linux package managers (wonderful things, really, once you discover their existance), put together a "remastered" version of 1337. It's sort of the same video, but with straight-from-DVD VOBs instead of re-encoded versions of fansubs found on Kazaa ("there's lots of DBZ footage on Kazaa...."), some restructuring to make it a little less "linear" and replace the relatively dull "bridge" scene with something more interesting, and sans the "blown-out" (film term) look. It
is still greyscale (though the color render was also quite attractive, in a different way), but this time it's
all greyscale. I also tried to simulate film grain and some "old film" junk with a noise generator and some compositing - but this threw the encoded bitrate through the roof at any reasonable quality, so the super-pristine version won out over the 16mm "experimental student film" wannabe-noir version. Really, it would make more sense to add the grain
during playback, but most players aren't really capable of doing that without just making the video look like a giant pile of poo with some flies swarming about it (inside joke involving a recent RIT student film screening, and one of my roommates' observations on a common theme in "experimentals").
It's amazing how fast you can edit when you've read a half dozen books on codec technology, live with a pair of filmmakers, finally have a "fast" computer, finally have a version of your editing app that doesn't crash every ten seconds, are working with clean footage, and already know pretty much where all the cuts have to fall (though only about half of them are the same, and a lot of fine-tuning has been done).
----
As a(nother) side note, it's actually kind of interesting to watch an anime in "black and white". Doing so brings out a lot of stylistic detail that seems to be masked by all the vibrent colours we've grown used to. Looking at the FLCL footage, I've actually come to appreciate the amount of care put into making the visual storytelling something that truely
needs to be animated, yet follows many of the critical aesthetic guidelines of filmaking that most animation seems to ignore. Nearly every scene has a white reference and black reference, the forms and motions of the characters pay close attention to silhoutte value, depth of field is actually emulated through the use of much softer tones and lines in the foreground and background (along with some actual loss of focus, in places), and so on. Just something I noticed and thought was cool.
----
Oh, hey, a message from Helen... I wonder what it says...
[wondering if the "topic review" is automatically updated, and, if so, how long I've had this little message window thingy open]
Yeah, I'm back. Disappeared for a while, but I finally get to put my schoolwork aside for a few days and take care of this (and a couple of other things I've been largely ignoring for quite some time now - like sleep and human contact outside of the whole five people I ever spend time with in Rochester [wondering what happened to the much more social lifestyle I enjoyed back in high school...]).
Anyhow, it's nice to hear that someone else has also produced a "new and improved" version of their segment. Though I'm actually a little surprized to hear that "Forbidden Memories" was that other segment. That's some serious attention to detail.
[Now realizing that there are quite a few posts in the "topic review" that I hadn't noticed, since I got here through an e-mail link and must have missed the last page of the thread - it's all making sense now].
----
About the DVD/Online distribution thing:
I'm thinking Bakadeshi and I should send a couple of e-mails to coordinate the DVD compilation and make sure each of us has the best possible set of files. I have a burner now, and have actually figured out the file system stuff needed to make a DVD (albiet one without any terribly fancy menus or other such interactive features).
I also have a couple of questions regarding submission formats and the like (and trying to save everyone from producing and uploading several different encodings of each segment), and whether it might be best to package the project as two DVDs (widescreen and fullscreen) to take advantage of anamorphic widescreen and twice the bitrate. It would add about $2 to each copy, but I don't think that's bad, especially with the cost of fancy packaging.
I've also decided that,
if the DV compilation in any way fails to work, I'll set up the project as an account with multiple entries - each containing the narrative, title, and video for a single segment. I've tested those smaller joins - but I'm not sure about how the full-length join will work (what with the 4GiB limitation on single files that my file system imposes and the fact that I
haven't tested a DV-based join of anything more than a few minutes long). I'm still going to try to get it to work, but I'm not going to hold up the release by another 10 weeks when there's an obvious workaround to this last critical problem. There would, of course, also be "Introduction" and "Credits" segments, if the multiple-file release is pursued.
----
About the aspect ratio problem from a while back:
It's solved. The new 13'37" is 4:3 (no subs in the new source footage, and the 16:9 wasn't really adding anything that wasn't offset by some of the awkward cropping it required). Thus, it can be switched with "A Boy I Knew" and "Ararat" - making those 16:9, as they should be.
The catch: Wendy (Songbird) is going to need to do a narrative for 13'37", and I'll have to come up with narratives for "A Boy I Knew" and "Ararat". Still, this solves a pretty big visual problem and maintains balance between the two sections of the project.
Improvements include
[/size]