Video Information


  • Member: ZaCloud
  • Studio: DZX Productions
  • Title: I Ain't Marryin' No Friggin' Horse!!!
  • Premiered: 2008-12-16
  • Categories:
  • Songs:
    • KT Tunstall Big Black Horse and the Cherry Tree
    • KT Tunstall Big Black Horse and the Cherry Tree (original version)
  • Anime:
  • Participation:
  • Comments:

    Note: This video is nothing but singing and dancing. If you don't feel like watching singing and dancing, then watch sumfin else! ;)

    Don't think of it as a typical AMV... It's basically me choreographing a dance, with only a small set of moves as available steps. Where other AMVs are football, this is figure skating. Or something...

    ...And feel free to skip all the below wangsting. :P

    Nobody will ever know how much work I put into this…

    Nearly all the effects are invisible, smoothened, or streamlined, or quick, and cannot easily be seen. But this puppy is chock-full of ‘em. They’re just humble. Their complexity is masked in simplicity.

    This was a project all about improvising, lip-syncing, and simple fun. When I heard this song, I just thought, “Wow, that kind of voice looks like it’d match Haruhi’s face… I can totally imagine her singing and dancing to it.” Got home, it hadn’t been done, so I got to work… That was about one year ago. @_@

    I was limited in opportunity to work on it, my job underpaying and overworking, in need of recuperation whenever I wasn’t working, so often helping out my family with my dad ailing so often. And I’m an emotional editor, so all the technical adjustments I needed to do were often daunting and exhausting.

    But I got through it, at last. First I composed the dancing for the most part, only placing a few native lip-syncs that already fit based not just on lips, but also head position and posture to fit the attitude of the singing. The majority was choreography first.

    That was what sets this video apart from most other dancy-vids; I made mostly unrelated dance moves appear contiguous using zooming, blurring, and just plain complex planning. I looked for frames where their hands or positions were similar to those of another, different pattern, and I spliced them together to change the sequences. Occasionally I exported individual frames and, in Photoshop, altered hand positions and such, or used a few frames from yet another part to string three sets together in less than a second. I sometimes shifted objects in the background just to make the transitions smoother, moving that sun on the wall and stuff during zooms, gradually over so it lined up with the sun in the new next-frames. Speed changes, reverses, loops, reverses blending back into forwards…

    Nobody will ever know how much work I put into this…

    Then getting Haruhi and her special-guest “star” on the same stage… I did that TWICE. First time, that clip had been larger than my project resolution, and Premiere ignored my “scale clips to project dimensions” option. So yeah, I thought it looked kind of off, then found out it was zoomed in too far, and I had to start all over.

    After I built Haruhi her side of the stage and made a spotlight layer and matched it to the other one (until the diversion), I thought I could just use chroma or alpha transparency to place Haruhi there, but with her white uniform, that was impossible, so all that was frame-by-frame exports, lots of MS Paint and Photoshop work to prepare the layers, lining them up properly (only using certain frames cuz I changed the speed of Konata’s dancing to match the beat), then exporting the whole sequence and manually fixing some outlines and shadows…

    Nobody will ever know how much work I put into this…

    And then the lip-syncing. Oooooh, the lip-syncing. There are so many legendary, awesome vids out there that pull it off in a sort of fake-looking manner, over mostly static characters, or try to cover up lip-flap and don’t quite pull it off right, and their vids are STILL friggin’ awesome. But, I didn’t want to settle for anything less than highly-passable and natural. I wanted it all to look clean and authentic, at very least while in motion. Well, I actually wanted it to look perfect even with it paused, but I did have to draw the line somewhere with how long it was taking me…

    Shutting up characters is often harder than making their lips match the speech. I so often had to quiet the background girls so Haruhi could do her thing, and then match Haruhi’s lips to the stuff going on. Or make them all sing when the background voices come in. Or shut all three of them up. Or create and add open mouths that were never open. All this on characters that are constantly moving, never standing still until the end. Every single frame had to be adjusted and lined up perfectly. One frame being just a tad off would make it look totally wrong. If you don’t put them perfect, their mouths seem to float around their face, and allowing Premiere to tween looks terrible. I even had to correct a few mouth frames where the animators themselves had a hard time keeping them steady.

    And I didn’t know about the lasso tool’s functionality and usefulness until I was five mouth edits from being finished… so all this time I’d been doing it with much harder methods. @_@ But, it looked fine so there was no need to go back, though at least now I know better for future projects…

    …Did I mention that nobody will ever know just how much work I put into this? ^_^()


    Oh and as far as improvising, I got 1:42 of footage to fit 2:53 of video (not counting credits) without it being all that repetitive. At least if there are repeated moves, they are quickly connected with totally different ones each time. Much more varied than standard dances nowadays at least. This is ALL I used:

    -The Hare-Hare Yukai dance
    -Six quick flashes from the Haruhi closing credits
    -Scenes from the Lucky Star opening
    -About 4 seconds of the Haruhi Cosplay scene from Lucky Star (episode 16)

    And then more of that scene for the closing credits, mostly unchanged (except for fixing Konata’s lips for a few frames, zooming in on one part so the audience could be seen better, and lip-flap being covered up as it fades out). (And the music in the closing credits is the original version of the song, before she did the official radio version featured in the vid)

    Yes, there were a couple parts where they go “whoo hoo” but don’t lip it… but this is a music-video style music-video, they don’t always lip in those either… and I think I’ve already put myself through enough anyway, yes? ^_^()

    Anyway, I really, REALLY hope you enjoy this! Cuz I worked darn hard to ensure that you get a good simple song-n-dance video that’s as natural as possible. HAVE FUN DAMN IT!!! :D


    -Adobe Premiere 7.0
    -Adobe Photoshop CS2
    -MS Paint
    -AviSynth 2.5 (tweaked Saturation, de-noised with Deen filter)

Opinions (7)

  • Orig
  • Visual
  • Sound
  • Synch
  • Lip
  • Effects
  • Effort
  • Re-View
  • Overall
  • 8.60
  • 9.60
  • 9.40
  • 9.80
  • 9.80
  • 7.40
  • 9.60
  • 8.60
  • 9.00