Some vids I learned from
May. 23rd, 2015 03:32 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
brainwaneWith some arbitrary categorization for ease of skimming, here are some vids I learned from:
"Around the Bend" by danegen demonstrated that mixing in still nonfiction photos could totally work in a massively multisource vid. From this I learned that it's often good to use a consistent speed and method of zoom on the still photos -- start with the person + environment/important object, and zoom in on the face.
"Still Alive" by counteragent often places a bit of video or web-based text into or on a bit of moving video source; I particularly noted the animation of grid of dresses, squares replaced arbitrarily by still photos. I like how counteragent layers one pan through a photo over another pan, zoomed out more and faster, of the same photo. And I was reassured to know that it's okay to pan or zoom through stills from fandom very quickly, i.e., 1/8 of a second each.
chaila's "Watershed" is practically a clinic in how to use comics stills. Stills are basically always in motion via a pan or a zoom, and you see no dialogue balloons or captions, which increases the visual impact of the image. (One sort of exception: an animated "bam!" followed by a comics still, zooming outwards, of Wonder Woman surrounded by her power aura or explosion.) Sometimes chaila begins showing you a comics still by focusing on a weapon, then zooming out and panning to also show Wonder Woman holding it. To show a small person in an epic or overwhelming environment, or to show stillness, she zooms out or quick-cuts through multiple comics stills that themselves "zoom out." To connect one event causally to its context, cause, or in-universe viewers, chaila briefly uses a translucent-going-to-nearly-opaque overlay, then fades it out (sometimes of a comics still over a clip of video). To show the viewer a character's perspective, chaila zooms from her face towards what she's looking at. To magnify the impact, close-crop and leave some of the screen black, e.g., showing only a pair of eyes while panning in the direction she's looking. And -- this is not stills-specific -- chaila sometimes starts an important motion or sequence on one strong beat, fills in the interim measure with other thematically related visuals, then finishes it on the next strong beat.
The way chaila uses different regions of the screen deserves special attention. Accompanying the lyric "while you were setting your woods afire," the top and bottom halves of the screen fade independently (and both are panning left to right), with a few pixels of black between them. First the screen shows up with the top already filled in. The bottom starts to fade in; as it finishes fading in, the top begins to fade away. We see the bottom by itself, then cut to the next shot. Similarly, a few times, chaila uses vertical pillars. Example: the frame is divided in half, the left side is already filled with comic, and then three pillars, following left-middle-right, fade in and out while panning downwards. In contrast, the next visual is a comics still panning upwards.
"Piece of Me" by obsessive24 demonstrated to me that it can work fine to mix very high-resolution source in with low-res source.
"coin operated boy" by counteragent includes new footage shot just for the vid, featuring a hired actress and a set. I thought, relieved, "What I am doing is nearly staid by comparison!"
This isn't a vid, but Rob Cantor's "Shia LaBeouf" demonstrates utter commitment to the spectacle of the joke; it escalates with every verse, and nice use of stereo for the end. For more in this vein that inspires me to strongly commit to over-the-top implementations of silly ideas, also see The Colbert Report generally, and the performance of Rebecca Black's "Friday" from Jimmy Fallon's TV show.
"We Didn't Start the Fire" reassured me that a scores-of-sources vid is doable, and spreadsheets help. And that it is okay to have super fast references fly past and not everyone will get everything.
chaila's "Parable" constructed vid trailer is like nothing I'd seen before. It includes lots of crossfading between clips or layering/montaging of images atop each other such that I more fully felt the unity of the world she made. I believe chaila also color-processed clips so they matched, which is cool.
"Broken Remnants" made me wish I had a video editing program that made visually intensive cropping and graphics placement easy. I at first felt put off by the extreme pillar usage and compositing, but then it grew on me. I liked how the vid faded into long stretches of full-frame black as a rest for the eye during slow instrumental bits. And at 1:32, accompanying the urgent strings, I love the gradual reveal of a fiery ship via pillars; we smash into each reveal but fade to black after each new reveal. The measure has 8 string notes, and the vidder uses that pillar reveal approach for 6 of the 8 string notes, then switches to a clip of a battle, and accompanies the last two string notes with photon torpedo hits. I also appreciated, near the end, the gradual enlargement of the the visual field to give a sense of epic expansiveness.
"One Night Fandoms" by eruthros and thingswithwings uses, in its intro credit sequence, successive fades into title, source, and the list of vidders. The vid cuts to black momentarily between sequences to reset (9 frames of black). I liked the use of "Roger Rabbit" to transition from animation to live action. And I appreciated seeing examples of very fast cutting or scrolling that works well. The "Baby, baby, baby, baby" lead-in to chorus has very hard cut-on-the-beat quick shots of couples, almost too fast to process. And in the end credits, the sources scroll by very fast, and can't really be read in a party context. And that's okay.
"Mr. Blue Sky" by odessie puts its opening credit over a relevant still for about 4 seconds, fading in and out. It then fades into the start of vid content (video) & starts audio at 5 seconds. A pan in the visuals accompanies the first measure of music, which serves sort of as orientation, and then we see very hard and consistent cutting-to-the-beats, one-two, one-two, as we switch back and forth between two different live-action Superman versions, and see the similarities between them. Then the vid expands, with somewhat looser cutting, to different sources, introducing them into our world. By 00:42 we've basically been introduced to all the sources. It was good to see that it's okay to use Superman flying up the shaft for 1 sequence and then use him taking off in the shaft in another sequence many seconds later; we don't need to strictly follow the chronology of that scene from the source movie. And in the end, to intensify the sense of wonder and fantasy during dreamy music and flight sequences, the vidder uses more fades rather than regular cuts from shot to shot. Another lesson: To prevent malfeasors from clipping out the credits and falsely claiming authorship of a vid, put a tiny bug of your initials in the lower right corner: white, black border, translucent background.
gianduja kiss's "A Different Kind of Love Song" uses a lot of interesting transitions between clips, such as a curtain-opening effect, a swipe cut, a door-opening effect, and sliding one clip out from under another clip, all of which work well thematically and with musical cues -- for instance, the swipe cut takes us into an anthemic chorus. The intro and the ending sequences use quick fades to signal transitions between different sources (in contrast to the regular cut between shots). I was heartened to see that it can be fine to speed up the middle of a pan. I also liked how the vidder matched the right-left (180-degree) orientation of different 2-person shots to analogize them. And gianduja kiss's quick-cutting over a stuttery bit of the music was so cool I watched it frame by frame. She uses a ton of 3-frame or 4-frame super quick shots -- stills almost -- cut with a single frame of full black screens in between, then rests the viewer's eye with 14 frames of full black, then fades into another shot over ~10 frames.
Both "A Different Kind of Love Song" and gianduja kiss's "Hourglass" use a credits technique I like: if there's a funny stinger at the end of vid, with actual relevant source video and audio, then reprise a bit of the chorus over the scrolling credits after that.
"Stay Awake" by Laura Shapiro starts by displaying its title credit over a relevant bit of video, instead of just on a field of black. The vidder also uses translucent overlays of relevant backstory in an available half or 1/3 of the screen while a character's face is in foreground, to give context, or overlays it on someone's body. In both the start and end credits, the words dissolve apart and the letters lose focus, which continues the vid's themes (the loss of bodily autonomy, creepy dreams and the scary things that happen to women if they fall asleep).
gwynethr's "Ready Steady Go" uses a stuttery transition between two clips, where the vidder intersperses frames or pairs of frames from each for a second or two.
Laura Shapiro's "Only a Lad", which argues similarities among three characters from different fandoms, shows us intriguing scenes featuring its three key characters at the start of the vid, then bookends or calls back to those scenes later in the vid. The vidder also uses two clips in which Character A and Character B make similar hand motions around their hair to connect them visually. The intro shows us a sort of palindrome of these three characters: A, B, C, C, B, A.
"Hey Ho" by thuviaptarth uses just the right song and just the right clips to show you a story that was already in the source text all along. It inspired me in that "I want to make something that good" way.
Other
And of course I learned from a ton of other sources, like a ton of meta about vidding, and Tony Zhou's "Every Frame a Painting" video series, and the scores of vids I've watched over the years. Anyway, hope someone finds this of interest.
