#i dunno how many people follow this blog when they're looking for fandom related things

[psychosinmatic]

Hello everyone! Just wanted to say if you followed me just to see fandom stuff please go to

@komoco

[vimesbootstheory]

How I Vid

(I wrote this about a month ago when I was extremely tired and a bit tipsy so I dunno how easy it is to follow (syntactically and also just in terms of me making too many assumptions about shared knowledge), but I think it’s worth putting on my blog for posterity.)

IDK if anyone is actually going to read this, I guess if nothing else it's a fun little writing exercise for me. But before I get underway I want to point to the title very insistently. This is how I vid. This is not how TO vid. My way is not the best way, and above all is D E F I N I T E L Y not the most time-efficient way to do it. I mean it. I've never finished a vid in less than three weeks, and 3 wks is really quick for me. This year I've made three vids and that's a 200% increase on the previous record. I'm slow as molasses in January. This also probably isn't how I will always vid, since I tend to tweak the process at least a little every time I make something. Anyway, let's get into it.

Tools:

- Sony Vegas Pro (14.0, but other versions work just as well)

- Avidemux

- Audacity

- VLC Media Player

STEP 0: Choosing a vid source/subject

(BTW this step numbering system is going to be 100% arbitrary. It only roughly approximates the order in which I do things, generally these steps overlap chronologically at least a little.)

I don't really consider this a step since if I didn't already have a vid source in mind, I wouldn't have the goal of making a vid in the first place. The order of operations never goes, "Hmm, I want to make a vid. What should I make a vid of? Oh this thing I like", it's more like "FUCK I love this thing. How should I express that? Well, one of the only things I'm half-way good at, that'll do, I'll make a vid".

The genres/formats I've done are character tributes, show-general vids and a couple of ship vids. The former two are my favourite -- character vids because they're probably easiest in terms of prep and funnelling passion into a narrow subject, and they also tend to be the most successful (in terms of views & comments & likes), and show vids because they're the most accessible for a general audience and are great for working out that "ugh I love this whole show and need people to understand this" energy.

But yeah, this generally isn't a question or process, I usually just know right off the bat what I want the source and/or subject (if the subject is narrower than the source) to be.

STEP 1: Acquire vid source

Usually I get the source either from ripping from DVDs (I use VLC; Media -> Convert/Save -> Open the video file on the DVD drive -> Video H.264 + MP3)  or downloading torrents. Gotta be logoless, logos are distracting. If I'm going to be including any diegetic sound from the source material in my vid, the video files have to have 6-channel sound, which is usually in the torrent information. I'm not going to describe my process for extracting dialogue from background music because it's exhausting and tbh I'm not a huge diegetic sound fan in the first place.

STEP 2: Choosing a vid song

In some ways, this step has the most potential for frustration. It sounds weird, but my head has gone to some weird, over-stimulated/under-stimulated places when I've been listening to song after song, looking for just the right convergence of factors and getting nothing back.

I get songs from a lot of different sources. my favourite starting point is 8tracks fanmixes. the song I'm using for my current project is straight out of a fanmix for the source material, which feels a bit bold -- usually, if I find inspiration in a song from a fanmix, I'll look into the artist's other work, or look for similar tracks on sites like spotalike, or look into similar artists on spotify. or if I use a song straight from a playlist, it's usually from a more general theme (like "rich kids", for my Bruce Wayne vid), or a fandom with similar concepts or characters (spent a lot of time listening to fanmixes relating to goodneighbor from fallout 4 trying to find a Z Nation s5 vidsong... no dice).

Factors I look for:

- appropriate tone/mood -- this might be the most important one. in particular it's important to avoid unnecessary melodrama, since I think it's very easy for the combination of video and music to create a melodramatic mood.

- a beat/rhythm that jumps out at me as something I can make clips dance to; bonus if it has unique, interesting sounds that I can match to specific shots or movements in my head

- lyrical accuracy -- of course it doesn't have to be a perfect match, but if there are lines that jump out as being completely divorced from the mood or tone or idea I want to convey about the source material, the song's not going to work

- the right level of exposure -- this is going to sound like such a hipster thing, but I can't bring myself to use a song that has too much radio play, and in fact I prefer not to use singles at all. this isn't something I apply when I watch vids, a ton of my favourite vids are to mega-popular, overplayed, over-vidded songs and I don't give a shit. but every once in a while I'll run into a vid where I'm like "... I would be in love with this video if I didn't already associate this song with another vid that used it better/before" and I can't risk my vid falling through that rare crack

STEP 2.5: Editing the song as needed

Didn't used to do this, but by now I tend to do at least a bit of trimming. I use Audacity. It's free and easy to use. I'll go through the song and figure out parts of it that I think will be a drag to vid -- if a segment repeats too many times, or there's a weird bridge that doesn't match the mood of the rest of the song, or there's a verse that doesn't match the source material lyrically as well as the rest of the song does. I note the time stamps for appropriate trimming points, eyeball the wave form, select and drag over the part I want to remove, and delete it. Sometimes there's a bit of trial and error here, if it's even a little obvious that something has been removed, but usually there isn't much trial & error, I must have a knack for eyeballing. if I can't make it sound to a novel ear (ie someone who's never heard the song before) as if nothing has been removed, then I just won't remove anything. I think it's incredibly jarring if you can tell a song has been cut to ribbons. I don't think I've ever ended a song early, and I don't anticipate doing so in the future. I always cut from the middle. Once I'm satisfied, I export the track as a .WAV to the folder where I'm keeping all the resources for that vid.

STEP 3: Clipping!

The first of the two most time-consuming steps of the vidding process, where the second is just... the actual vidding part lol. Listen, I do this step in possibly the most tedious and time-consuming way you could possibly do this. Just as an example, let's say I decide to do a show-general vid for one season of a show. I will load the first episode into Avidemux. Video Output set to Mpeg4 AVC x264, Format to MP4 Muxer. Unless cirumstances are extraordinary, I disable the audio track (Audio -> Select Track -> uncheck anything that's checked). I put my vidsong on a single-track loop on my media player. And then I watch the episode without sound, with the song playing on repeat in the background. Whenever I see a clip with eye-catching movement that I think will decently represent the mood or tone or idea I'm trying to sell (having the song on in the background helps me identify what the exact mood/tone is on a constant basis), I go to the keyframe immediately preceding the interesting bit (aka press the down arrow key), set start marker, then press up until I've passed the interesting bit, set end marker, and save. The clips are typically quite short: 2 sec would be on the short side, 30 sec would be relatively long. I save it with a filename that reference its episode and chronology in that episode, eg g2e05c15.mp4 will reference the 15th clip saved from gotham season 2 episode 5. I could probably be more or less precise about that, but I'm set in my ways. I like to set arbitrary rules for what I'm allowed to skip through quickly, eg scenes during which the opening credits are playing over the bottom of the screen I'll just fastforward, or scenes where the characters are talking to guest stars, or scenes with characters I don't care about, so I can try to clip through an episode in less than an hour. And yeah, I just keep doing that with each episode until I've reached the end, and then I have at least a couple hundred clips as a kind of starting palette to choose from. It's super boring and takes ages.

STEP 4: Vegas Palette Preview

Didn't really know what to call this step. It's a little step and maybe a bit useless in the long run but it's also very low stress and gives me a breather and makes me feel ready before I jump into the real work. I start a Vegas project and import all the clips, plus the song. Then I slap all the clips -- sometimes as much as 500+ of the things -- down on the timeline, and on the audio track, I place the vidsong, over and over end-to-end until it covers the duration of all the video clips. The last couple of projects, the resulting project has ranged from 45min to 1h45. And then I just watch it, multiple times if it's on the short side. It gives me a good idea of what clips are truly useful, so I'll just delete dud clips if they make themselves known at this point. Usually I'll get at least of couple of concrete moments of inspiration, of how certain clips interact with the music. But mostly it's useful for distilling my concept of the vid in my head, and really marrying together the audio and video agendas in my head. Also familiarizing myself with the whole of the clip options I have at my fingertips.

STEP 5: First Draft

Here's where it opens up, and where it gets significantly more creatively daunting. I always start at the beginning and go through chronologically to the end for the first draft. Maybe that's boring and stifling but it's good for my ability to feel that progress is being made. I can just zoom out and hey, there's my progress bar, that's how far through the draft I am, yay.

Titles and credits don't enter into the first draft. I often have an idea of what clips will have the titles overlain over them, e.g. the trundling record in "Still Feel Alive", the shot of Bruce underwater in "All That Money", but it doesn't get committed to timeline until later.

It's hard to describe my approach to placing clips moment-to-moment, because there's so much instinct involved in this part. I tend to listen to a very, very short segment of the song -- a couple of words of the lyrics, a couple beats -- several times until I have a very good idea of the movement that should match that portion. It's not like I put that into words, but I guess it might be something like "ok, I need a hard impact followed by constant motion from left to right, conveying a perky sort of mood, conveying X idea from the lyrics". so I'll start looking for clips that meet that description. I can usually just look at the clip thumbnails without having to rewatch them; another benefit of step 4 I suppose, just knowing them all by heart by this point.

My schtick is beat-matching. I do a lot of it, and I think it's fair to say I do it pretty well. I don't know if it's worthwhile to try to describe HOW I do it -- possibly it's obvious but just so time-consuming that people just opt not to. Anyway, just in case: the beat is often visually distinct in the waveform on the audio track, you can just, yknow. See it. So the idea is to align that peak in the waveform with the moment of impact in the video clip. I grab the front edge of the clip and pull it to the right until, in the preview window, I can see the moment of impact (eg a fist in initial contact with a face, the moment that a movement reaches its most extreme point before it starts to boomerang. Then it's just a matter of making the front edge of the clip line up with peak in the waveform, and then dragging the edge back to the left, making sure that the shot doesn't change. there's a lot of trial and error in this. like, a lot. I can be positive that it's lined up perfectly, but when I play it back, it doesn't quite work with how audio vs video are processed in the brain. I don't think of it in such concrete terms while working, but I suppose we would process visual information faster than auditory information, sooo that means the waveform peak should go before the visual impact moment? Honestly you can try to force science into this but then watching it back I might still be like "... yeah, no, still needs to budge up to the left".

If the beat ISN'T visually distinct in the wave form... man, it's just a lot of trial and error.

It doesn't do to just watch back certain limited portions of the video and expect the whole video to be cohesive, though. after about every ten seconds of placed clips, I'll render what I have so far and watch it back at least a dozen times. often a portion that seemed perfect just watching 1-2 seconds at a time will not be visually cohesive with the rest of the video at all, will feel like a hiccup in the established rhythm that proceeds it, and will have to be scrapped.

Aaand yeah I just keep doing that until the song is over. Every vid has its own little creative moments that will require a specific approach, and I can't very well go into all of those. I could dip into velocity envelopes, I suppose. And transitions, which is a very quick stop. I do hard cuts almost exclusively. Rarely, I'll work in a fade to black followed by fade up from black, if the tempo is appropriate, ie slower than the rest of the song. I've done a couple of hard cuts to black. But yeah, I'm not really a fan of transitions, both as someone who watches vids and someone who makes them. In particular, I always read one clip fading into another as like... the passage of time? you know, like if there was a montage in a film that was supposed to indicate that a girl and her beau passed several pleasant afternoons in the courtyard at her summer home, and the music is twinkly and slow and each clip fades into the next so we know time is passing. Anyway. Hard cuts. Hard cuts unless for some reason a hard cut really doesn't feel right, and even then, something pretty dang close to a hard cut.

Velocity envelopes are worth mentioning, and in a similar way, I don't tend to use these in a way where they're visually evident as added effects. I use them to force a clip to match the music -- to speed up part of a clip, or slow it down. Like in "Still Feel Alive", the bit with Doc jamming to the music, I'm particularly proud of this because in the show, the beat he's bopping to is completely different than the one in the song. it's not just a matter of a constant slow down or speed up, I split that clip every time his chin reached its most extreme point, and used Ctrl + dragging the front edge (I usually think of this as the Accordion effect because of how the zig-zag line scrunches and stretches, and it accomplishes a similar effect to velocity envelopes) to fit the space between beats.

During my last two projects, I tried to loosen up when it came to clip choice for a given moment. as long as the placement was accurate, if it wasn't the PERFECT clip for that moment, that's fine. fill the empty space and move on. substituting clips with higher calibre material is what the second draft is for.

STEP 5.5: Colouring

This step can happen anywhere from the Palette Preview to the Final Draft, really. In my current project, I've already got a colouring profile picked out and I haven't placed a single clip, whereas in other vids, I hadn't even attempted to colour anything until after I'd put down the opening titles and credits.

My attitude towards colouring also goes all over the place and I don't consider myself to be very good at it, second only in miserableness to my uselessness with using diegetic sound. My current attitude is to keep it pretty no-frills: a Convolution Kernel with the Sharpen preset, a Saturation Adjust (adjusted up), and a Curves layer. Don't ask me to explain the shape of my Curves layer. It's a wiggly line.

I don't think you can really go wrong, keeping it classic. You want to at least have some sharpening, though. Unsharpened footage looks sloppy to me, like you didn't care enough to at least polish it up a bit.

STEP 6: Neglected Faves

At this point, before I dig into the second draft, I like to go through my palette of clips again, super thoroughly, and find parts of clips that have not been included in the first draft, but that I really really like and would be crushed if they didn't make it into the final vid. I believe for the last project, I made a subfolder out of these neflected faves. This makes a narrower, more concentrated, easier-to-sift-through palette for the second draft.

STEP 7: Second Draft

In vidding, as in writing, second drafts are easier. Much less creatively daunting. First thing, I go through the vid and mark spots where the vid could be better. Maybe the clip feels a bit awkwardly-placed, maybe my eye is catching on something, maybe the clip is dull or doesn't represent the subject the way I want it represented. Vegas has a handy function here where (at least on Windows) you can press 'M' to put a little flag on the timeline. I'm given to understand that some people use these to mark where the beats in the song are. I just look at the audio waveform for that, which is much more accurate than my reaction time, so instead I use those little flags to mark the beginning and end of problem areas. Then I essentially just re-vid those parts, either with the clips already placed re-positioned, or the narrowed palette of neglected faves, or some combination of the two. If a neglected fave doesn't fit into any problem areas, I might also look at non-problem areas, though that happens less often. As problems areas get solved, I delete the flags. I keep going until all the flags are gone.

STEP 8: Optional Drafts

Sometimes, I have to do steps 6 and/or 7 multiple times through until I really have no objections to anything about the vid itself or what's been left on the cutting room floor.

STEP 9: Titles & Credits

This is pretty straightforward. These days the opening titles of my vids just have the title, and then in the closing credits I list the song artist, the song title, the source material (+ network), and my youtube channel username 'absinthespoons'. Sometimes I go through my fonts tag for a while, or resource blogs that I've bookmarked, or dafonts if I'm desperate, until I find something that speaks to me. I'm not all that picky. I like pretty no-frills fonts, nothing too garish.

In terms of how the titles are presented, it varies so much by vid that I'm not sure trying to describe the process would be useful because there isn't a consistent process. In "Still Feel Alive" and "All That Money", the title ran over a pre-song clip with an alternate audio source. In "Grit Spitter" and "Making It Mine", words appeared on the beat. I will say, as great as "All That Money" is, I hate the closing credits, they're so lazy. The closing credits for "Grit Spitter" and "Morningstar Coffee" are much better.

And there you have it! That's my process. Now I gotta sleep.

STEP 10: Betaing

Ideally, I like to have someone who isn't familiar with the source material look at it, as well as someone who is, so that I know the video is accessible to both audience types. Usually the kind of feedback I look for is, are there any moments where your eye snags on a moment that doesn't look quite right; and also, does this vid convey a perspective or message that I didn't intend. Like, for my current project, I am very keen on the vid not being interpreted as shippy but instead as a celebration of friendship. So beta feedback helps me feel better about that.