video - workflow

One has to back-out to a philosophical level to think about video work in common terms, so there is more of that in this post than my typical posts. There's recipes and there's ingredients, and there's logs of the work. Some of what might need to happen is to also have a database philosophy, since each edit could be a database entry. I should reference a database post.

Video is problematic. I outline further down the workflow elements which seem most consistent across projects. It's hackneyed though because I "meta-up and drill down" almost uniquely to each clip or overall project. So, I thought of listing similar project challenges rather than a list of postive steps. But even challenges shape-shift somewhat in each project. Difficult to patternize. Aieee.

Over the years, the one universal so far is -- the more vast and standardized one's storage and retrieval system is, the better. Storage for both video (ingredients) and production logs (recipes). An arguable second bedrock may be video vocabulary. Video thought seems unique to each person's visual imagination; a word, in place of an explanation, can be a small efficiency gain when dealing with others.

creative process

Sample a few other creative pursuits; writing, music, cooking. Of these, writing is the simplest, since we only need creativity and language. Creativity varies by time of day or time available, but compare this to video. Variation applies universally. Creative opportunties vary by time, inclination, premise, budget, location, expertise, chance, equipment, software etc. Also across functions. One video might have various percentages of creativty in its shooting, its editing, its vision of the outcome, and so on. And these variances might also differ in emphasis for a hobbiest versus a paid videographer. It's difficult to proceduralize video universally.

For me, most videos are sifting and reflection processes. The end result is a kind of report of an investigation upon the video, stills, text, sounds, and effects at hand. Rarely is the outcome (even in a commercial video) entirely known in advance, no matter how clear the sponsor's (implicit or explicit) instructions or desires.

project ingredients

Mostly there are 1) conceptual elements and 2) technical elements. The latter support the former. By analogy video creation is perhaps more like music or cooking than writing. The tech of writing is fairly simple; sometimes only pen and paper or simple software are necessary. Meanwhile the tech of video can become so complex that it prevents the video entirely, or mangles the intended outcome. Many many video ideas are thwarted by production limitations (software, skills, etc).

the shooting

We don't always know why we shoot. One pole of the continuum is, "shoot first and sift later". The other pole is "gather shots to support X concept". . Usually it's a mixture: a vague concept ("I'll bring my camera along on this trip") followed by a sifting process. Sometimes though it's a very clear concept ("I'm making a soda ad, and need to make several shots for my/producer's idea of the advert"). The latter might even turn the former on its head: "I need to make this trip with these cameras to get these shots for the project".

Ideally we'd prefer infinite latitude, time, and/or money to change our shooting on the fly. Creative whims develop in situ with the events and cameras. We know some things in advance of editing. We know the type of camera(s) we used. We know why we brought them -- even if multiple reasons (financial limits, waterproof, low-light, etc). And we often remember some of our intention(s) when shooting X thing, though often times these will be layered: "bored", "capture X event", "capture spirit of X event", "document X event beauty highlights", "wow". So the shooting varies from desirably undefined to a conscious attempt to capture the blocks to complete a pre-concieved schema.

Sometimes the theme is a trip, but only assembled afterwards. Perhaps I was able to capture several meaningful sunrises or I was struck by a person's expression and tried to capture it. Other times, when sifting through footage later, some surpise nuggets turn into a major or supporting theme. This can be organized into the final product (the highest meta level) and several versions with different emphasis or support might be necessary.

scratch sheet documenting

Whether a person organizes their work in advance or lets it organically develop while sifting through clips, some scratch sheet, sometimes many sheets, will be sketched with time, effects, and transitions. Like writing, some schemes are eliminated, some are modified in progress, some can be done nearly all in one's head, and very few can be rigidly determined in advance. To me, each video is more like a sifting and investigation process. the very few can be done on the fly with experience. plan timing and nature of cuts, transitions, and other effects. In each clip, they will need to know the timing of what's taking place, either by frame number, time, keyframe, or some combination.

1. first cut at organizing result:
2. normalize clips and record results: evaluate raw clips (eg. ffmpeg -i, personal knowledge). Relevant features vary by project, but typically at least timebase and fps must match from one clip to the next. Keyframe times or frame number may also be relevant. In the maximum case, enter relevant details about all clips into a spreadsheet or other form, since renders will need to be optimized to a minimum number and we'll want to combine events. Alternatively to step 1, an editor can just pick up the relevant info on each clip as they go along.

storage

pantry analogy. a chef not only produces food, but needs to save his recipes and needs to buy ingredients. The problem with video is it has an eternal shelf life and might have nuggets which want to be drawn from later. And as for the recipes, they need to be kept indefinitely and sometimes refined.

Where to store gigs and gigs of images and video, and the various versions of rendered results? Where to store the meta documents of each edit? How to even find clips again, or the method that went with a