I was asked by a friend to create a photo montage for her wedding. I agreed and talked with her about her picture selections and the style she wanted for the piece. We chatted about messaging and mood. Armed with a good sense of what I'd do with the video, I got ready to edit.
Then the groom-to-be said he'd like to try putting the piece together himself. He's been learning iMovie and would like the extra practice. Plus, with the latest software, he has all sorts of font styles, image movement options, pacing controls and automation tools whereby he can toss the thing together in a couple of hours... tops.
Me? I'm in the middle of learning an inexpensive but fairly powerful Non-Linear Editing program which I was going to use for this project. My program contains no fancy bells and whistles for automating photo montages. I'd be doing absolutely everything by hand, one click at a time... while wrestling a new NLE.
Honestly, iMovie is probably the better choice. I predict his version will be excellent and take him very little time.
This got me thinking about the future of photo montage editing.
Prediction 1: Automated programs will do a better job of cutting a photo montage if the machine is given a few seconds and the human an hour or so.
What needs to happen to make this a reality?
First, computers will need to use their face recognition ability to frame photos and--where applicable--apply a subtle zoom in. If I say I want a photo montage for my friend's wedding, the computer should have me identify which friends, then use the rule of thirds for framing and zoom in on the appropriate people in each photo. This should be doable within the year.
Second, the computer will need to find the beat of the song and trim the duration of each photo to fit the music. Considering graphical visualizers can already change style on the beat, there is no reason a computer couldn't do this right now.
Third, developers will need some simple tweaking tools which enable a human to show a computer where things need to change. A "slow down" button, for instance, may be needed if the computer decides the memorial video should move like an awards banquet montage. Of course, for starters, a few standard presets would give you most of the parameters you'd need.
Prediction 2: Computers will eventually be able to assemble photos, pick music, and put the whole thing together with almost no human interaction.
A friend posts: It's my birthday! You, inspired, click a "make montage" button for this announcement. The computer finds all the photos for this individual, selects the music he or she likes, and spits out a montage to celebrate. Cool! But not much different from above, assuming this is connected to a site that collects personal data about a person's preferences and tastes.
Even better: A distant relative--who distrusts the internet entirely--gets a promotion. You don't really know this person, but your mom thinks it would be nice to send them a celebratory video. You find a couple pictures in your family photo album, include something about the promotion, and feed that into your Photo Montage Tool. Within minutes it's uploaded a video with hundreds of pictures, fancy but obscure graphical elements, and a music style you didn't even know existed. Your relative contacts your mom praising your video as "spot on."
How'd the computer do it?
First, the computer found other pictures of your relative using something akin to Google's experimental "more like this" image search tool.
Second, the computer compared the look and feel of your relative to a "what these kinds of people like" database based on information collected from things like Facebook's Flair and Pinterest. This will also include music selections.
After that, it's cake.
A part of me mourns the loss of "creative expression" and the need for me to help out with these kinds of projects. On the other hand, the only thing that used to separate me from this engaged couple was the know-how to toss a photo montage together with my expensive program. I could crop and move and sync, and they couldn't. Today, most phones can do a pretty good job without any human input. My usefulness is quickly fading away.
This, of course, raises questions about the difference between art and assembly. How much of a production is governed by rules, and how much needs that spark of creativity? What is creativity anyway? And when well-proven rules can be programmed into a machine, to what should humans next apply their artistry and time?
Your Media Production Mentor