Tag: sampling

Video Lectures: From Concept to Reality

In my first entry, I discussed several of my sources of inspiration as a way of describing how I initially approached the project of making lecture videos. I’m going to leave most of the technical details – what applications I used, how I achieved particular effects – for a third entry. This entry, then, is an account of the awkward transition from abstract vision to deliverable course. Some of the awkwardness was the result of my working style. I don’t tend to systematically evaluate the options before moving forward; I dive in, try things, and, after hitting quite a few dead-ends, get a sense of what works and what doesn’t.

One of the first decisions involved determining what proportion of the visuals would be created “from scratch” and what proportion would be “borrowed.” I’d start with a voice-over script that was my own, certainly, but I didn’t necessarily have to write, draw, photograph, or film every visual element I’d use. I already integrated a handful of film clips and photos in my typical classroom lecture, but video lectures presented the opportunity to do this on a much grander scale. What if I just made the whole thing out of clips? Wouldn’t this be better than bullet points or some digital approximation of a whiteboard?

The Pedagogical Potential of Sampling

Here, I feel compelled to cite another point of inspiration: the sample-based hip-hop pioneered in the 1980’s. Listening to this music taught me that a skilled artist could re-purpose clips of existing creative works and make something that felt just as compelling and original as any new work. Sampled beats and horn blasts carried with them the expensive studios and skilled session musicians of the originals, yielding a sound that had a professional sheen to it which other do-it-yourself musical movements like punk and “indie rock” simply didn’t possess. I found echoes of the methods and aesthetics of sample-based hip-hop in online culture of the past decade. Nearly every bit of viral ephemera circulating through social media seems to include an existing photo or phrase re-purposed for ironic intent.

Once you see the creative possibilities of sampling, everything around you (particularly images and recorded bits of sound and video) becomes a potential component in whatever you’re trying to create. This is, mostly, how I approached the process of creating the visuals for my lecture videos. I say “mostly” because I created some visual elements (maybe 25% of those used in my videos) from scratch, typically by creating and animating Powerpoint slides with figures and text and then recording these animations using video editing software. But the vast majority of the visual elements in my videos were appropriated.

Sometimes, I had particular clips or photos I wanted to use, either because I explicitly referenced them in the script or because they were good illustrations of the general principle I was explaining in the script. In the case of film clips, I could just hunt down DVDs of the film and capture the relevant part of the DVD using video capture or editing software, but I quickly realized that this approach would be too slow and onerous. The amount of labor involved in this process would mean that it would take months to create a single clip-heavy video lecture (the math on this changes if you have assistants, as a documentarian might, but this is unlikely to be the case for many teachers). And so I turned to the internet.

Knowing What You’re Missing

Finding these clips or photos was as straight-forward as any online search: I typed various iterations into Google Images or YouTube until I found what I was looking for or until I became too frustrated to keep trying different iterations. This process continually reminded me of the limitations of the seemingly-limitless troves of content freely available on Google Images and YouTube. There were also times when I didn’t have a particular image or clip in mind; I needed an example of a montage from the silent film era, or an image of a director, any director, working with actors. In these instances, the limitations of freely available content weren’t as obvious. I could find exemplars easily enough, but they tended to be the same exemplars, over and over, reflecting a certain limited set of sensibilities: American, white, young, and straight. This worked against my intention to feature images of a diverse array of people in my videos.

What was true of YouTube was also true of Google Images. It’s quite easy to find images and video of young, attractive, white people engaged in a wide variety of behaviors, but much harder to find images and video of anyone else engaging in anything other than culturally stereotypical behavior. There are exceptions, of course, but they take time to find, and so I had to build in the extra time to dig deeper than the first layer of YouTube and Google Images.

Of course, the sampling approach to making videos raises all sorts of ethical issues. Here’s where I landed on those issues: as long as you provide proper attribution and don’t create a substitutable version of the original, you’re okay. There is a way to create a sample-heavy work that motivates the listener/viewer to track down the entire original work, thereby increasing the visibility of the original. This happened with me while watching Visions of Light and I hope it happens when I include a visually stunning clip from an obscure Iranian film to which most students would never have been exposed. I feel a bit less certain when defending my use of interviews and B-Roll from documentaries I found on YouTube. Arguably, I’m mooching off of others’ hard work, and because lecture videos exist outside of the classroom context, I can see how they could be substituting for the original in some sense. I’m certainly open to a debate on what it’s okay to sample in a lecture video.

From the Specific to the Abstract

The sampling method is great when you’re describing a concept and want an accompanying visual to act as a kind of concretization of the concept, but this isn’t always what you want. Many textbooks and classroom lectures alternate between explication of abstract concepts and descriptions of specific examples, but most online video “explainers” either commit to showing specific examples of concepts explained on the soundtrack or show simple animations corresponding to the words on the soundtrack.

In its specificity, video forecloses imagination in a way that print and audio tend not to. When listening or reading, individuals can imagine what characters and places look like; in the learning context, students can fill in your own examples of a concept. So, that’s one thing I discovered through trial and error: there are certain ideas that I wanted, in some sense, to remain abstract. In such instances, I created animated diagrams that used shapes to represent concepts and their relationships to one another.

Creating and Maintaining a Visual Style

When creating these diagrams, I tried to think in terms of color, shape, movement, and using the entire frame. I used similar colors for ideas or concepts that are proximate to one another, and contrasting colors or shapes for concepts that are not proximate to one another. I didn’t bunch things up at the top of the frame just because the program with which I was working did this by default. I avoided using small fonts, anticipating that some students would be watching these videos on smaller screens. And I maintained visual appeal by moving the shapes around the frame, not in a gratuitous way (i.e., because I could), but instead as a means of improving clarity and keeping students engaged.

While I used some variety from minute to minute to keep students engaged (in terms of editing pace, color, and amount of movement), I tried to maintain some level of consistency across the videos. I found that if you use a lot animation in the first half of a video and don’t use any in the second half of the video, it was distracting (the viewer is left wondering, “where did the movement go? Why did things slow down?”). Each lecture video and the series of lecture videos will have a kind of internal stylistic norm that guides students’ expectations, and violations of those expectations will distract the student.

Here, it’s worth paying attention to how long it takes you to create your first lecture video. Of course, some of the process will become easier and quicker as you become more experienced, but not all of it will. If you develop an aesthetic that requires forty hours of work per video, you can easily end up with inconsistent qualities across your videos that is borne out of the necessity of finishing all of your videos in the allotted time.

Often, I wouldn’t know if a certain moment of the lecture would work well with a particular image until I tried it out. As the accomplished film editor Walter Murch notes, stitching together a motion picture is a lot like learning how to dance: you just need to get out there and try it.

Here is an example of my approach to sampling in online lecture videos:

In the next entry, I’ll share more of the technical details of the video lecture creation process.