Making Direct Measurement Videos
By Peter Bohacek, Henry Sibley High School
Note: This page is under construction. But since I have had requests for information on making videos, I am posting this preliminary version.
Making videos for direct measurement is within the grasp of anyone with a digital camera and basic video editing software. Here are some of the techniques I use.
The overriding quality that I seek in these videos is clarity. I work to compose shots so that the motion we're analyzing is is vividly captured in a way that evokes curiosity. Ideally, these videos should make the viewer want to reach for a paper and pencil and begin working out a solution. Lighting, lens choice, and background color are key components to making a compelling video.
The second attribute I seek is the ability for the viewer to make unambiguous measurements. For example, during a collision, students must be able to clearly measure the velocities of the objects before and after the collision. Frame rate and lens focal length determine the time and distances the viewer can see. I aim for about 1-2% measurement uncertainty, but settle for as much as 5%.
Although there are less expensive cameras that can record high frame rates (such as the Nikon 1, and some Casio Exelim models), the Sony is the least expensive camera I have found that records high-resolution, high frame-rate video. I have not been able to get sharp clear images at high frame rates using less expensive cameras. However, less expensive cameras, like the Sony NEX6, will record beautiful 60 frames per second video, which is a sufficiently high frame rate for many interesting events. In many cases high resolution is at least as important as high frame rate. For example, when I made this video (which is not actually part of the Direct Measurement Video collection), I had several inexpensive high speed cameras, and a Canon T2i. Although the maximum frame rate for the Canon was 60 fps, the resolution and ability to control the shutter speed allowed clearer, sharper shots so that the position of the marbles is clearly visible.
I use Zeiss 35mm lenses. These are completely manual lenses, with wide apertures (f1.4 to f2.8) and superb image quality, particularly when compared to most zoom lenses. Wide apertures make the most of available light, which is allows shorter exposures minimizing motion induced blur. In addition, wide aperture reduces depth of field, pleasantly blurring the background. This has the effect of focusing the viewers attention on the moving objects in the video, rather than being distracted by the background. I use a 100 F2 most often. The longer focal length reduces angular distortion.
I use a pair of Octacool-9 softboxes for shots up to 60 frames per second. These provide pleasing light with soft shadows. They run cool, and have daylight balance. The downside is that they flicker at 60 Hz, which is clearly visible when recording at high frame rates. I also use halogen spot lights. These produce intense light, and do not flicker noticeably, but they also cause harsh bright spots and shadows. I often use reflectors to soften the light from these. We recently purchased three pairs of Generay LED arrays. Because they are powered by DC supplies, flicker is not noticeable at 240 fps, but is visible when using higher frame rates.
One wall of our lab is painted with several coats of low-gloss, super white paint. For many shots, this white backdrop provides ideal contrast for clear measurements. Black cloth provides a high contrast backdrop for lighter color objects.
Video is imported into Apple Final Cut Pro 7. Grids, rulers, and other overlays are drawn in Adobe Illustrator. These images are imported into Final Cut Pro as .png images. I have found that using Adobe Photoshop to convert the drawings into .png files works better than exporting directly from Illustrator.