Looking to Improve Educational Video? First Improve the Audio
In the July/August issue, I suggested a back-to-school shopping list for the teacher who is getting started with video production. Since there are excellent free tools available, my advice was to get to work using them and to spend your money only after you discover where your resourcefulness begins to fail. In this case, premature optimization is the root of all overspending.
Students tend to be generously tolerant of many shortcomings in otherwise-helpful instructional video, but not audio problems. Thus, it is likely that improving the audio quality is where teachers will first be looking to spend their money. While educational video is a small niche for the recording industry, our needs intersect with other market segments that collectively create adequate demand for a wide variety of recording tools.
Improving your audio quality is best done in preproduction, either by making the environment quieter or by improving the fidelity of the audio signal being recorded. Fixing audio beyond normalization in postproduction is always destructive. You can never add quality speech signal; you can only remove some amount of undesirable noise without noticeably damaging what quality speech signal was captured.
The traditional audio recording setup involves a microphone supplying a mic-level signal to a pre-amp, where the signal is substantially amplified to line-level while introducing minimal noise. The pre-amp additionally provides power to condenser microphones. This line-level signal can be captured by the input on your computer’s audio card to produce a discrete analog of the signal as a sequence of samples. Noise can be introduced at any of these points or in cables between them. In almost all cases, I’d recommend minimizing the points of failure by using a quality, all-in-one USB microphone.
When shopping for microphones, there are two critical specifications to look for: the frequency response and the polar pattern.
The frequency response is usually reported as either a range of frequencies that the microphone will capture or as a graph showing the sensitivity at different frequencies. Inexpensive microphones intended for telephony are designed to carry signals over the limited channel—typically only recording down to 100 Hz—that telephone lines handle, and you should avoid them. Unless you have a very-high-pitched speaking voice, that won’t work well. The human ear can hear sound at frequencies from around 20 Hz to 16,000 Hz; the lowest pitch notes demanded of professional opera singers are above 70 Hz (73.4162 Hz to be exact), so that’s a safe low range for instructional speech sounds.
The polar pattern describes the microphone’s sensitivity depending on the location of the sound source around the microphone. For recording yourself, you should choose a cardioid pattern, which means that it records what it’s aimed at very well and it depresses the ambient noise from all other directions. Some microphones that are popular with educators use a bidirectional polar pattern, which is great for interviewing a guest from directly across a desk.
There are three general microphone form factors that make sense for teachers: stand mics, headset mics, and lavalier mics. Historically, I’ve recommended either headsets or lavaliers, since incorrect mic placement is a common problem that can be avoided by using microphones attached to your body close to your mouth. I felt that tabletop stand microphones would get in the way of the camera and the teacher would vary their distance from the microphone too much, leading to uneven recording levels. I’ve since come around on this: A well-placed stand microphone tends to keep teachers from fidgeting away from the microphone and around the frame.
If you do want to spend more than $60 on improving your audio, one potentially helpful resource is your local guitar shop. Everyone who works there has had to teach guitar lessons, so they’re likely to be sympathetic to a fellow teacher. Many guitar stores have rental mics that they may let you try out to help decide which one works best for your voice.
[This article appears in the September 2018 issue of Streaming Media magazine as "Sound’s Good."]
A Sonic Foundry survey on higher education and artificial intelligence suggests personalized lessons could lead to better retention.
Bringing video into the classroom doesn't have to be expensive. Read this helpful list to discover the seven basic items every classroom should have.
Measuring quality of service is big business in the streaming video world, but educational video metrics lag far behind. Part of the reason is that educators need completely different metrics.