A common question that people send to Jeff and me as the support staff at Wildlife Acoustics deals with how people should adjust their recorder settings. In particular, how to set gain. For example, “I’m trying to record X Species in Y Environment. How should I set the gain on my recorder?” The documentation for our equipment provides some suggested starting points, but I always tell customers they should test out their settings and make adjustments if necessary.
So, how do you know if adjustments are necessary?
Setting the proper gain means understanding a little bit about what gain is and where it takes place in the recorder’s signal chain - the path from sound hitting the microphone to a digital file stored on an SD card.
When sound hits a microphone element, it is translated into an electrical signal - small changes in a voltage level over time that imitate the changes in air pressure that make up a sound wave. Adding gain means taking this voltage signal and multiplying it by a fixed value. For example, you might start with a signal that swings from -10 millivolts to 10 millivolts, then add gain to get a signal that swings from -100 millivolts to 100 millivolts. You can think of gain as making a signal louder.
All of this happens before the signal is turned into a digital representation that can be stored on an SD card. It’s still just a voltage wave traveling along a wire until it hits the analog-to-digital converter (ADC). The job of the ADC is to break up the voltage signal it receives into a series of numbers. ADCs are built with fixed voltage ranges they can convert, and a fixed number of different values they can spit out. For example, in the SM4 recorder, the ADC can accept Voltage signals between -1 Volt and +1 Volt.
Now, remember that gain determines how much louder a signal is made. What happens if, after adding gain, a signal reaches values beyond the range of the ADC? In our SM4 example, what happens if you have a signal that swings between -1.5 Volts and +1.5 Volts?
In the pictures below I have generated a 440 Hz sine wave, a pure waveform that is just about the simplest signal you could record, and am viewing it in Kaleidoscope. In the first image, the waveform view at the top of the screen shows that the sine wave fits entirely within the waveform view when I’m zoomed all the way out. The signal is loud enough to be clearly read without exceeding the range of the ADC. Notice that the spectrogram view shows a single line at 440 Hz, with no other frequency content.
In this next image, I have increased the gain on the sine wave until it has exceeded the range of the ADC. You can tell this is the case because the whole shape of the sine wave does not fit into the waveform view. It is instead “chopped off” at the top and bottom. Notice that instead of a single frequency being shown in the spectrogram, we see a lot of higher frequency content. By flattening the peaks and valleys of my sine wave, I have changed the shape of the waveform, and added a bunch of frequency content that wasn’t there in the first place. (Why this happens is a topic for another day!)
So, if you notice that your recordings are frequently being “chopped off,” this means that the signal hitting the ADC is too loud, and the recorder is unable to accurately capture the full shape of your sounds. To remedy this, simply try lowering the gain.