Measurement 101: Types of Measurement

Created by Jake Bedard, Modified on Tue, 5 Nov at 4:46 PM by Jake Bedard

Measurement 101: Types of Measurement

Single-Channel vs. Dual-Channel Measurement

There are several different ways to measure and visualize sound information. However, all measurement methods can be broken down into two key categories: single-channel and dual channel. Single-channel measurement aims to analyze the frequency content and/or level of an individual signal. Single-channel methods can include RTA, spectrograph, and SPL measurements.

Dual-channel measurements, however, use a Transfer Function to compare two signals to each other: a reference signal and a measurement signal. Most commonly, the reference signal is a copy of the signal going into a system, and the measurement signal is the system's output. Since the two signals are correlated, the input signal is effectively "subtracted" from the output signal and we are able to view the system's response in isolation.

Since the source signal is known in a dual-channel measurement, we're able to get timing, relative level, and phase information. While frequency response and level are important, the inclusion of timing data adds a third dimension to understanding what happens to a signal when it passes through a system. This data is typically viewed as two traces: magnitude and phase, shown together on what is known as a Bode plot. Time-domain issues with single sources (such as reflections) and multiple sources (such as poor alignment) can lead to issues in the system's frequency response, but can't be fixed solely with EQ.

Impulse Response measurements, which analyze the acoustical response of a room to a sound source, can be measured either way: direct IR measurements are single-channel, and indirect IR measurements are dual-channel.

More information on Impulse Response measurement can be found in our 2-part guide:

Real-Time Analyzers (RTA)

A Real-Time Analyzer is possibly the simplest measurement method. It takes a signal and breaks it down mathematically by frequency, then displays the frequency content of that signal.

This is done via what is effectively a series of bandpass filters with level meters. Digitally, the data comes from an FFT, or Fast Fourier Transform (a mathematical operation that breaks a signal down into its individual frequencies) and is then banded into octaves or fractional octaves.

Spectrum data is most often banded into 1/3 octave bands and viewed on a logarithmic scale. Smaart offers higher resolution options including 1/6, 1/12, 1/24, and 1/48 octave. Octave banding is another commonly utilized option, although it offers the lowest resolution. If we band the entire spectrum together, we have a regular signal level meter. In fact, that's a good way to think of the RTA - having a bunch of level meters for different frequency ranges. More information on banding can be found here.

Smaart also has different options for averaging, other wise known as integration time. The higher the averaging value, the less "jumpy" the meters get. This can help produce a clearer picture of the signal's tonal balance as we view it over time.

Being a single-channel measurement type, an RTA only displays information contained within the signal itself. It is unable to account for any processing or time-related information. It is frequently used in order to visualize the tonal balance of a mix.

Spectrograph

If we were to view an RTA measurement "from above" over time, we can gain valuable insight into how the measured signal can change from moment to moment. This can provide context, especially when measuring a dynamic signal such as music.

As shown in this screenshot, brighter colors correspond to higher levels. Utilizing a spectrograph,

it is easy to spot feedback and ringing, as well as trends in frequency over time.

If we view a series of RTA measurements back-to-back, we can get a better context on how the signal's levels change over time at specific frequencies. Here's a split view showing an RTA below and a spectrograph above. In a spectrograph, brighter colors correspond to higher levels. This dual view is a big help when keeping track of mixing decisions. It's also the easiest way to spot feedback and ringing in a system. It can prove difficult on a simple RTA because once an obvious peak forms, it's likely very audibly loud in the room. The spectrograph can reveal a trend of ringing over time as a bright (high-level) vertical line.

Transfer Function/Dual-Channel FFT

A Transfer Function (or dual-channel FFT) shows the difference between two signals. As such, comparing a signal to itself via loopback or other method will result in a flat magnitude response. Any stimulus type can be used for a dual-channel Transfer Function measurement, but random or pseudorandom noise signals are the most common. (More information on noise types and contours can be found here)

It's possible to take several real-time measurements within a couple of seconds for averaging purposes, which can help with noise reduction. A Coherence metric compares the correlation of the measurement and reference signals as a percentage. More information pertaining to coherence in Smaart can be found here. This is helpful for spotting noise, reverberant energy, reflections, and other auditory phenomena that can lead to inaccurate measurements.

The end result is that we can measure more quickly and at a lower level using our choice of program material, so a dual-channel system is probably the friendliest choice for working on a system when other stuff is happening around us (both for us and for them).

Time Delay Spectrometry (TDS)

Time Delay Spectrometry (TDS) is a dual-channel measurement method that utilizes a sine wave "sweep" as its test signal. This signal "sweeps" at a steady rate from the low end of the frequency spectrum upwards towards the high end. While this sweep originally ascended at a linear rate, current methodology instead utilizes a logarithmic sweep, like you'd see from Smaart's signal generator. As such, it is commonly referred to as a "pink sweep" or a "swept" measurement.

Being a sine wave, only one frequency is passing through the system at any given point in time. As such, capturing the system response at any given moment during the sweep will show the system response at that specific frequency, in isolation. This allows for the visualization of any potential harmonic distortion caused by the system. (Note: Smaart is not intended for harmonic distortion measurements)

If you use a fast sweep, it's also possible to window out any reflections caused by the space in which the system is installed. A drawback of doing so, however, is that a shorter measurement time will limit the frequency resolution of your measurement. A shorter measurement time will also limit the lowest frequency you can measure due to the time required for lower frequencies to resolve.

Utilizing a slower sweep and a longer time window will allow you to measure lower frequencies, but at the cost of letting reflections into your measurement. It is important to consider the pros and cons of each and how they correlate to which elements you're specifically interested in measuring. Averaging multiple measurements together can help reduce environmental noise and reflections, but isn't completely foolproof, either. You might be interested in measuring the response of both the system and the room together, as in nearly all cases the system will be used in a space that will also affect what the audience hears.