IR Measurements, Part 2: Utilizing Smaart

IR Measurements, Part 2: Utilizing Smaart

Once you have determined your source and measurement positions and chosen a measurement technique, Smaart's involvement is quite straightforward.

It follows these basic steps:

• Select your measurement parameters, including:
  • Input source
  • Excitation level
  • Measurement duration (time window)
  • FFT Size
  • Averaging & overlap
  • Delay compensation
• Turn on the signal generator or other stimulus signal source.
• Start the measurement and save it.

From this point, you are free to analyze the impulse response data you've collected. A list of common IR measurement settings can be viewed here. The following guide includes a more detailed guide to each setting as well as how to make and save the measurement in Smaart.

Measurement Settings in Smaart

The two main things you need to concern yourself with at this point are stimulus excitation level and measurement duration, which are determined by a combination of FFT size and the number of averages.

Excitation Level

As a starting point, it is recommended that you measure your background noise level. Your excitation level will be set in relation to this baseline. Since this level is used to set a relative threshold, precise SPL calibration isn't necessary (unless you're measuring STI). To do so, set the sound level meter in Smaart to "Slow SPL" and let the meter run for 10-20 seconds with no output signal running. This will give you a general idea of the background noise level, against which to set your excitation level. Then, start the signal generator at a low level and gradually increase the gain until you reach your target level (or as close as possible).

Ideally, your excitation level should be at least 40-50 dB higher than your background noise level. In reverberation time (RT60) measurements, reverberant decay is over a range. This range starts 5 dB below the arrival of direct sound (typically the highest peak on the plot) and extending down another 20 or 30 dB from the start point (while a 30 dB range is preferred, 20 dB is also acceptable). The lower end of the range should be at least 10 dB above the measured noise floor.

Following these guidelines, you should see a minimum of 45 dB of dynamic range for a 30 dB evaluation range, or at least 35 dB for a 20 dB range (in an ideal situation with no noise artifacts from the measurement process itself). In more realistic situations, adding another 5 to 10 dB on is advantageous (so long as it doesn't drive the system into distortion or cause damage).

Measurement Duration (Time Window) and FFT Size

For dual-channel measurements, the time window (measurement duration) is determined by the FFT time constant - that is, the time required to record enough samples for a given FFT size at whatever sample rate you are using. In Impulse Response mode, Smaart gives you the time constant in milliseconds, along with the frame size in samples for each available FFT/DFT size.

If you are measuring:

• Delay times only: 
  • The measurement should be 3 times the longest delay time you want to measure. 
• Reverberation time and early-to-late energy ratios:
  • The 60 dB decay time (RT60) of the system is a good target. 

This is not only a requirement for period-matched/dual-channel measurements, but a good general guideline in general. Ideally, you would like to measure 30 dB of reverberant decay with the lower end of the evaluation range +10 dB above the noise floor, which would be 40 dB (2/3 of 60). By the time you factor in propagation delay, early decay and maybe enough of a noise tail to see the dynamic range of the measurement, chances are you have eaten well into the remaining third.

Both rules require either knowing either the delay time or RT60 before measuring, however. This typically means you need to guess, then measure, then possibly adjust and measure again. For delay times, you can divide the distance to the source by the speed of sound, to start. The speed of sound,  1130ft/sec (345m/sec) at typical room temperatures, increases with temperature, so you may need to adjust this value if measuring a particularly hot or cold room.

For reverberation times, one to two seconds is a useful starting point for most theaters/auditoriums. Indoor stadiums and other large structures tend to have much longer reverb times. You may want to make longer measurements then you think you might need, as there is no downside to do so. Lower frequencies tend to decay more slowly than highs, so lower octave decay time is the key deciding factor in measurement length.

Averaging and Overlap

Averaging is remarkably more effective when using random stimulus signals. With random (or effectively random) signals, typical averaging settings are in the 4-16 range. Noisier environments may require a larger value (or just resorting to using a period-matched signal). When measuring with period-matched noise or sweeps, averaging is normally set to "None" or 2, although it is still possible that a higher setting could prove helpful if measuring in an extremely noisy environment.

The Overlap setting also affects averaging, and can be adjusted in the Impulse Response Graph Options menu (accessible by either the [Alt/Opt] + [i] hotkey or by navigating to Options>Graph Settings>Impulse Response). When set to 0%, each FFT is calculated from unique data for maximum noise reduction in that averaging setting. When set to a non-zero value, some data carries over to successive FFTs, similar to spectrograph overlap.

If overlap is set to 50%, it takes slightly longer to record 16 averages than it would for 8 at 0% overlap. You don't get the full benefit of averaging 16 unique FFTs in that case (with increased processing time) but you will experience better signal-to-noise than you would get 8 with some net time savings.

Delay Compensation

When using random signal sources your results will improve if you compensate for the delay time through the SUT. It is typical to measure twice if the delay time isn't already known: once to find the delay and a second time utilizing the delay. Clicking the circular gray button to the right of the delay field in the measurement engine control block sets reference signal delay to the highest peak in the impulse response.

Note: Setting the delay to a non-zero value does not affect the peak location in the resulting IR. Smaart will apply delay compensation while making the IR measurement and then back out the delay time upon completion. This preserves the flight time for direct sound in the time record.

Finally Making the Measurement

At this point, all of the groundwork has been laid out. Outside of Smaart you've chosen your measurement technique, excitation sources, and measurement positions. Within Smaart, you've set your input/output levels, FFT length, and number of averages, if applicable. All that's left is to take the measurement itself. 

• For a dual-channel (indirect IR) measurement: 
  • Start your excitation signal (unless using a triggered sweep, in which case the generator will start automatically when you start the measurement).
  • Click the start (play) button at the top of the Control Bar. 
  • Smaart take the measurement and display the results when complete.

• For a single-channel (direct IR) measurement: 
  • Click the record button.
  • Click the start (play) button. 
  • Trigger your excitation source (pop your balloon/fire your blank pistol/etc.)
  • Give the system a moment to ring out.
  • Click the stop button to end the recording and display the results.

Saving Your Work

When you make a new IR measurement, it is not saved automatically. Since you may need to make several IR measurements while adjusting your settings, this would create unnecessary data bar clutter. To save your measurement either press [Space] or click the Save button on the Data Bar. Doing so will only save the top trace.

You will then be prompted to name your measurement. When you click OK on the Save Trace dialog, the new file will immediately appear in the current session folder in the Data Bar. If you have cropped the file for display purposes using the Crop function, only the displayed portion of the time record is written to file.

Additionally, you can save all displayed measurements by using the shortcut [Shift] + [Space] or by clicking the Save All button on the Data Bar. In this case, you will be prompted to name the destination folder and each measurement will be saved there as their measurement engine names.