SPL measurement far-field | Soundeasy tutorial
How to make a SPL measurement using Soundeasy?
Anyone can do a far-field SPL measurement, if you have an appropriate microphone. However, trying your luck with Soundeasy may prove quite frustrating. We all know that it’s not user friendly, but I encourage you to soldier on, until you get the hang of it. First of all, you need to know the basics on how measuring is done. Normally, this would be done in an anechoic chamber, but odds are excellent that you don’t have access to one. The solution is to make quasi-anechoic measurements. Therefore, I encourage you to get familiar with these types of measurements here, because I will not explain the terminology again. I would also invite you to check out some quasi-anechoic speaker measurements using ARTA. Although not mandatory, it will get you more familiar with the process. For now, we will do a gated far-field frequency response measurement.
Before we start the application and get all messy with Soundeasy, we need to setup some hardware. Things you will need :
- A microphone stand.
- A speaker stand.
- The Soundeasy probe jig.
- Measuring equipment (I’m using a Dayton Audio EMM-6 microphone and a Focusrite Scarlet 2i2 audio interface).
- A large, tall, quiet room with no obstacles / objects (in the vicinity of the speaker).
As I am sure you have read the article about the types of measurement, I will be redundant and do a brief run-down regarding the gated SPL measurement. The problem with measuring inside a normal room is that the sound waves will bounce off the walls, and you will measure the speaker in conjunction with the response of the room. To get around this, we do a gated measurement. This means we calculate how long it takes for the first room reflection to hit the microphone, and ignore everything after that. This way, we get only the speaker’s response.
Why do I need an anechoic chamber then? This method has its limitations. Long sound waves (bass) can’t be recorded without room reflections. The bigger and taller the room, the lower you can go in frequency with the measurement. For the low frequencies, we will use the near-field method, and splice the two responses.
Depending on the size of your room, there is a certain time frame where the sound recorded by the microphone is reflection free. Usually, the first reflection will come from the floor or ceiling. You have to place the speaker on an adjustable stand. This is to place the speaker exactly half way between the floor and ceiling (maximizing the distance between the speaker and the first object).
As you can see, the reflected wave takes a longer time to reach the microphone. The first reflected wave that reaches the microphone will contaminate the response. So we need to make it travel longer, by having a larger room. Also, jacking up the speaker and microphone stands too much, will make the first wave bounce off the ceiling, that’s why you need to place them half way up. As a result, make sure there are no obstacles around that will generate a premature reflected wave.
Calculate window width
Now let’s calculate the gate window. First of all, we need to calculate the distance the reflected wave travels.
We know the dimensions stated in the picture above, because we measured them with a measuring tape. The distance between the speaker and microphone should be 1 meter. You can set any distance you want, but 1 meter is industry standard, and we will go along. I measured the distance from the microphone to the floor and it’s 1.2 meters (this will depend on your room size). From now on it’s basic Pythagoras : c2 = a2 + b2. Doing the calculation c = 1.3 meters. The reflected wave travels twice that distance, so 2.6 meters.
How to design loudspeakers - video courses
Now, you have to realize that until the direct wave reaches the microphone, no sound is recorded. In conclusion, we are interested in the interval after the direct wave hits and before the reflected wave arrives. We subtract the distance traveled by the direct wave : 2.6 – 1 = 1.6 m. By knowing the speed of sound (343 m/s). we can calculate the time we have available for a SPL measurement with no room reflections.
Reflection free window = 1.6 / 343 = 0.0047 s = 4.7 ms
Probe and microphone setup
Before we do anything, we need to make the appropriate connections. If you made a Soundeasy probe jig like mine, it’s all modular, and this is handy right now. We don’t need the 10 ohms resistor so we take it out, and we also need just one probe. Here is a chart with which goes where :
In the chart above you can see there is an optional 20 μF capacitor. This is only used when measuring tweeters. When you do a frequency sweep, the tweeter doesn’t like low frequencies very much, and you might damage it. The capacitor is basically a 1st order Butterworth high pass filter. If the tweeter is 8 ohm the crossover point is around 1000 Hz. However, if you are measuring some other type of speaker, that doesn’t suffer from low frequency damage easily, make sure you don’t use the capacitor. This will be our case (no capacitor), since we are measuring the SEAS CA 18 RNX.
First of all, we need to upload a calibration file for the microphone. What’s nice about the Dayton microphone, is that you can download the calibration file for your particular microphone. As a result, simply go to their website, and enter the serial number, which you can find at the base of your microphone. After you download it, open the file with notepad and delete any lines at the beginning and/or the end of the file that are not frequency adjustments. It should start with 20 and end with 20 000. My file had a line at the beginning : “*1000Hz -40.2”, which I deleted. How to upload the file and do all the necessary settings, please check the Soundeasy design guide (User Help -> SoundEasy Help).
Sound levels calibration
Now go to EasyLab -> Digital MLS and do the following settings :
- Set the resolution of the MLS. I usually go for max.
- Sample rate, depending on your sound card. Usually 48000.
- The output you set by trial and error, in conjunction with the volume on your amplifier. You increase it in small steps and check the levels each time. Please observe on the right (Ref / In) an example of appropriate levels. As a recommendation, check out the impedance measurement article to learn how to set up your levels correctly without damaging your speaker / sound card (it’s the same process, only for a different purpose). Also, make sure the phantom power (48 V button) is on, as the microphone requires it. I would also like to mention that the gain (physical knob on the audio interface) for the microphone is 100%, and for the probe is around 75%, because I couldn’t get a decent level for the microphone without the probe going off the chart.
- Enter the time window we calculated previously : 4.6 ms.
- This is the button you use to check the levels (for point number 3).
- You can add decibels to shift your frequency response chart upward. People like when the graph is around 90 dB. Make sure you add the same number of decibels to all the measuring you are doing for that particular project. This is possible because we are doing a relative SPL measurement, not an absolute SPL measurement.
- Smoothing of 1/6 should be fine.
- Check box to use the microphone calibration file.
Now, to do the actual SPL measurement do the following instructions :
- This is not an actual instruction for measuring SPL, just a mention I would like to make. If you click the “Show” check box, found to the right of “Window Width”, you will see this dotted area on the graph. Depending on how big your window is, the area will be smaller or bigger. It tells us that we can make accurate gated frequency response measurements above 250 Hz. Below that point, the response is not valid, hence the dotted area.
- Click on the IR vs Time window just before the 1 meter mark. After you click the graph, you can use your arrows to fine tune it, till you get it right below 100 cm.
- Click the MLS button to take the measurement.
- Click the IR -> SPL to get the frequency response chart.
That concludes our SPL measurement, and here is how the response looks like :
It’s always nice if you can check your results, so you know you didn’t do any mistakes when setting up SoundEasy. Therefore, I fired Room EQ Wizard and did a similar measurement. Here is the result :
Looks good! The shapes of the graphs are almost identical. As a conclusion, I want to say that this is not the full frequency response of the SEAS midbass driver, because it misses the low end. We need to do a near-field measurement for that, and merge it with the far-field response. But that is for next time.
- Image source : link.