Acoustics 201 : Loudspeaker measurements | Tanasescu Marius | Skillshare

Acoustics 201 : Loudspeaker measurements

Tanasescu Marius

Play Speed
  • 0.5x
  • 1x (Normal)
  • 1.25x
  • 1.5x
  • 2x
41 Lessons (3h 35m)
    • 1. 0 - Introduction

      2:14
    • 2. 1.1 - Contents

      3:20
    • 3. 1.2.1 - Equipment

      6:34
    • 4. 1.2.2 Equipment - 2020 Update

      3:56
    • 5. 1.3 - Software pick

      3:36
    • 6. 2.1 - UMIK-1 wiring diagram

      2:08
    • 7. 2.2 - Dayton Audio wiring diagram

      2:33
    • 8. 2.3 - Calibration file

      5:33
    • 9. 2.4 - UMIK-1 Room EQ Wizard setup

      5:09
    • 10. 2.5 - Dayton Audio Room EQ Wizard setup

      2:52
    • 11. 3.1 - Types of measurement and Far field in detail

      7:42
    • 12. 3.2 - Near field measurement

      4:19
    • 13. 3.3 - Half space and ground plane

      2:12
    • 14. 4.1 - 2-way speaker measurement | Presenting the setup

      4:27
    • 15. 4.2 - Measuring distances

      1:48
    • 16. 4.3.1 REW Interface Update

      1:36
    • 17. 4.3.2 - Far field midbass

      7:27
    • 18. 4.4 - Far field tweeter

      2:48
    • 19. 4.5 - Near field midbass

      3:29
    • 20. 4.6 - Near field port

      2:00
    • 21. 4.7 - Error, heads up!

      2:34
    • 22. 4.8.1 REW Interface update 2

      2:07
    • 23. 4.8.2 - Combine midbass + port

      3:07
    • 24. 4.9 - Box diffraction for near field measurement

      5:12
    • 25. 4.10 - Splice near field with far field

      4:04
    • 26. 4.11 - Create FRD files

      1:57
    • 27. 5.1 - 3-way speaker measurement | Presenting the setup

      2:24
    • 28. 5.2 - Describing the measurement process

      7:12
    • 29. 5.3 - Raw speaker measurements

      9:39
    • 30. 5.4 - Processing of the measurements

      18:07
    • 31. 6.1 - Speaker full frequency response - anechoic

      11:47
    • 32. 6.2 - Speaker full frequency response - at listening position

      11:37
    • 33. 6.3 - Distortion measurements

      10:47
    • 34. 7.1 Presenting DATS v2/v3

      2:48
    • 35. 7.2 - How to measure the Thiele / Small parameters

      13:00
    • 36. 7.3 - How to create ZMA files

      5:17
    • 37. 7.4 - Measure electrical components

      4:33
    • 38. 8.1 - XSim presentation

      7:33
    • 39. 8.2 - Acoustical offset - measurements

      8:36
    • 40. 8.3 - Acoustical offset - processing of data

      3:52
    • 41. 8.4 - XSim - finishing touches

      3:02

About This Class

Learn how to measure frequency response

The first thing you will learn from this course is to measure the frequency response of a certain speaker driver, loudspeaker or sound system. The difficult part is to get an anechoic response without an anechoic chamber. Using different measuring techniques and with the help of intuitive software, you will learn how to measure the full frequency bandwidth anechoically in your own living room. This will help you set up your audio system using measured data, or it can be used to design crossovers for your particular DIY loudspeaker.

A little bit of electrical engineering

A sound system is a combination of both acoustical and electrical parts. You will also learn how to measure impedance and create impedance plots. Measure different electrical components like capacitors, inductors and resistors. Also, if you have an unknown speaker driver you will learn how to measure the Thiele / Small parameters without much fuss.

Dip your feet into crossover design

While this course will not teach you how to design a crossover, it will show you how to create FRD and ZMA files. These are frequency response and impedance files which are used in crossover design apps. At the end of the course I will show you how to load these files into such program, and tryout different circuits and see how adding crossover components behave. A comprehensive course about crossovers will be released in the near future.

Do I need certain equipment?

Yes, you will need 2 core pieces of gear :

  • A measurement microphone (like miniDSP UMIK-1).
  • A device for measuring impedance (like Dayton Audio DATS v2).

These are roughly 100$ each, but you will get a more comprehensive list of the equipment needed inside the course. With alternatives that you can buy, or more cheaper options. Also, some accessories that will make your life easier are included in the list.

Software option

For the different acoustical measurements we are going to use Room EQ Wizard (or REW). This is a free to use software. Usually, acoustics software are quite expensive and difficult to follow. However, REW is free and quite intuitive to use. Pretty much a complete package. It only misses one function, which I consider important. Don't worry, we're are not going to step up to those complicated application. I devised an Excel spreadsheet which handles this shortcoming. You will find the spreadsheet inside the course.

Transcripts

1. 0 - Introduction: Hello, guys, and welcome to acoustics to one. The second course in the acoustic Siri's I'm Marios any most likely know me from the acoustics. Wanna one? Of course, you're probably wondering what this course is all about. Well, it serves two purposes. First, you can view it as a standalone course, where you will learn how to make frequency response measurements, and you will measure the audio system from your home or from your car. And you can make adjustments using measured data rather than just by year. You will also learn how to make distortion, measurements and troubleshoot problems. Impedance measurements. You will learn how to measure the feel small parameters off a speaker and much more. The second purpose of this course is that it will serve as a foundation for the third course, which will be about crossovers. You cannot design a crossover If you don't know how to make this measurements, you need to make the frequency response in the impedance measurements off your exact enclosure. Otherwise, you cannot design a crossover properly. Now, this is a brief rundown of the course, and I sense that you might have some questions first would be doing need to take acoustics . Wanna one first? Well, if you know the basics and know how toe design an enclosure, then you might as well skip it. But, um, if you find yourself a bit confused, then you don't understand some terms that I'm saying. Then that might mean you should go ahead and take acoustics one on one first. The second question you might be having is a Dwight need a certain equipment to make these measurements. And in short, yes, you do you need, Ah, a special type of microphone to make the frequency response measurements and, ah, a small device to make the impedance measurements. I will make a briefly stir inside the course description and a more detailed one inside the course. See you there. 2. 1.1 - Contents: Let's talk about what should you expect from this course. First, we're going to talk about the equipment you will need. You will have different options to choose from. I'm going to show you what I use but also offer a cheaper alternative. Then we will talk about the software we will use and the advantages and disadvantages that comes with this choice. Secondly, we're going to talk about how to set up the equipment. Depending on which type of gear you choose the set up. It's slightly different. After everything is old wired up, we can get toe sitting up the software. And finally, after all the settings are complete, we can get to actual measurements. But before we get to the measurements, I'm going to briefly review the theory behind every measurement technique. You have to realize that depending on what type of speaker you are measuring, you need to adapt. Each type of speaker is more or less unique on how you will measure it. You need to know the theory so you can choose the appropriate the measurement technique. Then we're going to do an actual measurement. First, we're going to measure a two way bass reflex speaker We're going to do on an earthquake measurement for each individual speaker the Tweeter in the mid base and create corresponding if RD files next, we're going to measure a three way based replace speaker with the dual wolfers. I cannot cover all the possible configurations off loudspeaker design. Instead, I choose another example, which is very high in difficulty. If you get this one right, you will have no issues with the simpler ones after we learn how to create a farty, files were going to do some other types of a critical measurement like a full frequency responsible speaker. And it quickly how to measure speaker with room response at listening position and try to get the linear response by altering the position of the speakers or by digital correction. Also, we will do some distortion measurements as well. Then we will switch to the electrical measurements. I'm going to show you how the Dayton Audio test system works and how to measure. Do you yes, parameters and different electrical components and, of course, how to do any impedance measurement and create Azia may file. Lastly, we're going to play around in Exim, a crossover design application. I'm going to show you how to load the Frd and Zia May files. Also, I dedicated lesson in which I show you how to calculate the acoustical offset between the drivers, which is in this city for this application. After everything is in place in Exim with the basic knowledge you have from acoustics one on one, you can design some basic filters and see in real time what happens when you change the values off the components. 3. 1.2.1 - Equipment: No. Let's talk about the audio equipment you will need attached to this lesson. You will find a pdf file where I listed a couple of items that you will need for making these measurements. So, um, when you open it, you will find the first the the options for making frequency response measurements. I listed two options and an alternative also the device for making the impedance measurement and, ah, a couple of extra items that you will need for making the frequency response measurements. Now, if you look at this table, you will find that I made a brief description of the item of picture. I also insert on Amazon link where you can buy it linked to the manufacturer website if you want some mixed rain for and, um, a price, which is the price when this course was filmed. So with my very depending on the time you check the price. Now let's start with the second option for the frequency response measurement. We have the Dayton audio microphone. This is a frequency response measurement microphone, and it uses three pin XLR connector. This is because it the needs phantom power toe work. This means you have to connect it to a no odor interface which has phantom power. This this means it needs at constant stream off 48 votes for the microphone to work. So, um, my option for the audio interfaces the focus, right, Scarlett? Too white, too. This ah audio interface has two inputs. It has two outputs. It uses ah, USB connection and it doesn't need additional power. So you can use it on the go using your laptop, for example, this device and not only is a microphone pre amplifier that phantom part which I talked about earlier, but it's basically also on external sound card and we will use the outputs for making the frequency response measurements Um, additional uses for this device. Maybe you are using studio monitors and you know So your monitors have, ah, well integrated amplifiers in some kind of knob, which you control the volume on the back so you can sit the that volume and then used this volume to control it remotely and more, more conveniently. So this is ah not only just for frequency response measurements, you can use it for other stuff as well. The other option for frequency response measurements. Is the medias be you, Mike? One. This is the same as the Dayton audio microphone. The only difference is it doesn't need phantom power. Well, it actually does, but it draws it from the USB connection. You connect Lee, you connected directly to your computer or laptop. You don't need another new interface. So if you look at the prices, you can see that the U mike one is $106. If you access the manufacturer link, you will find that the price listed there is $75. But bear in mind that the item is located in Hong Kong. So if, for example, you are from the United States, you have to add the transport and the import duties, and most likely you will exceed $106. Anyway, if you look at the second option, the diet and audio microphone in the focus right, the audio interface, you can see that it adds up a little bit more. Maybe some of you already have an audience audio interface, and you can buy the cheaper die tone audio microphone. If you go to the next page, you will find a cheaper alternative, and if you go in from further, you will have the impedance measurement device. This is also a Dayton audio device and it uses ah USB connection. And at the other end you will find two alligator clips which you will use to connect to the speaker. With this device, you can measure the impedance and ah, you can also measure the fetal small parameters of the speaker. You can measure the resistance and the induct ins and capacitance off electrical components . Now, if you have ah, multi meter, you can measure the resistance for sure. But, uh, most of them cannot measure ah capacitance and certainly cannot measuring in darkness. So that is a plus. It also has other functions, like signal generator or Robin bus tests, but the we won't use them in the discourse. If you go to the end of the pdf file, you will find two additional items, which we will use four frequency response measurements. You will need a microphone stand because the microphone needs to be a certain distance from the floor and, uh, a speaker stand. I found them in pairs, but you only need one stand. So if you find a place where you can buy just one stand. That's enough. This is because if we are measuring a bookshelf speaker, uh, in the same fresh and the speaker needs to be a certain distance from the floor. Also, the stand needs to be height adjustable. 4. 1.2.2 Equipment - 2020 Update: now, since two years have passed since I made this course, I felt the need to update this equipment list file because some things have changed in some heaven up. So Ah, the microphones didn't change. But this product, the focus, right Scarlett has been discontinued. But don't worry, It has been replaced with another version which is basically similar. This is the third generation and if we look at the lot layout, we can see that it's basically the same. They have replaced those two way switches with the individual buttons and it looks more premium. They probably made some modifications inside, so it has better chips or something like that. But the important factories that it works the same and it doesn't affect them. How things will go on in our course if you chose to buy this equipment, for example. Now, another thing that has changed. Um, this Berenger audio interface I changed with another one because the last one has risen price by quite some margin, so I felt the need to replace it with something else that is cheaper. And ah, the Dayton audio test system has also been discontinued. And now Ah, new version is out. Uh um, the operation is the same, so the software is the same. So don't worry. Even if you see me working with the version two when you buy the version three, it's, ah, the same. It's the same thing and I will briefly show you the differences between the two, which I assure you are just a few. So we're looking at both the devices right now, and it's not much of a difference. So besides the fact that this one uses different electron ICS, and they say the measurement is more accurate, let's look at the subtle differences. So this one has the wires attached so you cannot remove the alligator clips or the USB cable, and you can see this one. You can remove the USB cable, and also, if you damage your cables, you can replace them. And, um, later, in the course, you will see me calibrating this device, and in the deaths V to box, you will find this resistor, and you will have ah, you'll have to connect the two alligator clips to so this resistor and then you will make the calibration procedure. And in the deaths be three box you will not find this resistor when said instead that the resistor is inside the device itself. So you have to connect the alligator clips to these polls like that. And now you can do the calibration procedure. I will show you again this image when we will do the extra calibration. Now the rest of the equipment remains the same. So the accessories so not much has changed. Don't worry about that. Just continue the course as normal. 5. 1.3 - Software pick: I know that we finished talking about the audio equipment we need. Let's talk about the software we will use. We will use room, make you wizard. Now this is a free software, and despite this, it's easy to use an intuitive. It's perfect for the beginner we can use Ah paid software, which is, ah, more complete package. But it's very hard to understand and has all this additional options, which we will probably never use. The downside for Roma Que Wizard is that it doesn't have a function which I considered mandatory for frequency response measurements. And to get around this, we're not going to go into the paid applications own. I devised the An Excel spreadsheet, which we will use to get around this. So basically we make a measurement in room, make you wizard export the measurement in a text file, then copy the values from the text file into the Excel spreadsheet, apply the function, then copy back the new values into the text file and then import back into the room a coup wizard. Now I know it sounds like a long process, but trust me, it's easy to do so to get rooming que wizard? You open up your browser. I opened up edge browser because I have some personal stuff in my chrome. So we go to Google and type room. Thank you, wizard. And the first result. Now you can see they recommend the exactly the same equipment we talked about in the earlier lesson. So we can use the mini DSP Mike or the Dayton Rodeo microphone with some audio interface, including the focus, right, Scarlett. Now both of these microphones are calibrated microphones. We'll talk about this when we will set up the room me to wizard because it's very important . If your microphone is not calibrated, then you My advice is that you shouldn't buy it in the first place. Now, to download the software, you simply click the down long link and then you will run. The is taller. I'm not gonna run it because I already installed the room, make you wizard after you completed the downloading the installer and install the application. When you fire it up, it should look like this Now, depending on the audio equipment you have, for example, you might have the immunity SB mike or the Dayton audio, plus the audio interface combo. The actual set up is quite different, and you have to set up room McCue wizard differently now, To be honest, I didn't even have the immunity is being microphone, and I bought it, especially for this course, because I presume that the set up would be different and the you guys might have some questions and I couldn't be able to answer him because I didn't have the equipment. Now we're gonna talk about more about setting up the equipment in the next lesson. 6. 2.1 - UMIK-1 wiring diagram: Let's talk about how to set up the mini DSP microphone, so you need the the microphone. And inside the package you will also find cable, which is Ah, mini USB to USB cable. This is what you will use to connect the microphone to a computer. And also you will need cable to hook up your computer to your amplifier, and you will have it one in and mini jet blood stereo Jack and a pair of R C at the other end to hook up to your amplifier, included. With this lesson, you will find AH picture file, which is available for download, where you will find a graphic representation of the set up and how you connected. It should device the which also I will show you in real time how to do the connections so everything will be crystal clear. Now here's a view of the room where I will make the measurements and you can see the microphone stand with the minute es Be mike in front of it. We have ah speaker stand with a bookshelf speaker, which I will use for various testing's. So you you were simply hook up the many idiots be Michael to the microphone stand. The Mounties included with both the media, ESPN, the data, nobody a microphone. And you simply use the cable to connect it to your left up or my desktop computer in my ugliness of wires over here and then for them. Unity is be Mike. You simply use the output, as you can see here, the output off the on board sound card, which goes to the Arcia inputs into the template fire and 40 amplifier outputs. You get up to your speaker. 7. 2.2 - Dayton Audio wiring diagram: if you have the second option. So the data Nodia microphone in an audio interface, your set up connections are a bit different. So again you will find a picture file with a graphic, a schematic off, how you connect the it should device the which so you will have Ah, of course, The audio interface and the data Nodia microphone You will need Ah, the USB cable which you will find in the focus Right, The package. This is a USB type B to type A, which you will connect to your computer you will also need This is a dual mono jack plug which you will connect into your audio interface, the outputs and a pair of R c A at the other end, which you will connect to the amplifier. These are full 5.25 millimeter jacks Mano. You'll also need a cable to connect the microphone to the audio interface. This is ah, three pin XLR cable male to female, and I suggest the take a three meter long cable so you have enough slack from the microphone to the audio interface. Now again, you can look at the schematic and figure it all out. But I will also show you how to do it in real time. Again, we have the same set up is before. But now we have the Dayton microphone in place and it's hooked up to the XLR cable, which is three meters long. This is for convenience and this goes all the way to the audio interface. I connected it to the left side input you couldn't connected to this one. But I prefer this one. I either one is okay. And then this is connected to the computer or laptop. And then the outputs are used. The Do Amano jacks into the amplifier with the pair of RCs. And then again, the speaker outputs from the amplifier goes to the speaker. 8. 2.3 - Calibration file: before we actually gettinto the settings of the room. McCue Wizard, I have to talk about the microphone calibration First. Both of the microphones were using are calibrated microphones. This means that they come with a calibration file. You have to understand that when we measure the frequency response, the microphone itself has its own frequency response. And if we go further down the chain, um, the sound card has its own frequency response. So if the microphone doesn't have a perfectly fat response, it will introduce its own sound signature. And then again, if the song card doesn't have a flat response again, it will deteriorate the sound even more so. Our measurement will not be accurate. So what does that mean? Well, a sound card. It's easy to have a flat response. If, ah, if your sound card is not a defective one, odds are excellent that you have a ruler fret frequency response. You can calibrate this the sound card, but that we will not do that in this course. Because first of all, you cannot do that with you, Mike. One. You can, however, do it with the focus, right? But it's usually not the case. because the response is very flat, maybe at the ends of the frequency response band with like 10 hertz or at the other end, at 20 kilohertz, you might find some small deviations from flat response. But then again, we are strictly talking about the microphone calibration. So the question is, does the microphone have a perfectly fret response? Well, it doesn't even if you have ah, very expensive microphone. Your response will be very close to flat, but it will not be perfectly flat. So what does the calibration file do? Well, let's see. The microphone has a flat response up until 200 Hertz, where it has a dip off minus two db, you will find the calibration file that it will say 200 hertz plus two db so it compensates for the defects or the deviation from flat response. So you have the frequency response of the microphone, plus the calibration file. It will equal a perfectly flat response. So each individual microphone comes with its own unique calibration file because no microphone has the same frequency response as the other. So to get your frequency response calibration file, you have to go to the website off the manufacturer. So in the case of the media DSP Mike, you will find a serial number and for the day to know your microphone again. A serial number on the side of the my paper. So if we fire up our trusty browser, we can go to Google. So let's type Mini, Do you speak you like one? And if you scroll down, you will find the calibration file downloads section. So if we introduce our Silien were serial number, which is seven or three toe 166 then we can save it Donors on our desktop and again for the data node your microphone. Let's go to Google and type to date and audio go to test and measurement. And here is our microphone introduced the serial number, which is It's 103 15 Again, this is unique for your microphone. And then we can save our calibration fire for the Dayton audio microphone as well. So now that we figured out what the calibration file does and we saved them to our computer , then we can head down to setting up the room, make you wizard 9. 2.4 - UMIK-1 Room EQ Wizard setup: when we were setting up the you might one or the Dayton Audio with the audio interface. There is a slight difference because the focus right Scarlett has a CEO support. Uh, you have to install 1/3 party application for the U Mike one because it doesn't, and you have to take in consideration. You are using the own board, the sound card as well, or some other external, some car. What this issue application does pretty much at the bypasses all the queuing off the Windows operating system. So you have a clean signal. Open up your browser and type in google dot com and search for a CEO for all. Go to their website and download the latest version. Just click next next, and make sure you check this offline settings and install it. Then the easiest way would be toe type in the surgeon bar is your for all offline settings . Now it may take some time before this application starts up, so please be patient When it opens up, However, just click on the little wrench icon and then de select anything that you are not using. So the scarlet we're not using it so they selected and chick your ah sound card in my case , really take high definition audio because we are using the own board sound card as the output in the input is the you Mike one. So check that out as well and that's it. You can close it up and then we will start the room. Acu wizard as you started this application, just goto preferences. And first of all, we will load the calibration file. So he had the mic and then Rose. And then, ah, I rename the files, so I couldn't know which one is which. So we're gonna load the you Mike calibration file and then we go back to some card. Now we select the ace, your drivers. And now I have a lot of devices here that's poured a show, drivers, but we will select the issue for all. And yes, we will continue using the calibration file. Now, on the output section, we see that we have eight channels now. This is because the on board the sound card is a 7.1 Suncor and I really don't know which one is the correct one. So we'll select the 1st 1 and we will see afterwards which one is correct. And for the import we have to select the you Mike one. Yes, and that's it. I'm just gonna do a test measurement. So we selected our outputs and inputs, and now we're gonna check levels. Now, you have to hear some sound for your speaker. And because we're not hearing any sound that that means that the output is not correct. So we have to select another one. Let's try to I don't know if my work for me speaking up, but I can hear some for the speakers. So I know the Output two is the correct one. Let me increase the sound just a little bit because I think minus 41 is kind of low, and I'm gonna do a taste measurement wars. This is I just did this this measurement to compare it to the Dayton and focus right, the combo. So I'm gonna do just a bit of it's moving and apply window. Don't worry about the sittings will talk about them later on. Now I'm gonna save this measurement and then continue with the Dayton plus focus right audio interface and compare the two 10. 2.5 - Dayton Audio Room EQ Wizard setup: Now when we launch room, make you wizard the set up for the data, nobody, a microphone and the focus right. The audio interface is much more simpler. So go the preferences. Let's load the calibration file for the Dayton audio microphone and then go back to some card. Select the issue drivers and the correct audio interface. Yes, we want to use the calibration and select ah, out with one and input one. We know that the PhoCusWright Scarlett has two outputs into inputs, and that should be enough. Now let's do a taste measurement and a spiel again, Uh, it The tries to calibrate the microphone to, uh, give absolute measurements in decibels, but we're only interested in the relative measurements, So I measure a curve. I measure another curve and compare it between them. It doesn't matter if it's 50 decibels or minus 20. We talked about this in acoustics 11 So we'll continue anyway. Let's check the levels. We don't hear anything. This means the output is incorrect. We can hear sound. No, let's do a measurement. Worry. Let's repeat the steps. Apply this moving and a gate window. Now let's load the measurement. We did with the you might microphone and you can see the difference in amplitude rather the magnitude of the amplitude. This one is way lower, but that is not important. We can, uh, adjusted. So we take this measurement, add like general 50 db, and let's try to overlap them, as you can see, even though I used to completely different devices the measurement almost that match perfectly. We're not going to get a one on one match because there are measurement errors in ever in the equipment, but you can see it is quite decorate. 11. 3.1 - Types of measurement and Far field in detail: Now it's time for some of theory. Before we get into actual measurements, you need toe. Learn some theory because not old measurements are the same. You might be measuring two way sealed speaker or a three way based reflects Speaker. Maybe you have to base drivers. Your port is on the backwards on the front, you have ah, floor standing speaker. Every measurement will be unique and you know to and you need to know the theory so you can adapt to each situation individually. So there are four types of measurements you have, ah, full space or free field. You have ah half space. You have ground plane and near field. Out of all, these four types of measurements were gonna use free field and the near field. Now let's start with the 1st 1 When you do a free field measurement, you are referring basically at outdoor measurement because you have to suspend the speaker somehow in the air so you don't have any boundaries around it. You have to understand that you are trying to measure the speaker and it quickly, so any reflection will corrupt the measurement. To do so, you have to suspend the speaker up in the air somehow, so it doesn't have any boundaries around it. And the microphone captures on Lee the signal generated by the speaker and not any reflections from the ground or any other object. You can do this in two ways. You either suspend the speaker in the air and do a measurement outdoor where you will not have any boundaries, or you measure the speaker in an earthquake chamber, which is either a very big room or it has absorbing material on the walls, and it doesn't reflect the any sound from the walls. Now, uh, I'm guessing you don't have access to an anti quake chamber, and the fact you have to suspend the speaker up in the year is quite difficult and challenging to pull off while it does result in very precise measurements. So what can we do about how can we measure a speaker indoors and also, and it quickly? Well, there is some solutions. First of all, when you measure the speaker inside the room, you will also get the room response with the speaker response. So it's no good. So how can we do this first of all, place your speaker inside the biggest room you have with the least amount of boundaries near the speaker. So if you are measuring a bookshelf speaker, you have to place it on a stand. So you have to place it about halfway from the ceiling and from the floor, because the stealing and the floor are usually the first reflections that appear on the measurement. So now you probably have another question. So if I measure it indoors and I get the reflections from the ceiling in from the floor, how is that any quake? Well, let's take a look at this picture so you have a speaker on the stand and a microphone understand as well, and their place in line and halfway distance from the floor and ceiling. When you are doing the measurement, a signal will play for the speaker. The first wave which will hit the microphone will be the direct wave, and after that, the first reflection which will corrupt our measurement and the reflection which will reach the earliest will be from either the floor or from the ceiling. And if you draw two diagonal lines at halfway point from the direct way of distance, you have the distance travelled by the reflected wave, and if we do, some simple might have. If you place the microphone at one meter distance, you have 0.5 meters at the halfway. And let's say the speaker is placed 1.2 meters away from the floor. So this is the B side and we need to calculate See using the Pythagorean theory. Um, we can calculate it easily, so it's 1.3 meters and if we double that, we know that the reflected wave travels 2.6 meters and the direct wave travels one meter. So after the sound has traveled one meter, it has reached the microphone and the microphone starts picking up the signal. Now this type of sound is truly an equipped because the reflected way from the floor hasn't reached the microphone yet. So now the sound has to travel an additional 1.6 meters before you get the room reflections . And if we take the distance off 1.6 meters and see how long it takes in time to travel this distance, we have the speed of sound, which is 343 meters per second so were 1.6 meters. It takes about four point 65 milliseconds. So basically, when you are doing a measurement in room ACU wizard, you are telling the application to do the following. After you received the first signal. Onley measure the next 4.65 milliseconds and the rest disregarded because it contains the room response. So everything before that is an a quick measurement. No, this seems pretty easy and straightforward, and you might think there's a catch and in the various by doing this type of measurement, this gated free field measurement. The problem is with the low frequency sounds because they have very longer wave lings. Depending on the size of the room, you will not get the full spectrum and it quickly now, in our case, when you have, ah, the speaker 1.2 meters above the floor, so the distance traveled by the sound and quickly is only 1.6 meters. If you divide the speed of sound, which is 343 divided by 1.6, you will get 214 hurts. So this is the lower limit off true and a quick measurements. So in our room with the speaker 1.2 meters above the floor, we have ah, true antiquing Frequency response measurement from 214 hurts all the way up to 20,000 hertz . Now, to solve this problem, you have to employ a near field measurement which we will talk in a bit. So you make this measurement, you make the near field measurement and then combined them together to get a full, uh, frequency band with a response. 12. 3.2 - Near field measurement: Now let's talk about the near field measurement. This type of measurement is used to find out the low frequency response of that particular device under test. In this type of measurement, you place the microphone very close to the speaker and the depending on the size of the speaker. This measurement will be accurate up until a certain point. So if you're measuring, let's say, uh, six inch will, for it'll be accurate about up until 500 her Selous E. But try to use the near field measurement as high is your far field measurement permits. So if your room allows you to make a far field measurement, UH, which is, Ah, accurate. Up until 250 Hertz try to taking consideration. The near field measurement on Lee up until 300. So 300 below, even though it might seem accurate, up until 500 hurts more on that later. Now, because the microphone is so close to the speaker, it doesn't pick up any effects from the baffle. So any baffle diffraction or any room reflections that my appear the downside is it's not accurate in the upper frequencies now, depending on the size of the speaker. If the speaker is larger, your problem you are permitted toe. Place it a bit for their way from the speaker. But in most cases, just place it as close as possible to the speaker without the speaker diaphragm actually touching the microphone when you pass the measurements single for the through the speaker and when you're measuring a port, for example, plays the microphone flush with the inside of the port right in the centre. Don't stick the microphone inside the port, so when we are measuring near field, you have to take into consideration each base driver. So let's see if you have a base reflect speaker, you will have to do a near field measurement of the speaker and also for the port you. Because the port will have a different size compared to the speaker, you will have the you will have to scale the port according to the size of the speaker in the just the level. Then you have to combined the two responses, and after that you have to scale it too far field, because when the microphone is really close to the speaker, you can imagine the sound is way louder, so have to scale it so that the microphone is about one meter away, exactly like in the far field measurement. Now I'm going through the steps don't stress too much. It will be more clear when we actually do the measurement. So when you measure like a two base driver speaker in with the baseball export, you can measure only one speaker near field and apply 60 Beat to death curve. Or you can measure each individual speaker and add them together, but you will get the same result. Then again, measure the base reflects separately, scale it, combine it with the speaker response and then scale it too far field. Now, after you've done the near field measurement and the far field measurement, you stepped them up together and spliced them in the part where they overlap. So let's say you have the far field measurement, which is accurate from 250 toe 20,000 hertz and the near feel measurement, which is accurate from 20 hertz toe 500 hertz. You have the span from the span from the 250 hertz all the way to 500 where you can spice the responses and then you get a full frequency response, and it quickly 13. 3.3 - Half space and ground plane: Now let's also talk about the half space measurement and the ground plane measurement. We're not going to use these types of measurement in our course, but I'm going to tell you something about them. So you know that they exist in. You might implement them if you want or not. So, firstly, let's talk about the half space mission. This type of measurement has the same boundaries limitation as the anti quick measurement except for the floor to measure the speaker frequency response. Using this method, you will have to go outdoors. You have to place a speaker on a baffle and dig a hole in the ground. Make sure that the baffle is perfectly flush with the ground. By doing so, you have on unlimited vertical space and the only obstacles you need to care about are those on the horizontal plane. Depending on how wide the frequency response we are measuring, make sure you don't have any obstructions on the ground for the appropriate distance. This type of measurement is called half space or two pi because the speaker radiating 100 and 80 degrees hemisphere now, if we switch over to the ground plane measurement, this type of measurement is done by placing the speaker onto hard reflective surface and the microphone on the ground as well. Usually, this is done in a parking lot with an asphalt or cement floor. This ensures the reflective surface and no nearby boundaries for at least nine meters foran accurate low frequency measures. By doing this, the measurement includes the sonic mirror image off the speaker. For this reason, the microphone placement is on the ground at the apex of the direct signal, and it's reflected sound image to perform. This type of measurement plays the microphone two meters away from the speaker. This is because the sound picked up by the microphone contains the sound of the speaker, plus the sound of the mirror image, which adds up to a plus 60 be game for every doubling of the distance from the speaker, the sound loses 60 be so by placing the microphone two meters away and not one meter, it compensates for the extra game 14. 4.1 - 2-way speaker measurement | Presenting the setup: Now let's talk about the set up a bit so you can see the microphone, Understand? Which is placed the one meter away from the bookshelf speaker. Understand? Of course. And you can see I have some stuff lying around you. You probably have the same at your home. So what is important to know is that in front of the speaker, you have nothing to reflect the sound off. So the fact that I have something over there doesn't interfere with the earliest reflected wave. So the earliest way will be bouncing from the floor. So the sound traveled to that object and then to be picked up by the microphone will come at, um, or later time after the earliest reflection. And we know that after the earliest reflection, with disregard everything because we're interested Onley in the Anna quick response. What concerns the back of the speaker? Uh, you have to place it not close to the wall. But it's not that you don't have to place it too very far from the back wall. You have to imagine that we calculated that earliest reflection will take two point six meters off travel. So if you can imagine the sound traveling from this front of the speaker to the back wall and then back to the front and the microphone. It has to travel like so we have one meter. We have to travel like 0.8 meters to the nearest boundary, then 0.8 meters back. So we have 1.6 meters and then another one meter to the microphone. So we have 2.6 meters the same time as the earliest reflection. So if you place it closer than 0.8 meters, toe the back wall, you will have some problems, because your, uh, time window with no reflections will be smaller. So you have to take that consideration when you taking account the back boundary. Now the microphone is hooked up to another audio interface because I had some problems when using the focus right interface and this one I used to power up the Michael, which I'm speaking on now, and from now on, I'm going to use it for the measurements as well. So you're going to see another device in room reserved but is basically the same thing. Now let's look on how you can wire up the speakers so the speakers are placed inside the enclosure and you have to somehow wire them to the outside to measure them so you can use the base reflects hole to rob a cable from the speaker to your amplifier. I just used some cheaper elevator clips to easily hook up this. The wires and the fancy way to do it is to use a by amplification binding posts so you can hook up the tweeter here in the mid base here, and you have so much freedom to taste your crossover and such, but not always you will use by amplification binding post, and you will have just one pair. So you can, for example, hook the tweeter toe a pair of binding posts and wire the other through the base reflect sport, or where both of them for the base reflects sport. It's, uh, however you choose on this example. I wired the meat base for the base, reflect sport, and the Twitter is connected to these terminals over here 15. 4.2 - Measuring distances: now for the next step, you will need a measuring tape and the first thing we will do is measure the distance from the floor to the ceiling. I already did that and it's 2.5 meters. So we need to set the speaker stand toe a certain height which is half the distance between the floor and the ceiling. So we need to set it at 1.25 meters. Now, of course, the speaker has a certain size, so I like to take the point of reference. The point between the mid bass driver and the Twitter. Most of the most of usually people take the point of reference that the treater if it's a three week speaker, I like to take the reference to the mid range driver it if it's between the Twitter in the mid maze again, most of the guys usually like to take the Twitter as well, even for the three way design. Now let's take this point of reference and measure from the floor and set your speaker stand accordingly so it matches the half distance from the floor to the ceiling. So 1.25 meters same thing from the microphone. Set up your stand so that the so that the microphone is 1.25 meters away from the floor, perfect, and then measure the distance from the speaker to the microphone and check if it's exactly one meter. 16. 4.3.1 REW Interface Update: now, Another thing that changed in the past two years is that, um a room a que wizard has bean updated quite a few times. And the thing that stands up the most is the interface when you make a measurement, So when you click the measure button, you will be greeted with this new interface. And actually, if you look at the old interface because that is what you will you will see in this course because I cannot update all the video files with the new interface, you will have to make these associations because they are the same buttons. But they are in different placements, so you can see the SPL in the impedance buttons at the top left. But now the start frequency in the end frequency and the level which is, uh, off interest, it's in different produced position. Now, I don't know why they changed it. You also have the length of the sea, the sequence and how many sweeps here in here. So you have to be aware of the new position. You also have the output and airports now below. And of course, the three buttons that check levels. The start measuring and the cancel button. So basically it's the same thing, but uh, in different positions, and this is going to be a bit of an inconvenience, but that I'm sure you will get over it quite easily. 17. 4.3.2 - Far field midbass: since we finished setting everything up. Now we can actually take a measurement. So let's start up a room, make you wizard and all you have to do is click the measure button. Here you will select the frequency spam you will be measuring. So from 20 hertz to 20 kilohertz. If, however, you are measuring the tweeter, you are advice to go a bit higher on the lower part because you might damage the Tweeter. Here is the level. So when you check the levels, if it's too low, if it's too low, you might increase it. And here is the length of the measurements. So the higher number will be a measurement which will take longer, but it will be more accurate. So 2 56 key will be sufficient. So this is the number of sweeps. One is okay, so let's check the levels again. Everything is fine. And before I clicked the start measuring button, I want to talk about what we're actually measuring. So when you hit the start measuring, but you are actually measuring the impulse response. So this is gonna be a bit difficult to explain what the impulse responses, but we're measuring it in. Ah, we're extracting from the impose response all that we need. So we extract the using various equations and mathematical formulas. We expect the frequency response, the phase response, distortion and all kinds of measurements. So instead of me explaining what the impulse response is, I want to show you the graph off the impulse response after measuring it and what we can deduce from. So let's start measuring. Or that was the measuring sweep in. Uh, usually, you start off with this window, but you can click the impulse. And here is how the impulse response looks like pretty cool, right? So the time zero here we have the time scale measured in milliseconds and here in microseconds. So here you have the time zero time zero coin sides with when the first sound wave hits the microphone. So before that, you have a whole lot of nothing. So you play the sound for the speaker, and there is nothing to record until the first sound reaches the microphone. And this is the 0.0 in time, and it will look like a large spike in the impulse response. Then you will have a lot of squiggly lines, which represent the first sound wave decaying. And somewhere along the line you will see another rise in impulse, which represents the first room reflection. So if you remembered, we calculated the first room reflection to be at 4.6 milliseconds, and I advise you to take the gate window right when it starts to rise. So at 4.4 milliseconds, so if you switch toe SPL, we can see the frequency response. Let's smooth it out a bill because we are not understanding anything from this. So this is the frequency response off the mid based picker. But because we have not applied any gate window to the impulse response, it actually contains the room response as well. So you will see a lot of a lot of peaks and dips and valleys, because when the mic picks up the room reflections as well, it will create the scenarios where sound waves cancel each other out or reinforce each other. Here, you can see a huge cancellation at around 87 hurts now to exclude the room reflections. We look at the impulse response, and we are interested only from time zero to 4.4 milliseconds. So let's go to your spiel and apply er windows. ER stands for impulse response, and we are interested in the right window. So from time zero, which is this the reference time to the right, 4.4 milliseconds, an inter four point for and use common not dot because I don't know. It doesn't work that well. Let's go back toe SPL and see what happens when we play the window. So let's hit the play. And as you can see, this is the response of the speaker without the room. But as we know when we make far field gated measurement, the low frequency response is not the representation of reality. So then let's look back to the impulse response, and if if we apply a left window when you're applying, the left window never used zero because you will ruin the measurement. Instead, go for number, which is before any deviation from zero. So 0.3. This is 300 microseconds, so 0.3 million seconds. So use common zero comma three zero comma, three milliseconds and let's see what happens when we apply this window. It cuts off all the low frequency response because it's not accurate. So the response is accurate on Lee from 213 hurts and above. So this is the response of the mid bass driver majored Fart Field. We are still missing the low frequencies, but we will have to do a near field measurement to have the bass notes as well in this chart. So let's rename this measurement too far field, mid base. 18. 4.4 - Far field tweeter: Now let's do a far field measurement off the Twitter, so the microphone is left in the same position. The speaker in the same position. We only have to connect the Tweeter to the amplifier. So, like I said, I used the top binding posts to hook up the Tweeter. And then we seem to do a measurement. So let's get back to Roma Que wizard his measure now because we are measuring the Tweeter, you have to be a bit careful. So let's go for 350 Hertz because anything below it's not label for a tweeter, anyway. Solar start measuring. That's a play that's moving, and next we shoot applied the Let's check the impulse response first So we see the time zero, which is the first wave hitting the microphone. In here, you can see a large spike, which corresponds to the first room reflection. Now the squiggly lines 0.6 seconds. So let's a play impose response window off zero point six and 4.4. Great. As you can see, it already highlighted the part we are interested in, so go to SPL apply windows rate. Now we finished measuring both to the speaker's far field. Now this is actually the complete response of the treater because we are not interested in the lower frequencies. And we don't do near field measurements for the Twitter anyway. So to get a complete the frequency response bandwidth for the mid base, we have to do the near field measurement first. 19. 4.5 - Near field midbass: So now we have to do the near field measurement of the speaker, the mid bass speaker. So place the microphone very close to the speaker, right in the center of it. So you have to take it as close as you can to the speaker. But you have to take into consideration that when you are playing the measurement signal, the speaker will move. So you have to make sure the speaker doesn't touches the microphone when you are playing the measurement signal. Now let's do the measurement. So now we're measuring from 20 hertz to 20 kilohertz because we're interested in the low octaves now. But since the microphone is so close to the speaker, you have to imagine that output level is very high now. So why turn down the volume knob and, uh, back down the D. B's over here as well? So you're probably not gonna hear much picking up from the microphone in which I'm speaking . But you can see the level is okay, so let's start measuring. Let's smooth this out and let's unsure of the far field measurements. So this is it. We're not gonna do anything to this response yet. There is no need to apply any window. And there is, ah, some particularity for this type of measurement when the speaker is in a bass reflex box. So as we learn from acoustics one a one when ah speaker is placed in the base for flex box and it reaches the resonant frequency off the box, the speaker barely moves. So the port takes old energy away from the speaker. It's at its maximum output. Speaking about the port and in the same time the speaker barely moves. So when we're looking at the frequency response chart, the fact that the speaker isn't moving it shows up as a dip into the response. So you can see the deep over here. And this tells us that the resonant frequency off the box is at 51 hurts. There are other ways to measure the resonant frequency off the box, like from the impedance charts. But, uh, the acoustic measurement, the one with the microphone. The near field measurement is the most accurate in determining the resonant frequency off the box. So let's rename this into near field speaker, and now we're gonna measure the ah port and place the microphone near field as well 20. 4.6 - Near field port: Now we have to do a near field measurement of the port. So switch the speaker around and, uh, place the microphone in the center of the port. But don't stick it in. If the Portis flared, maybe go inside just a little bit. So after the positioning is complete, we can do the measurement. So using the same sittings as we did the an earful measurement off the speaker, it's important not to change them. So go ahead and start measuring. So let's apply the smoothing. As you can see, the maximum output of the port corresponds with the dip in the response of the speaker. Here we have, ah, awkward anomaly. I don't know what this is. This might suggest inside residence of the Bucks, because it's a easily picked up through the port. But the right now we're not concerned with the troubleshooting enclosure problems. I'm just trying to show you how to measure the speakers. So now we finished the near feel measurement of the port and of the speaker, and we have the far field measurement of the mid base and the treater. And now a we have to do is do some processing and some calculations. Let's rename this toe near field poor and let's get the work 21. 4.7 - Error, heads up!: So I just realized I made the measurement error. So it's nice that I did, because I can give you a heads up so you don't make it as well. So it's important to make all the far field measurements at the same level. So your amplifier, your audio interface, the volume from room ACU wizard needs to be the same for all the far field measurements. So in our case, we did the Twitter in the mid base. They all have to be the same, and the same goes for the near field measurements. So when you're making near field measurements, you can fiddle with the settings because we're going to scale it so far field anyway. But all the measurements which are made using the near field technique must have the same level. So in my case, what I did wrong is when I measured the Twitter, I just moved the volume up from the amplifier and they were not made in the same live. So if you can look at this chart if he closed down the near field measurements, we can see that the Twitter is blower in level than the mid bass, and this almost never happens. So I know that I did something wrong over here. So instead of doing the measurement again because we're not gonna do anything with these measurements, I'm just showing you how to do them. I'm just gonna do a small correction. So we take the Twitter and increase the output, and I would looks like more like a a normal measurement where the output is a bit higher. It's not always this, but I know in this case it's the Twitter is a bit hiring output than the mid base. So again, make sure when you do the far field measurements off each individual driver that you do them at the same level. When you switch the near field, you can change Thea output because the microphone is really close to the speaker and then do all the near Philly measurements at the same level as the first near full measure. 22. 4.8.1 REW Interface update 2: So since this course has been released a couple of years ago, Roma que wizard has gone through numerous updates until now. And ah, the functionality hasn't changed. But the the interface has changed slightly. So you will see buttons in different positions and you will have to take different actions . So normally, when I am, apply some functions to the frequency response grafs You will see me click this button over here. The control button, this little cog wheel and we have different options over here. Now, in the older version, you have, ah, along the functionality readily available. Now we have to click some buttons so this moving is already here. So if you want to apply some smoothing, just do it from here directly. But, uh, when we Well, for example, if we want to modify the SPL off a certain curve, we have to click this measurement actions button. And now you can select the curve and add or six course obstruct the the decibel rating, and then you can add offset to data. And if you want to toe add two curves, we will do this with ah near field measurement, not the far field measurement like in this example, and we can add them together like this. We select the two curves we want to add to get her and select a Polsby, and then we generate and ah, we will also use the function to merge, had a certain frequency spot. So you have to take into consideration that it will be different in the lessons which are about to follow. But you have to do this extra step, so press these two buttons when we need to. 23. 4.8.2 - Combine midbass + port: Now that we finished making the near field measurement off the port in of the speaker, we have toe add them up together, but before we add them up, we need to scale the response of the port. Now, because the diameter of the port is usually smaller than the diameter of the speaker, you need to scale it up to size before you can add it up with the response off the off the speaker. So to do this, we need to apply Ah, small mathematical formula. And we need to know the area of the speaker in the area off the port and the theory of the speaker. We can easily find it in the specification shit of the speaker and it's ah, 136 square centimeters. And for the area of the port, we know that the diameter is 6.2 centimeters and ah, the area is pi times the radius squared, so 3.14 times 3.1 squared, which is equal to 30.17 square centimeters. Now let's bring up the calculator and we have ah, 30 points 17 which is the area off the port, and we divide this by 136 which is the area of the speaker, and we get this number and then we extract the square root, then apply luxury and multiplied by 20 and we get minus 6.5. So we need to adjust the level off the near field measurement of the port by minus 6.5 decibels. So we go to controls and the select near field port and add minus 6.5 decibels. Now you can see the curve has been lowered by the amount with just entered and tow. Add up the two curves. We again go to controls and select Neil Field speaker in near Field Port and then a plus B . So we add the two curves to get her and let's generate the new curve. And, as you can see, here is the new curve generated. Let's close up the far field measurements and the near field wounds, and here is the response off the port, plus the speaker. Now, this is the time where we need to add the function I talked about earlier in the course and this Ah, software, the room ACU wizard doesn't have, and in the next lesson, I will show you how to apply dysfunction. I'm talking about 24. 4.9 - Box diffraction for near field measurement: Okay, so now we finished adding up the near field measurement off the port with the near field measurement of the speaker. So let's rename this speaker plus board near now. The problem with this response is that when you make a near field measurement, since the microphone is so close to the speaker, it, um it simulates on infinite baffle. Let me give you an analogy. So let's say you are looking and a wall from a distance. So you're you're 10 meters away from a wall and you can see it. Let's say it's two meters by two meters. Now if you go near it and put your face right on it and I ask you what is the dimension of the wall? And you see, it basically extends into all directions because you do not have the perspective and you don't know the size of it, but you feel like it's going into the old directions. This is somewhat what happens with the microphone, So since it's so close to the speaker, all the frequencies get boosted by six decibels. The infinite baffle effect right? But in reality the enclosure is finite. It has certain dimensions and on Lee. Some frequencies get boosted and some not especially the lower ones. So we need to compensate for this. That's why I devised the Excel spreadsheet. So what we're going to do, we're going to export this measurement into a text file, so just go to file. Export export measurement has text. Ah, select all these. What's important is that you choose the export text the limiter tab. So it has the frequency, the phase in the SPL value. Right? So between each of these, you have tab and not space. So this is important. Now select. Okay, give it a name near field Save now India's folder. I have the measurement and the Excel spreadsheet. Now open up the file. You can delete all this and then press control A to select everything and control. See to copy the values. Then you can close it up. Then open up the excess. Appreciate it might take it well, because it's a big file. Then you based the values here control V enter the dimensions of the baffle. So the height 32 centimeters by 20 centimeters and then you get all the values here Now to select all the values easily. Genest, Scroll, Scroll to the end. Click here, press control and shifted the same time. Then right there, Up arrow. Let go off control, but still hold shift and then down Arrow and then copy control. See? Then go back to our text document Control a control V and save. Now we go back to the room, make your wizard import import frequency response and select the newly created file. And you can see the lower frequencies are at a lower amplitude compared to the initial response. And that is to be expected because here it's like a simulated infinite baffle. So old frequency get boosted by six decibels, but according to the dimensions of the bear full not all of them. And it should progressively lose amplitude. As you can see the overlap as soon as the actual dimensions of the baffled do increase the output by six decibels and that happens starting with 650 hearts 25. 4.10 - Splice near field with far field: now to get the full frequency response of the mid based driver, we have to merge the near field measurement with the far field mission. Let's rename this correct near field and let's take the far field Mint base. So we need to combine these curves to combine them. We need to first match their levels. Now, if we look at the graph lines, they pretty much overlap. So this is, ah, fortunate coincidence. But normally they have quite a large visible difference between them. So, um, how do we match their levels? So we know that the near field measurement is accurate up until, like, uh, 500 hertz, maybe even more. But it say it's better toe keep the near field measurement accuracy as low as possible. So let's say 500 hertz. And we know that the far field measurement is accurate from 214 and up. So we have a portion where they overlap in ah, in accuracy. So to see so between 214 and 500 hertz, they should look similar in ah, in amplitude in frequency response. So we tried toe overlap the graphs in somewhere in that area usually in one point we choose one point where we were. We like them to match. And after we did that, then we can splice the two responses together. So let's, ah, always always adjust the response off the near field. Stay away from the far field. Fire field needs to have a fixed response because it has to have the same response as the Twitter. So don't move that about You have to move the response of the near field, so go to the near field. So we renamed it the correct near field and adjust the level until you have a matching response so we can see they match in a lot of points in, ah, the area where they share accuracy. So this seems about right. So now we have to merge them together. So select the correct near field in a point. A. You select the part which has the low frequencies and important be you select the one which has the upper frequencies. So here we sit at the far field, mid base and select merge B to A. I think it's the other way around. Are for mid bass incorrect near film? Yeah, so Let's emerge them at? I don't know. Here. 346 hertz. 346. Let's generate. So if we untech this and this there we have it made base frequency. 26. 4.11 - Create FRD files: now that would be generated the full frequency response of the mid bass. And we also have the full frequency response of the Tweeter because it only needs one measurement, one far field measurement we can generate. If our defiles for each of these drivers, the effort, the files, you can use them late around in crossover design applications. While this course is not about designing crossovers, you can load them into such applications and have a taste on how to design crossovers even though you might not be knowing what you're doing. So let's secrete the files. We have selected the mid base, so go to export in the export measurement. That's text just like before. Okay, okay. Made base. And let's say the Twitter export measurement. Okay, Tweeter, when you go to the folder, we received the files. You need to modify the extension now to see the extension. Just do this. Go to view and, uh, file NYT of file name extensions. Now the extensions are visible and simply press F two to rename the file and type Frd. Yes, I want to change it. And the same for the Midwest. Perfect. Now we can load these files into Exim, for example, 27. 5.1 - 3-way speaker measurement | Presenting the setup: Now we stand for a more complicated project where we will measure floor standing speaker, and this is a three way design with a dual wolf for and a base reflect sport. This is one of the more complicated designs, so if you understand this one, surely you will manage with the more simpler ones. So again we have, the microphone said at one meter distance. And the reference point is, Ah, I taken the reference point somewhere above the mid range. Like I said in a three way design, I like to take the mid range driver as a reference point. Usually people take the tweeter because that is where usually the Twitter is in line with the year of the listener. But either one is fine, so in this case we want to measure it each driver. So for a crossover design, we want to measure the tweeter, the mid range in the base and the base side. So for the base, I robbed the cables through the base, tripling sport and for the mid range in for the twitter. If you go on the other end, you can see here a different type of binding posts and before placing the binding posts, you can rob the cables. This one is the mid range in this one is the tweeter both wires and you have to place some silicone, some silicon cock. So this is a perfect seal. After you finished designing the crossover, then you can take the wires out and place the binding posts as as normal. Another particularity for this type of speaker is that the midrange driver uses a face blood. This this is a Facebook, so we will see if you want to do a near field measurement. Uh, it might be a bit trick. 28. 5.2 - Describing the measurement process: we're going to switch it up. Toe amore complicated design. So this is a three way with a tweeter, a mid range driver, too subwoofers and a base here for export. So the purpose off measuring each individual driver is to create if our defiles for the highs, mids and lows to design a three way crossover. So if we look at our speaker so the tweeter represents the highs. The mid range represents the mids, and the two sub workforce, plus the port represent the lows. So when we do the actual measurement, I don't want any confusion on how to set up the levels. So when you are doing the far field measurements, you have to have the same levels on your amplifier. On your audio interface, the output level on roomie que wizard all have to be the same. Because when you introduce them into your application for designing crossovers, you have to adjust the levels so that each speaker response blends with each other and create on a world lena response. So if the Twitter is hiring output than the mid range, you have to tame it down a bit. But if you fiddle around with the volume on the amplifier and measured the Twitter on a level and the mid ridge on some other output level. Then you will have errors in your measurements. So again, for the Tweeter, for the mid range and for the base drivers, all far field levels must be the same. Now, when we switch it upto the near field measurement, we can changed the output level. Why is that? Because we have to scale the response toe far field anyway. So even if we change the output, it won't do any wrong because we have to take that near field measurement and decrease the output. So it matches the far field scenario. Now, when we need the measurements for the two way design, I told you that all near field measurements have to have the same levels. Why did I say that? Because when you combine the response off the wolf for with the response of the port, they both have to have the same measurement output, so you can add them up. Otherwise, the response the some The response is not correct. However, in this case, since we have on additional driver the mid range, you can have a totally different output level for the near field measurement when compared to the near field measurement off the wall for because you are not adding any near field, the responses so for their wool, for you make a near field measurement off the wolf for and of the port, and you add them up together so they have to have the same output. So the some the response is correct. But when considering the midrange speaker, it's just one speaker. The output is irrelevant with the measurement output off the wolfers, so as a conclusion or far field output levels must be the same. When you're doing the measurements and when you are doing the near field measurements, you you can have another outward level. But the near field, the measurement off the wolf for and off the off the port must have the same output level because you add them up together in the mid range can have any near field output. It doesn't matter. Some like to have the same levels for old the measurements, including the near field and the far field. But if you do that, you have tow start with near field measurement because that is the loudest. Otherwise, if you start with the far field and then go to the and measure near field, you will clip the microphone. So you have to start with your loudest speaker and do and start with the near field measurement. Usually the loudest near field is the mid range driver and then the port. The port is usually louder than the wolf for itself, and the mid range is louder than the port. So if you want to keep the same output levels for all the measurements were doing toe complete system, start with the near field measurement and with the loudest driver, usually the mid range or the port. I tried to explain how this works so you can understand the big picture because you will have your own scenario with your own speaker, with who knows how many drivers or which design you are choosing. And you have to make the correct choice depending on your design. So in our case, we're going to do far field measurement for each individual speaker so far. Field, tweeter, Far field, mid range and far Field wolf for then we're going to switch it up toe near field in change . The output level measured the mid range. Then we are going to measure one will for and we're going to add six decibels. So we're not going toe measure each individual for you could do that. You will get the same results. So we're measuring the new field of the wolf for and for the port. And then we're going to do our little processing where we merge the near field with the far field responses and in the end, we have to have three FRD files for the highs for the mids and for the lows. That is what we are aiming for. 29. 5.3 - Raw speaker measurements: first of what we're gonna hook up the wolf for. It's advice to start with the wool for first, because when you are sitting the levels of your amplifier and in rooming que wizard if you have ah, high enough output. If you start with the Twitter, for example, you might burn it up. So it's always safer to start with the wolf for so now the wolf, or is hooked up to the AMP microphone is placed in the far field position. So let's, uh, check the levels. You might go a little bit higher and let's do a measurement. Or now let's go all the measurements first and then switch toe processing. So let's rename them so we know which one is which. Far field face. Let's go ahead and connect the Tweeter. Now we hooked up the Twitter. The microphone is still in the far field position. All the settings remain. The same hit measure increased this frequency so we don't do any damage to the Twitter. Let's go toe 300 hertz. Start measuring. Let's rename this toe far field return. Now let's go and hook up the mid range driver. Now the mid range is hooked up to the AMP so my performance. Same position, same settings. Let's go. The measurement measure now from 20 to 20,000 hertz. Start measuring. Let's rename this toe Far field, mid Range, and now we have to switch it up to the near field measurements. Now let's put the microphone in the near field, the position for measuring the mid range driver. Now there's a certain particularity about this driver because it has a face blood design. Now, when you employed the near field measurement, you have to place the microphone in the center of the speaker. But because this driver uses a face plug, the face plug is not part off the actual cone. So, uh, the distance between the microphone and the actual cone of the speaker is pretty is pretty large compared to the size of the speaker. If this would have been a 12 inch were for, you could have placed the microphone in the center of the speaker because the distance between the microphone and the cone isn't that large compared to the overall size of the speaker. But because this driver is just a for each driver, you have to place the microphone closer to the comb. As a result, we have to place it just near the face block and closer to the cone, just like you see in the pictures and after the positioning is complete, then we have to check the lifts. So let's go and measure now. Obviously, the level will be way high, so let's, uh, decrease the level. Let's check the levels. This might be a bit high bullets to a measurement. So let's rename this near field mid range. And now let's hook up the wolf for and change the microphone position. Now we hooked up the wolf for to the amplifier, and the microphone is in the near field position. So in the center of the war, for very close, the speaker just the usual one thing that I would like to mention here, since we have to wolfers in the base triplex port on the same front, baffled. When you do the near field measurement, do it on the top most speaker. Since you have the speakers so close to one another, you might the pick up the signal from the second speaker and which is most likely the case to pick up the signal from the port. So, uh, measured the top most speaker So which is the furthest away from the port? Sometimes you get the near field measurement corrupted by the sound off the port. And usually when you do that, you will see your near field measurement without that deep in the response, which corresponds with the resonant frequency off the enclosure. Since you are picking up the Portnoy's, it compensates for that dip, so you understand why it doesn't appear on the graph. If this does happen, you have to do something to separate the speakers. Maybe put a panel between the speakers or between the speaker and the port. Maybe use some pillows or something like that. Well, let's do the measure. We are going to use the same settings as we did with the mid range. It is not mandatory. It is mandatory when we do the near field measurement of the port that we use the same settings we did when we measured the wolf for. So let's start measuring with the same settings as you can see. We have the dip over here. This enclosure is tuned to 24 hertz, so the measurement is okay. Now let's which to measure the port year near Field Wolf for And now for the measurement off the port, the near field measurement of the port. You have to get a bit creative with how you place the microphone and what is Ah, super nice about the minute DSP microphone is that it comes with the this little stand, and I used to some ice Texan books and use this tripod many tripod to place the microphone right in front of the port. Now, since the pork has a big flare, place the microphone just a bit inside the port like halfway through the flares radius. And after the microphone is in position, we can do the measurement. So let's hit measure now. This is important to have the same settings as the measurement off the wolf for near field , and indeed it is. Now, let's start measuring and let's rename this to Near Field fourth. Now we have all the measurements made. Now all we gotta do is do some processing and create the three files, which we need 30. 5.4 - Processing of the measurements: So now we have all the measurements in place. So let's begin the processing first. Let's do some smoothing so we can understand anything from these graphs. Much better. Now. Let's start with the far field measurements and let's apply the windows. So let's, uh, on take the near field measurements because the graph chart is too cluttered. So let's leave just the base and check the impulse response. Now you have to understand that here we have a shorter window. This is because in our case, we have to wolf for drivers and the port and the port is very close to the floor. Since we said the microphone to the mid range level, the port is quite close to the floor. So the we will have on earlier reflection compared toa the mid range or the Tweeter, and we have to check the impulse response for that. So, uh, the the first reflection is this one. We can clearly see it, so our window should be 3.3 milliseconds and the left side and let's go for 0.4. So now here, windows zero comma four and three comma. Three. Apply windows. Now, this is our response. Our far field gated response for the base. Now let's switch the tweeter far field treater. Go to impulse. You can see characteristics for the high frequency drivers they have really high. Ah, spike in the impulse and the ah reflections is also very clear, so we can see the windows is much larger. So now we have 4.3 milliseconds gated the measurement and on the left side, let's go for 0.6. So there's the all SPL window 0.6 and I just forgot 4.3 4.3. There we go. This is the treater and this is the final response. Now let's go for the far field. Mid range. Let's check the impulse response. As you can see, the first reflection is later this. This is because we place it right in the middle of the room, the mid range driver and on this side we have 0.4 and 4.6. Now, again, this left window is not really mandatory. If you want to leave that four response even though it's not decorate, you might as well leave it on now. We finished fiddling with the far field responses. Now that we have finished with the far field measurements, let's start feeling with the near field measurements and let's begin with the midrange speaker. So let's close up all the far, field the curves and take the near field measurement off the mid bridge. Now all we have to do to this measure wound is applied the Bethel Diffraction. So we have to use the Excel spreadsheet. So let's export the measurement. Okay, mid range your field save now open the text file. Delete this control A to select everything control. See to copy, then close this. Open the Excel spreadsheet based everything now enter the dimensions of the baffle, which are 108 centimeters told by 27.5 centimeters. Now let's scroll down to the and off these values. Click here. Control shift right up. Let's go on control, but still holds. Shift down and copy the values Control. See? No, He has maintained the information to the clipboard. Then go to mid range near field control A to select everything. Chicken deleted and control V paste. Save the file. Then go back to room acu wizard file import frequency response and mid range near field and you can see we lost the output for the low frequencies. And now we have to splice the two responses. So let's uncheck the near field mid range and maybe rename this near field may range the French and let's ah, check for field mid range, and now we have to a splice them into the correct location. So somewhere just above where the far field measurement starts being accurate. So anything about above 200 hertz, so anything in this area will be great. So now let's modified the output off the near field measurement. I emphasize this. Do not touch the far field measurement, so let's go to near filled me range diffraction and try to adjust the level until we overlap in the recently mentioned area. So as you can see the overlap nicely over here, so let's splice them together. We got the far field, mid range and the near field mid range with depression, and we go merge. B two a. Let's say, at 270 hurts 270 hurts. Let's generate Let's take these two responses, and this is the full frequency response off the mid range driver. So let's rename this the mid range complete. And now let's switch over to the wall for so in God than your field before in the near field port. So the first thing we gotta do is we got to start with the port and we have to scale the response of the port toe the actual size of the wolf for And since we have to wolfers the actual area off the wall for is the added area off the two wolfers. So to scale the response, let's bring up the calculator and we have the area off the port, which we know to be 17.5 squared centimeters, and we divide this by two times the area off the speaker. The area off the speaker is 208 0.7, and we multiply that by two and we get 417.4 square centimetres. So we divide this by 417.4, and the result way apply a square root, then la Guerry and multiply by 20 and the result is minus seven point 25 decibels. So we go to the near field port, and we scaled it by minus seven point to decibels and we add, offset to debt. Then we have to switch over to the the near field response off the speaker because we measured one speaker and we actually have to. So in this case, you either measure each speaker individually and have two responses and add those responses together or the way we are going to do it. We are going to measure just one speaker near Field and add six decibels to that curve because we will get the same result. So let's, uh, switch over to near Field Wolf for and add six decibels. Now all we have to do is add these two responses together and let's do just debt. So we have the near field well for and the near field work, and we generate this new curve. Now all we have to do is, um, let's rename this Toe Wolf for plus Port, and now we have to export Theis Measurement export measurement this text and let's rename this to work for Plus. Now let's open the file, deleted the first part. Select everything. Control a and copy control. See, now let's open up the Excel spreadsheet based on the Values Control V. Then enter the Baffle Dimensions 108th and 27.5. No, let's copy these values. Let's go to the end page down. Click here. Control shift right there. Over a barrel. Let go of control and down control. See, to copy, close this file. Don't save. And yes, I want to go to play board and then pays the values over the old values. So control a dilly debriefing Control V. Now let's save this safe and close the file. Then head back to Room e que wizard and import, then newly created file. And as we can see, the measurement looks correct. So this is the near field Well, for plus sport and this response, we gotta spice it with the far field response. So we have the far field based response and we need to supply sit. Some were just above where the far field response starts to become accurate. So in our case above ah, 300 between 304 100 hertz should be fine. So now we have to adjust the level off the near field measurement. Don't touch the far field measurement. So We're just the near field measurement level until it matches in the designated area. So let's, uh, choose our near field measurement and for just the level. Now. This seems about right if we splice it in this point at 349 hearts. So let's choose for a field base in the near field measurement and select merge. B two A In the 0.349 hearts. It's generate the new curve. Well, it's on Chick the far field in the near field and this is the base complete. Now we have all the curves for each individual drivers. So we have the base we have Ah, the mid range and the far field. Twitter is, uh, already done. So let's export these files. Export. Yes. Next. Okay, this is treater. Save now. Let's go for the mid range and let's export the file. It's text. Okay, this is the mid range save. And now let's go for the base driver and export file as well. So hit export export measurement. This text okay, its name, this base. And now we find the our text files and we have to rename them into if our defiles. So let's go to view and change extensions into if our deep is they want to change it and do the same for full of your files. Now we have three FRD files for the Twitter, the midrange and bass driver, and we can use these files, toe a, blow them into exim and then design a cross over from there. 31. 6.1 - Speaker full frequency response - anechoic: since we learn how to make frequency response measurements for each individual driver in create If rd files for these drivers now it's time to do a frequency response measurement for a complete enclosure. So after you finished your crossover, you want to measure you're a loudspeaker or you have ah, on aftermarket loudspeaker and you want to measure the frequency response and it quickly. So in the same fashion, we have to do ah, far field measurement. And then we're going to do a near field measurement. And since ah, the speaker I chosen for this experiment is, um, stood your monitor and it has a base triplex enclosure. We have to do to near field measurements one for the wolf for and another for the port. So basically, when you were doing the far field measurement, we are measuring the twitter and part of the mid bass response, and then we're doing the near field measurements toe, find out the low end response and then combined them together to get a full frequency response. Now, as always, we are going to start with the far field measurements. So place the microphone at one with their way, and, um halfway distance from the floor and ceiling. And since this is Ah, studio monitor, it has a built in amplifier. So we keep the amplifier from the A circuit diagram. So I connected the focus right interface directly into the monitor. Now let's open up, roomie, Kill Wizard. And now let's go ahead and do the measurement. Let's check the levels now. If you don't hear any sound, then that might mean you have to select output one. But, uh, we were like you from the start, so we're going to do a full frequency measurement because there is no need to protect the Twitter since it has crossover on it. So let's do the measurement more. Let's go ahead and check the impulse response so we have the first reflection at point 4.5 milliseconds. So let's apply the window. 4.5 milliseconds and the left window zero point supply some smoothing and outlets which toe the near field measurements. Now let's move the microphone into the near field position, so place it real close to the speaker. We are measuring the speaker first in the center of the speaker and as close as possible, and then we will do the measurement. First, let's renamed Far Field Measurement, Far Field. And then let's do the measurement. Let's, uh, decrease the output because the microphone is really close to the speaker. Let's check the levels. This should be fine. Let's do a measurement. Let's apply this morning. And by looking at this graph, we can tell that the resonant frequency of the box is Ah, somewhere at 63 hurts because we have the dip in the near field response measurement off the speaker. Now we have to do the near field measurement of the port. So let's rename this near field speaker. And now let's switch the microphone to the port Now that replaced the microphone in the correct position. So flush with the port in the center of the port, we can do the measurement using the same the same settings as the near field measurement off the speaker. So let's ah, start measuring. Now, let's apply this molding now we need to scale uh, the response of the poor to this response of the speaker according to their size. So I calculated the the area of the speaker in the area of the port. There is no way to find the diameter of the speaker, but you can measure it using a ruler. So from the apex off this around from one side to the apex of the surround on the other side. So after a calculating the areas we have, the let's bring up the calculator we have, Ah, the area of the port, which is 12 0.56 squared centimeters, divided by the area off the speaker, which is 86 point 59 square centimeters. Then we extract the square root, apply longer it and multiply by 20. So we have to reduce the level of the port by 8.4 decibels. So we take the near field measurement of the port, which is this one, and apply minus 8.4 disciples. Let's rename the measurement Near Field. For now, we have to combine the two together so we have the near field speaker in the near field port, and a plus B is generate. So let's rename this near field for plus speaker, and now we have to apply the box diffraction effects. So we have the near field, who for plus port they leave the first thing then said it control eight to select everything. Control, See to copy, Close the file, then open the Excel spreadsheets. Wait for it to load up paste the values entered the baffle dimensions. We have the height at 25 centimeters and the with at 17. 45 centimeters. Now let's scroll to the end. Control shift right up. Let go off control. Still hold shift and damn! Then copy everything, control. See? Then hit back to the text file. Control a control V and save the file. Then head back to room acu wizard and import the file import frequency response. And if we check these, we can see the lower output in based response. And now we have to splice this one and the far field measurement. So first we have to equalize. The response is somewhere in this area. So from 200 to 300 will be, uh, the best option. And again, always move the near field response. The far field stays the same. So take Ah, the near field response and tried to match the levels in the designated area. So this looks fine to this place right over here at 224 Hertz. It's had offset the data. So let's plies the far field with near field merge. B two a. It 224 hearts that's generating your curve. And if we close the far field and the near field there you have, this is the full frequency response off the M o d o b X five studio monitor. 32. 6.2 - Speaker full frequency response - at listening position: Now it's time to do a frequency response Measurement off your speaker with your room response. So actually, this is the easiest want to do because you don't have to do any gimmicks. Toe. Get the and a quick response. You simply place the microphone into your listening position and just do a normal measurement and that's it. Now, in our case, I have, ah, my studio monitors in the left and the right off my monitor. And, ah, I will have to place the microphone right where my head will be so right over here because I'm listening at my computer and we simply do the measurements. Now. After you made the measurement, you can start sitting up your audio using graphic equaliser, also playing with the position of the speakers or any boundaries that you might have some objects that you could move. For example, did my be doing some annoying reflections? Or you can place sound absorbing material in different places and see how your ah sound system measures after you place the those materials. But first, let's start with the easiest method, so with a graphic equalizer. So let's ah download the a nice application for windows. Go ahead and open up your browser. Let's go to Google and type a pew equalizer. Now you simply downloaded from here. They may try to walk you through it. So click next degree and ah, after you. I'm not going to stall it. But after you install it, it'll ask you to choose which sound device they want toe, Uh, the equalizer to effect. So you have to select your output sound card. In my case, I use for my studio monitors. I used the ASIS zone. Our essence one, which is this over here. So make sure you select your output sound card for your speakers that you are measuring. Now, after the installation is complete, you will find in the installation folder Uh, this, uh, editor. This is the program itself. So entered the editor. And here we can adjust the different frequencies with the gains and the band with but for its the But before we do that, let's head over to room, make you wizard and go to preferences. I understand. If you are using Ah, graphic Equalizer, you are messing up with the audio. So if you're using a CEO drivers, these drivers simply bypassed everything. So if you're the graphic, equaliser is nullified. So we have to use Jumma drivers and select your output device. The device which you connect the speaker, too. So in my case, I have the Arizona Essence one, like I said, and the input is our microphone. So in my case, ah, the microphone is coked up to the scarlet audio interface. And yes, we want to use the calibration file. One other thing that is important at the output. Make sure you use left plus right because we want to measure both the speakers simultaneously, and that should be it. Now you simply put your mic microphone into your listening position. Now let's click measure in. Let's check the levels So left, left plus right, let's go bit higher. And usually for these types of measurements, you can go for a shorter length because it will take less time since you're going to do repeated measurements. So let's start measuring war. No fancy stuff. We're just going to apply some smoothing so we can understand something from this graph and this is our response for our speakers with the room response so we can use the graphic equaliser to flat another response. But since I see this huge dip over here, this is, Ah, some unwanted reflections that that cancel each other out. And that's why you see this big dip over here. And for sure, you cannot resolve this just by equalizing things because it's it's too deep. It's like 25 decibels, which is way too much. If you compensate the, uh that much in the equalizer, you were surely introduced. Distortion. Well, let's have a feel of how it works. So let's see this pick over here. We have a little bump it 133 hurts and let's go to the equalizer. So we have various filters at the different frequencies. We can change that you want. So we have 124. Let's change this to 133 and we want to reduce the output. Let's say by five decibels and again see here how this effects the curve and the band with , um, 1.2. If we reduce this with we sharpen up this speak, so it's not so broad. Let's see how this Let me let me just increase this a bit more so you can see how how the band with affects the graph. So if I reduce this it the focus it focuses on 133 hurts and less on the neighboring frequencies. So I wanted a bit sharper, but minus 15 is way too much. Let's go for minus seven for eight and let's see what this has done toe our response. So now we seem to go on a measure again. Warren Place moving. As you can see, the response has lost a bitter amplitude in that peak. If we want, we can reduce it even more. Let's go for minus 15 and see what happens. Warren even more out with the lost. So now it looks much more linear and you can do this toe all of the frequencies you are interested. So if you want to reduce this speak or you want to increase these frequencies, just simply enter another frequency over here and do your modifications. So let's see if we can do anything about this dip, which is at somewhere around the 96 Hertz. So I'm going to use this filter at the 96 Hertz and trying to increase the gain when you are boosting frequencies. Um, it is wise not to exceed 10 decibels because you will introduce distortion. So let's, uh, decrease the band with as well. And let's see what the this did that the overall frequency response as do another measurement. Warren, Let me close everything else. Let's leave only the initial response. So we fixed this over here. It's flatter now, and this dip, while it's clearly not, is deep as before, it's still an issue. Like I said, that this is clearly some annoying reflection that introduces cancellation between the waves. And you cannot solve this by by equalizing, you have to play around with the positioning or with different boundaries that you might have and tried to fix if you can. So now that I showed you this, you can do how many alterations you want until you find the response that you are pleased with 33. 6.3 - Distortion measurements: Now I'm going to show you how to measure the stores. But before we get to the actual measurements, I want to tell you something about what distortion is. So, um, when you, Dr Ah, Speaker, you play a sound through the speaker If the speaker plays anything but that sound, that other thing is considered distortion. And the thing we are interested in here is the harmonic distortion. So if you play for example, 1 100 hertz tone, the speaker will play not only the 100 hertz don't. But it also will play 203 104 100 hertz. Ah, at a lower output, that is. But these are called harmonics. So any, uh, any whole multiplier off the initial tone are considered harmonics. So if you, for example, play uh 12 hurts tone, you will have the second harmonic at 24 herds, the third harmonic at the 36 hertz and so on. This happens for every speaker. So the issue here is Ah, you don't want toe avoid this completely, But you have to minimize the harmonic distortion. So if we start up roomie que wizard, I'm not gonna do the measurements again. I just opened the last project we did with the two way speaker, the 1st 1 where we did the measurements for each individual driver. So, like I said, when you're doing the measurement, you are measuring the impulse response. So from the impose response, you can derive all sorts off like frequency response, phase response, the distortion and whole other stuff. So to measure the distortion for each individual driver, we do the far field measurements. And if we are interested in the distortion for the lower octaves, we do the near field measurements as well. Besides troubleshooting any problems that the speaker might have When you are doing a distortion measurement, you can accurately choose a crossover point. So if you are looking at the distortion chart, then you are seeing that the speaker is distorting heavily at a certain frequency. You know that you need to set the crossover point somewhere above that, so let's take a look at the distortion sharp. So let's start with the Tweeter Far Field Tweeter and click the Distortion Tab. Now there are certain ways to look at the distortion chart. You can either view it from the perspective off decibels or percentage. So let's close the noise floor and the total harmonic distortion. And let's leave only the fundamental. So the initial tone and the harmonics. So here is our fundamental and we have ah, harmonic. So we're here so we can deduce that Let's say, at 1000 Hertz, the fundamental is Ah, 135 decibels and we have the second harmonic at 97 decibels. So we have the second harmonic, about 40 decibels below the fundamental. So which is very good to make our lives easier. We can normalize the plot to the fundamental. We can see that this response is not linear. But we can change this and make the fundamental flatline. And the the harmonics will change according to this, and we can easily interpret the data. So if we check this blood normalize, the fundamental we can see that the fundamental is a flat line at zero decibels and we can see how much below the fundamental the harmonics are. Now when you are judging certain harmonics, you have to understand that the human year perceives them differently, so even order harmonics are perceived the better than odd order harmonics. So a second order. Harmonic won't sound as bad as 1/3 order. Harmonic, for example, and lower order harmonics are also better compared. Toa the Tall order harmonics. So what? Sixth order Harmonic will sound harsher than, Ah, third order Harmonic, for example. So, for example, when you're looking at the second order harmonic, since it's ah ah, low warder and it's also on even order Harmonic. It's not so bad compared toa the other harmonics. So let's talk numbers. So if we look at our graph over here, when we have the scale on decibels, because we can change this to percentages will. Well, let's leave it at this symbols for now, everything below 40 decibels. So over here it's It's a good sign if it starts to go over 40 somewhere between 30 and forties, I guess. Okay, if it's going over 30 then it's pretty bad. So you have to take a look at this. If you are looking at percentage, for example now, depending on who you're asking, some say that anything below 1% so 1% it's over here. Anything below 1% is not curable distortion. Some say that below 2% is not curable, but some might hear it. Some might not. So to draw a conclusion, anything below 2% distortion is good. Anything, 1% is great. So now that we are judging the distortion of this tweeter, let's choose its crossover point. So judging by the distortion numbers, let's set the threshold for 1% distortion so we can see that anything above one kilohertz has under 1% distortion. So if the Twitter plays from one killer, hurt and above, everything is fine. But that is true Onley. If you have a crossover that blocks everything under one killer, and we know that is not true when the crossover point is met, then you have, ah, slope and you still get the some of the lower frequencies. So to get around this, when you choose the crossover point, you choose it one active above this point, so too kilohertz. And this might be true if you take a steeper slope. So for 44 there crossover, for example, and maybe for a second order crossover go for like 2500 words. Now let's look at the mid based driver. So as you can see, we are still in a percentage. So if we look under 2% first of all, we have to disregard everything below 250 Hertz because this is the far field measurement, so we can see that the driver is playing pretty well. It starts to get, um, rising distortion as it reaches the higher frequencies. Now, to check the distortion of the Bayside, we have to look at the near field measurements. So if we look at the near field measurement of the speaker, we can check the distortion here. So, um, we are looking on Lee at the lower octaves so we can see the distortion is quite high. So if we take the 2% mark anything 80 hertz start to rise. This is because it's a base reflects board. And, uh, consequently, the port will have lower distortion in this area because, as we know, the speaker doesn't move so much at the resonant frequency of the box. So if you look at the port, we can see a new inverse graphs. So we have lower distortion in the resonant frequency over the port and rising Ah, at the other ends. So as you can see, it's even even under 1% here and again. We have lower distortion here as we get into the port area that we have rising distortion because this is the job for the port and not for the speaker, as we can see over here. 34. 7.1 Presenting DATS v2/v3: as we're finished with the frequency response measurements, let's move on to the electrical measurements. So the device we are going to use for these measurements is the Dayton Audio Test system, or DATs for short. And it is just a simple device which connects to your computer using USB plug. And at the other end you have two alligator clips, which you used to connect your speaker or any other electrical components you want to measure. So let's hook this up to the computer and let's fire up the application. The first thing you want to do after you start the application for the first time, we need to calibrate the device and included with the package, you will find one kilo ohm resistor, 1000 homes and uh, you simply hook up to the clips like this, and then you go to impedance analyzer and impedance calibration. So we have the value off one clone and then click. OK, Now the device is calibrated, and another thing we need to do is is measure the resistance off the taste leads. So when you are measuring a device, let's say they're measuring resistor. The test leads themselves. Have ah resistance for their own. So you have to substructure it from the actual result. So you don't have to do that for every measurement. You just do it once and then compensate for the resistance of the leads. So what you need to do is short of the leads like this, and then go to impedance analyzer and test the leads calibration. So you will get a message to short the test leads which we already done, and now click OK, and you can see over here that the resistance is 0.37. So now, for each measurement we are doing the application automatics obstructs the resistance off the leads. Now we're done with the set up and we can move on to the actual measurements. 35. 7.2 - How to measure the Thiele / Small parameters: the first measurements. The first electric measurements I want to show you is how to measure the fields. Moral parameters of a speaker. So I'll measure the field small parameters of ah Tweeter and for mid based driver. So let's start with a simpler one. We have here V for Twitter, and all we have to do is connect the alligator clips toe the connectors. So if you look, you have to check the polarity. Here is the negative terminal you can see over here. And if I stop over here, you can see the plus sign here. Otherwise they might be color coded like the red one is the positive. So now all you have to do is clinic the alligator clips to the, uh, correct terminals. So this is the negative. And this is the positive having some trouble here, the magnet attracting the clips. So now you simply lay it flat on the table. En ifwe switch toe our application, you're simply click measure free air parameters and we're going to hear a sweep from the speaker and that is it. Here we have the impedance response and the red line is the phase response. So we know that the sharp impedance speak matches with the resonant frequency of the driver . So if we try to find the top most value 501 hurts, we have 18.48 domes. So if we look here at the parameters, we see that the resident frequency is at 500 hertz, which matches with the peak in impedance, and we have the rest of the parameters also the D C resistance and the cuse, the mechanical and electrical Q and the q t s that total Q and we also have l e which is the induct. Its we don't have emus NVs because there is no way to measure the mass, the moving mass and the V s. Since this is a Twitter, this is a parameter which does not interest us. So now let's switch to ah, different driver. Let's unhook up the tweeter and now we have ah, bigger driver, a mid bass driver from SAS. And now since ah, since this design has a vented pole piece, you don't simply place it on the table because you will obstruct the you will obstruct this event. This needs to be unobstructed. Eso with the speaker moves freely. So to solve this issue, I have over here to random books. But you can use just about anything and simply place the speaker between the books. So the vented pole piece is not obstructed by the table. Now, you simply hook up the Dayton audio device to the correct, uh, terminals positive to positive and negative to negative. And then we can switch back to the application in measure free air parameters. Now, if we look at the chart, we can see that the Peking impedance, uh, is the resonant frequency. But we can check the numbers on the right side. So the resonant frequency is 40.30. It hurts the resonant frequency in free air. And we have ah, the D C Resistance. Also, the Q t s is 0.35 But, uh, compared to the tweeter here we are interested in the V s and they're certain methods. We can achieve this. So, first of all, I want to give you a tip when you are doing impedance measurements, because when you are measuring the three year parameters, you're actually doing in impedance measurement and extracting the parameters from the impedance chart. Make sure the room is quite if you do and you hear a loud noises where you are doing, Um, don't do any noises because any sound ways that touch the cone will slightly move the cone . And since when you are measuring the impedance, it's actually a low signal measurement. The fact that this even the slight lis movement of the Speaker Cone, will we'll give you errors in your measurement. So make sure the room is quiet and you don't do any noises when you do the actual measure. So now that we got that out of the way, we can do the ves measurement. So in our application, we have four methods off measuring V s and we can find them on the right side. So first of all, you have to input the diameter off the driver, and then you can. Either first method is to place the speaker into a sealed box, and you have to know the volume precisely off that sealed box. Then you do another measurement with the speaker inside the box, and the program compares the two measurements and then it extracts the V s value in a similar fashion. You can do the edit mass method where you add ah, precisely known mass to the cone. So now you are changing the moving mass number and since we know the number, the the mass of the object and doing another measurement, then the program compares the two and extract the V s value and the other two methods that the deaths application proposed to us is the specified the spell method. So you have to know the SPL of the driver, the efficiency rating of the driver and the other method is by knowing the moving mass number, and today we are going to cover the easiest one. So I presume that you do not know the SPL or the moving mass number. So you have we have to place the driver into a sealed enclosure with a known volume, or we are going to place a mass on the cone and do another measurement to find out the V s . Now we're not going to do the the one with the enclosure, but if you want to do it, it's it's actually more precise. The accuracy of the measurement is better than the added mass omitted, but it's not so convenient to do so. If you want to build a sealed enclosure, make sure you know the volume precisely. And when you place the speaker inside the enclosure, make sure you place it with the magnet outside so it doesn't disturb the internal volume. Then you can do you can apply the method. But now we're going to do the added mass method and, ah, we're going to place. We are going to place this Scotch tape and we're simply going to place it on the cone like this. You can use any other object that won't bounce around when you are doing the measurement. So before we do before we do the second measurement with the added mass first, we need to place this on a precise scale. So have ah, don't use those kitchen scales. Make sure it has, um, the grams with one a decimal point with at least one decimal point. So I have my scale over here, So I simply turned the scale on in place. Place it on the scale, and the scale shows 120.6 grams. Now we simply place the object on the cone and then we head back to our application. Let's, um, enter the diameter. I know it s Oh, I see it in inches. I know the diameter is 13.16 centimeters, so I have to divide this by two point 54 5.18 inches. Five points, 18 inches. And let's select the added mass method and we have 120.6 grams. Now we can see that we have the measure of es button available and we simply click it and the master a driver and then click. OK, We already had the the Scotch tape on the cone and then let's click. OK, what? It simply doesn't other measurement. And now we have the V s value as well in the moving mass. Now, this is a bit confusing to me and cubic feet, and I'm going to use the metric units. Yeah, 100 years? Yeah, No, it's much better for me. So, as you can see, we have finished the measuring the fields more parameters over Twitter, which is much simpler, and the feels more parameters off base driver with a vented poor piece. So these are the parameters in free air when you I want to design a crossover, you are interested in the impedance measurements. But with the speaker placed inside your enclosure, so we will see that in the next lesson. 36. 7.3 - How to create ZMA files: Now it's time for me to show you how to create Zia me files. These are the impedance files. So to design a crossover, you need two types of files the frequency response, which is the FRG file and Z m a file, which is the impedance measurement. So now, uh, I have, ah, the same speaker set up, like in our first example where we measured the frequency response. So we have a two way bookshelf speaker in a base reflects enclosure. So the set up is nothing fancy like we did in the frequency response measurement. I simply place the speaker on the stand, and we have to connect the Twitter and the mid bases separately. So, like before I am going to measure the Twitter first. And we connected the treater to the first the upward, the terminals and the mid based driver I route. I wrote it. The cables through the base reflect sport. So let's hook up the treat her first, and let's go to our application, and we simply do on impedance sweep. So we have the impedance we button over here, simply click it what? You will hear the sweep, and this is the impedance over the tweeter. It shouldn't look much different from the impedance in free air because the tweeter is encapsulated and the enclosure won't effect the Twitter parameters in any way, or at least in a negligible way. So all we have to do is save this file. So we go to file export impedance data debts for much of text. Or is it a me? So let's rename this into Twitter and select Zia Main Peten data file and click safe. Now we have to do is switch to the mid base. Now the base is connected. Toa are Dayton audio test system and we simply need to do on impeding sweep. What now? We can see. This is much different from the three year, uh, impedance measurement you can see we have, ah, dual spikes. Now, this is you will see this kind of graph in a base triplex enclosure and to check the resident frequency off the box, it corresponds to the deep between the peaks. And if you remember, when we did the near field measurement, the resonant frequency of the box was 51 hurts and bear in mind that when you are doing a near field measurement with the microphone Dead measurement is much more accurate than determining the resonant frequency of the box from the impedance chart. So let's find ah, the point with the least impedance the so we find 5.56 point six homes. This is 26 7 six, 6.671 So, actually, they do match. So that's ah, really nice. So the impedance measurement tells us the exact same thing that the resonant frequency of the box is 51 hurts. So now let's create the Z m a file. So let's export impedance. So, like, Zia May and let's go for a base and Zia me And now we have, ah, the to see em a files. And if we combine these files with therefore defiles, we can load them up into Exim, for example, or any other application for designing crossovers. And, uh, you can start to model, uh, some crossover designs 37. 7.4 - Measure electrical components: Another useful feature off the data nor your test system is that you can use it to measure various electrical components. So I have here in in Dr on Air Corps in Doctor Ah, resistor and like a pastor. So, um, if you actually look on the components so if I showed up close, it would focus. There we go. You can see the value. So disk a pastor is 20 micro favorites. So why would you want to measure component if you know the value beforehand? Well, depending on the quality of the component, the manufacturer even quotes the error, so it might be 1% off the quoted value or 5% or even more for cheaper components. So if you want to know the exact value, you have to measure it. So let's start with the capacitor. Simply hook it up to the leads like this. Then let's go to the application. And on the left side, we have capacitor. Simply click the pastor and then click test so you can see the actual value is 19.7. Now, this is a good quality capacitor, so 19.7 is actually really close to 20 micro Farage's. So now let's test the in doctor. So if we look at the values, let me do a close up. You can see it's 0.82 Milli Henry's and zero point 44 homes. And now let's hook it up to the test leads and see. How exactly does this in Dr Measure? So let's close this up and choose in doctor and test so we can see that the resistance is spot on. But the inductive value is, uh, just a bit off. So it's a bit higher. Zero point 85 million Henry's and now let's ah, measure the resistor. I have a resistor here, so let me don't close up so you can see the value there. You can see it's 8.2 homes. Let's go back to the application. But before that, let's hook up the leads to the resistor, close this up and select resistor and then test so we'll see just a bit off. This one is 8.1 homes. Like I said, these are very good components, So if they are high quality ones, you can expect to have, ah, the values very close to the rated ones but if you're using some cheap components, it's, ah better to measure them. So you know the exact rating, so the exact value off the components. 38. 8.1 - XSim presentation: Now I'm going to show you how to use. Except I'm not going to show you how to design crossovers because that is left for another course. But I'm going to show you how toe set up to drivers and upload the necessary files and have a feel for the application and play around with crossover designs. Now, first we have to download the application. So let's start our browser. Let's go to Google and type of Exim crossover. Now go to this website and delete. Except so here you can download the application. Now, I'm not gonna do that because I already installed it. So let me start the application. Now here we have a window where we will place our electrical components. So here is the crossover diagram and on the right, we have the frequency response and the impedance. So in our case, we have ah, a two way speaker. So we're going to place to drivers you have here the components, the resisters in doctors and capacitors, drivers and ground. So we're going to place two drivers. The top one usually is the tweeter in the lower one is the midway's, So let's connect to the input. This is the positive and negative input. And let's draw the wires, set up the grounds for each speaker and for the amplifier. And now we have to our blow the files into the specific drivers. So the point is the tweeter. So let's right click and click Tune And here you are going toe upload for defile And here the Jeremy. So here we have the files which we created the previously in the two way design. So let's all blow the the Tweeter Frd and down we are blow the treater. Zia may Let's do the same for the mid base base and Bass said Great! Now you can see here we have the phase response and we don't see the frequency response yet . I'll show you why in a minute and then we have the combined impedance in the phase as well . So now we have actually the the frequency response. The combined frequency response is off the scale, so we need to fix that. There is there. We have our combined response when you are blow the frequency response. I advise you to not use the as measured check mark here, but the derive the one. So if you click derived here, you will apply the Hilbert bowed transformation. So what this essentially does is correct any measurement errors and extends the response to off the chart. Let's see. So now we measure the Tweeter from 20 from free 100 hertz because we wanted to protect the Tweeter. So the measurement is from 300 to 20 1000 hearts. Now, when you play the Hilbert bow with the transformation, the response will go down toe, then hurts and probably above 50 kilohertz. I don't know. So now we have to sit the tails so we know that the response was accurate up until t 300 hertz. So let's increase the lower tail toe 300 hertz, 313. That's great. So now we have to anticipate the the slope, the slope of the roll off, so usually try to Now, this is actually looks quite correct. If we go even lower, you have to keep the trend. Okay, So tried toe match to predict and match the trend, right? So for the upper tail Ah, 20 kilohertz with 12 db per octave slope looks fine. So let's it okay close and do the same for the mid base derived now. Ah, the higher tail. We can see the response going down. Now. This is actually measured. So it's correct. So let's go to 10 kilohertz where the response starts to roll off and try to match the actual trend. So this looks about right minus trench for disciples proactive. And for the lower tail, we actually have the near field response here. So somewhere about 60 hertz, it's OK. And try to match the trend minus three. This suppose practice. So this looks fine. Hit. Okay. And now we uploaded the FRG files, applied the Hilbert boat transformation and uploaded the ZESTIMATE files. Now all we have to do now, before we start adding components, toe our circuit, we have to add the acoustical off, sit off the mid bass driver, and now you can make an educated guess, or you can actually measure it. You are going to input the delay over here, so I'm going to leave this circuit as it is. I'm going to save it, and then we're going to load it afterwards and show you how to find the acoustical center 39. 8.2 - Acoustical offset - measurements: Now, before you start playing in Exim, uh, you have to do three things, so you have to a blow the 40. For else you have trouble the CME A files, and you have to know the acoustical offset between the drivers. So even though the drivers are flush on the baffle, we're talking about the two way speaker. Um, the acoustical centers of the speakers are not on the same spot horizontally, so we know that the sound radiated from the speaker is somewhere near the voice coil. So since the base has a bigger motor, the voice coil is somewhere in the rear. So even though the speakers are on the same baffle, if you play a sound through the Twitter, it will arrive earlier to the microphone compared to if you would play that same sound through the mid mace, you have to understand that exim is a basic application. If it would have been a more complex one, you could have input the position of the speaker on the X axis on the Y and on the Z axis. But instead we're gonna tell the acoustical difference between the speakers relative to the microphone. So you measure the distance from the Twitter to the microphone, the distance from the mid based of the microphone and the difference is considered the acoustical offset. Now, to do this procedure, we are going to do Ah far field measure for the twitter of far field measurement for the mid base. And then we are going to hook up the two speakers in peril and do another far field measurement. So after we made these measurements, we are going to create a far defiles for each individual one, and we are going to upload the Twitter in the mid base into Exim. So we will get the some response and then we are going toe overlay the FRG file with the drivers in parallel, and they will be slight mismatch as we increase or decrease the acoustical offset in Exim. The Kurds will start to match when they match perfectly. There is the acoustical difference between the two drivers and after that we will no to that number down and we can start playing with. Except so let's go to Roma que wizard and do the measurements. Let's do the far field measurement for the mid base. So, like usual said the microphone in At one meter distance, the speaker is halfway between the floor and the ceiling and let's do the measurement. Start measuring. Or now let's hook up the Twitter. Let's rename this to maybe base. Is that the measurement of the Twitter? But before we actually start measuring, let's increase the start frequency toe about 300 hurts so we don't damage the Twitter. Start measuring Great. This is the Tweeter. And now let's wired both off the drivers in parallel and let's do another far field measurement. Okay, now that we hooked up the Twitter in the big bass in a barrel, let's do the measurement. Start measuring. Let's rename this I mean base Plus No, let's do some smoothing and let's chicken the impulse. Three. I want to apply the same gate, the same window to all the measurements. So the short just one is the mid base somewhere around for milliseconds. No. Okay, so let's apply the window. 07 and four milliseconds fly window. Let's do the same for the tweeter my window and the same for for both drivers connected in peril. As you can see, the drivers are quite out of phase because when you, some the to they cancel each other out. But that shouldn't be a problem, because you will sort that out in the crossover. So let's export the each of these measurements into, if our defiles. So first, let's export the mid base sport measurement that's fixed. Okay, this is the I mean, the base. Now let's export the Tweeter export measurement. This text okay to eat there, save and the last one Tweeter plus base. Now we have to change the extensions, Farney. Now we need to head over to exam and use these files to find out the acoustical center. 40. 8.3 - Acoustical offset - processing of data: you know that we finished the doing the measurements. Let's start Exim and add two drivers. Let's connect them with the input at the ground. And now let's add the FRG files, which we just made. And let's add the Tweeter here and let's click derived to it toe Apply the hill Bourbeau The transformation. So just like before, that seems about right. We're not interested in the impedance file. So let's close this up and let's at the 40 file for the men Bees derived. So let's go to 10 kilohertz and match the slope minus 24. Now, here we don't have the near field response, So this looks a bit off. Uh, but, uh so here we have the response accurate from 300 hertz, something like that. And we need toe and the really shallow because we know the response extends all the way here and now click. OK, so let's ah scale the graph correctly. Here's our curve. And now we have to add Ah, the curve on this chart with both of the drivers wired in parallel. So click get filed and add the Twitter plus mid bees. There we have. This is ah the blue curve is, Ah, the one represented by this circuit over here, and the green one is with the drivers wired in parallel. Now all we have to do is tune the mid bass driver and add delay until the two curves match up. So let's increase the delay. As you can see, they start toe overlap and try to find a sweet spot because you are never going to get the perfect match. But at least try it as best as you can. So this looks about right. They pretty much overlap, so the critical difference between the mid range and the treater is 0.56 inches. So you have to remember this number as we will enter toe when we design a circuit a crossover circuit for this two way design. 41. 8.4 - XSim - finishing touches: Now that we found the the acoustical center, We are back toe our initial design. So we have Ah, the treater here with the 40 in the M E files loaded and ah, the derived for D file the same for the mid base. And as we found, the a critical difference. We have to enter the number here zero point 56 inches. Great. Now you are all set up toe add components as you wish. So since this is a twitter, you can It's a deal it this you can add, for example, capacitor, maybe at, ah, second order filter So on in Dr Toe change the directions place press space. So I had, uh, second order filter for the Tweeter and you can change the values here and see how the response changes. You cannot, uh, that's a first order filter on the mid base, and then you can adjust the value here and see how the response changes. And again, all you have to do is aim of for a linear response. As you can see, the graph is only up until 100 hertz, so you can fix that bite going toe 20 hertz for 10 So you have the full representation. Now we can have some fun with Exim. You already know the crossover basics from acoustics one on one so you can apply them here in ad components toe the circuit and see how they interact in the sea in real time. How the frequency response changes while you add components or change the values to the existing components. Now this marks the end of our course, and I hope you will have fun in Exim with the the basic knowledge that you have. And I also hope that you will have patients for me to release the third course, which will be about crossovers.