CAT-BUG*

*Car audio tuning

Beginners Ultimate Guide*

Or an easy-to-follow step-by-step highly detailed tuning guide for people that never did this before, but have a
lot of patience and want to learn.
 Table of contents
1. Introduction .......................................................................................... 4
2. What is tuning and why we need to tune or cars ............................. 5
3. Gear overview ...................................................................................... 6
3.1 Microphones ................................................................................................................................................................. 6
3.2 Audio interfaces ............................................................................................................................................................ 8
3.3 Software ........................................................................................................................................................................ 9

4. Initial settings .................................................................................... 11

4.1 Windows settings for microphone ............................................................................................................................. 11
4.2 Microphone calibration file ........................................................................................................................................ 13
4.3 REW settings ............................................................................................................................................................... 16
4.4 Pink noise .................................................................................................................................................................... 18

5. Measurements ................................................................................... 19
5.1 Sweeps vs RTA ............................................................................................................................................................ 19
5.2 Spatial averaging vs single mic position...................................................................................................................... 19
5.3 Crossover selection ..................................................................................................................................................... 25
5.3.1 Purpose of a crossover......................................................................................................................................... 25
5.3.2 Crossover types and slopes.................................................................................................................................. 25
5.3.3 Electrical vs acoustical crossovers ....................................................................................................................... 26
5.3.4 Why 24dB/oct Linkwitz-Riley? ............................................................................................................................. 28
5.3.5 Speaker “beaming” .............................................................................................................................................. 28
5.3.6 Starting typical crossover points.......................................................................................................................... 29
5.3.7 Optimal crossovers .............................................................................................................................................. 29
5.3.8 Using individual target curves.............................................................................................................................. 30

6. EQ ........................................................................................................... 32
6.1 What is EQ............................................................................................................................................................... 32
6.2 Cut vs boost ............................................................................................................................................................ 32
6.3 How flat do you need to go?................................................................................................................................... 33
6.4 Peaks vs Valleys and human perception ................................................................................................................. 34
6.5 Auto EQ with REW .................................................................................................................................................. 35
6.6 A note about smoothing ......................................................................................................................................... 41

7. Level matching drivers ......................................................................... 43

8. Checking for clipping and compensating gains if needed ............... 45
9. Time alignment ...................................................................................... 46
9.1 Why time align? .......................................................................................................................................................... 46
9.2 Time alignment using tape measure .......................................................................................................................... 46
9.3 Time alignment using Impulse Response ................................................................................................................... 47
9.3.1 What is Impulse response? .................................................................................................................................. 47
9.3.2 Microphone position ........................................................................................................................................... 47
9.3.3 Loopback .............................................................................................................................................................. 47
9.3.4 Phantom power ................................................................................................................................................... 48
9.3.5 Using HolmIMPULSE ............................................................................................................................................ 49

10. Checking for summing ....................................................................... 57

 1. Introduction
When I started my journey into car audio many… ahem 1.5 years ago, the car audio community greeted me with
heaps of (useless, contradictory and mystical) information and all the best advices on a sliver plate. Of course,
I am being ironic, the best advice when I started asking about tuning was - take it to a professional… I didn’t
want that, I wanted to learn everything myself, so I started to dig deep. I asked ALL the questions, there were
(and still are) many angry people that couldn’t (can’t) stand my multiple posts a day, in some of them asking
ridiculous or plain stupid questions. But after many heated arguments, after a few bans from FB groups I started
to get somewhere. Some people (lol, literally two) started to guide me to the right direction answering my
questions and finally I started to understand stuff. Kind of. Just to make it clear - by no means I consider myself
an expert or a pro at tuning, I am nowhere near that level and probably will never be, but that is not my goal. My
goal since I remember myself was to learn new things and skills. With tuning I just kind of know where to click
and what to do to make my system sound good to me.
A HUGE DISCLAIMER – this is not MY guide. I do not claim copyrights or any kind of ownership to the
information in this document. ALL the information provided here is easily accessible and obtainable online and
is free. I will list all the information sources that I used in creating this guide, the only thing that is kind of MINE
is the interpretation, understanding and formatting of the information.
I am a DIYer, a hobbyist having no interest in going pro. I’ve spent countless hours watching YouTube videos,
listening to podcasts, reading thousands of Facebook (or should I say Meta?) posts… All the things that people
do when they are obsessed with the hobby, you know? So, who is this guide for and why did I make it? It’s for
me. I mean for people like myself. For passionate enthusiasts that discovered this hobby but can’t find the
information. For people that struggle to weed out fact from fiction. Now I know what some of you are going to
say – why should I trust you and your guide? Who are you to make a guide? You shouldn’t, I am a nobody. But
here you are, reading this and looking how long this document is.
The purpose of this guide is to help people with tuning their own cars, to help them learn to click buttons and
hopefully in the process to make the systems sound amazing (at least better than stock). And best of all, if
English is not your first language – use google translate and read the document in your language. Print it and
have it in your lap while sitting in the car. I am a foreigner myself, so in this guide I haven’t used any of the fancy
English expressions that would be difficult to translate into other languages. There are quite a lot of technical
terms, so consult a dictionary if needed.
 2. What is tuning and why we need to
tune or cars
Tuning is a process which allows us to make a coherent audio system out of discrete components. Instead of
having a bunch of speakers making loud noises we can be a conductor via the DSP and tell individual drivers to
do particular things we want them to do and to do them exactly how we need them to be done. In this day and
age with cheap gear and widely available information it’s relatively easy to make your system in your car to
sound good, it just takes time. Car environment, the cabin with all the reflective surfaces and far from ideal
speaker locations contribute to the fact that a speaker by itself won’t sound the best. The transfer function of the
car will alter the frequency response so much that the speaker won’t sound good without manipulating the signal
that is going to it. If the install is finished, the only other way to change how the system sounds is with tuning.
For a speaker to sound nice it has to have “linear” or close to “flat” frequency response (amongst other things),
but instead we have this:

This is a typical 6.5in midbass driver frequency response from the driver side door. So how do we get from that
to something that actually sounds good? That is the whole process of tuning.
Before embarking on this whole journey, I have to make some assumptions.
First of all, I assume that you – the reader is somewhat technically literate. Then I will assume that your install
is finalised and nothing will change, because if you change one piece of gear in the middle of the tuning process,
most likely you will have to start over. I assume that your gains are set correctly (not up to clipping), that
input/output routing in the DSP is done and that your system is fully active, that means that you have a separate
channel of DSP and amplification for each driver in the system. And just to clarify – in this guide we won’t be
talking about installation, about driver locations, drivers themselves. This is a guide for a one seat tune.
Now it doesn’t really matter if you have a 2, 3, 4 or even a 5-way system, as the principles used in this guide
can be applied to any combination of drivers, just scale up or down. The results of this process, will depend on
your install and gear you are using. You can only polish a turd so much…
 3. Gear overview
Now let’s talk about everything that we will need to have a successful tuning session, here we will cover:

• Microphones
• Audio interfaces
• Software

3.1 Microphones
Let’s start with the absolute must – a measuring microphone. Professionals in the trade were using ears for
tuning since the dawn of time with very good results, but let’s be honest, how many of us have golden ears,
years of practice as live sound engineer and spend weekends training our ears to listen for miniscule differences
in sounds? Not many. For us amateurs, that have no real reference and experience a microphone is a gift from
the gods. Now days these measuring mics got so cheap that I am really surprised that not everyone has them.
I mean people pay over 200 dollars for SMD DD1 just to set the gains, but can’t buy a mic for 20 dollars?
Seriously?
To clarify we are talking about calibrated measuring microphones. There are 3 types of microphones defined by
the connection it is used: XLR, 3.5mm and USB. I can perfectly illustrate this with mic selection from Dayton
Audio.
The most basic, smallest and cheapest is the iMM-6 (iMM-6S for Apple devices) that uses a 3.5mm TRRS
connector. You can use this mic with a phone or any laptop that has TRRS jack. I use this mic with great success.
USB microphones like UMM-6 plugs directly into your laptop USB port, they have an integrated soundcard, so
you are not relying on your laptop. Benefit of USB mics is that they are plug and play and they will have calibrated
SPL reading.
XLR microphones like EMM-6 considered kind of “pro”, as they are used in the pro audio world utilising 3pin
balanced audio XLR connection for superior performance and noise rejection. They need “phantom power” to
work so, if you want to use them you will need a thing called an audio interface. Now why would you want to
complicate everything with an XLR mic and an audio interface if a simple iMM-6 does the job? You NEED an
XLR mix if you want to measure impulse response. We will need that to time and phase align the drivers.
Let’s have a look at the most popular microphones used in the car audio community for tuning. Firstly, we have
all the Dayton audio microphones mentioned above, the iMM-6, EMM-6 and UMM-6.

Audiofrog UMI-1. Very well regarded 3.5mm (however used

with an external USB sound card, so technically a USB mic)
small form factor microphone from no other but the famous
Andy.

MiniDSP UMIK-1 and UMIK-2

USB microphones from a
company specialising in home
audio DSPs.

Behringer ECM8000 is a cheap XLR microphone from a company known

worldwide for specialising in pro audio gear. Note: this microphone doesn’t
have a calibration file.
 3.2 Audio interfaces
An audio interface is just a fancy name for a sound card. Technically
you could use your laptop sound card, but to be honest – they are not
the best. In most cases it’s beneficial to use an External USB sound
card for your iMM-6, you can get one from eBay for less than a tenner
and it will be better than the one in your laptop. You don’t need an
audio interface for a USB microphone though.
Now if you want to tune with all the bells and whistles – get an XLR
microphone with an audio interface. This will allow you to measure
impulse response and phase for time and phase alignment.
Probably the most popular one is Focusrite Scarlett 2i2. A cheaper alternative is Behringer UMC202HD. If you
are on an even tighter budget look at Behringer UMC22 or even UM2.

We will be talking more about these things in the section about time alignment. Everything before that can be
done with a simple 3.5mm or a USB microphone.
 3.3 Software
During the tuning process we will need to use a few pieces of software: for measuring frequency response, for
targets and for measuring impulse response.
For measuring the frequency response and EQ work most people use REW – Room EQ Wizard. REW is a
software package that measures the transfer functions of acoustic systems and displays the corresponding
frequency, phase and impulse responses and various quantities derived from them.
You can download REW here: https://www.roomeqwizard.com/

Another very popular option is All-In-One JL developed software called TUN. It’s amazing if you are invested in
JL ecosphere, however you can still use it even without the MAX kit. Nothing against TUN, I just feel more
comfortable with REW. Same like all my life I was an Android guy.
You can download TUN here: https://www.jlaudio.com/pages/tun-4-jl-audio-s-dsp-tuning-software

The benefit of TUN software is that it has target mode integrated. REW has basic targets in EQ mode and it
works good for home audio, but for a car we need something more sophisticated and with more flexibility. For
setting our targets we will be using an Excel spreadsheet called “Jazzi’s tuning companion”.
You can download Jazzi’s tuning companion here:
https://www.dropbox.com/s/7ko9kl8n03ttpwb/Jazzi%27s%20tuning%20companion%20for%20REW%205.7.xls
m?dl=0
For measuring impulse response, you can use several different programs. You can even use REW, but my
favourite is “HolmIMPULSE” (again, just find in on Google to download). Bare in mind that HolmIMPULSE is an
older program that in order to work might need .NET framework to be installed in your computer first.
You can download HolmIMPULSE here:
https://download.cnet.com/HOLMImpulse/3000-2170_4-75910359.html
https://download.cnet.com/HOLMImpulse-64-Bit/3000-2170_4-75910360.html (64 Bit version)
Other people use “Rephase” or “Open sound meter”. King of measuring live phase is “SMAART”, but this might
give you a clue why is used only in pro audio world:

Ouch, right? All of these options (with the exception of SMAART) are totally FREE to download and use. All you
need to spend money on is the physical gear. Software needs just time to learn to use it, but that’s what’s this
guide is for!
For this guide I will use equipment that I personally own. However, you can easily adapt the techniques if using
different pieces of gear, as fundamentally they are the same. Here is the list of equipment that I have:

• Microphones - Dayton iMM-6, Dayton EMM-6, Behringer ECM8000.

• Audio interfaces – Focusrite Scarlet 2i2 and an external USB soundcard (no name from eBay).
• Software – REW, Jazzi’s tuning companion and HolmIMPULSE.
• DSP – Helix DSP Pro KM2.
• Needless to say, I will be using a Windows device.
Let me explain why I have 3 different microphones. Dayton iMM-6 is probably the cheapest measuring
microphone. When I was just starting to measure, I didn’t want to spent 5x the price on a USB mic like UMIK-1,
so got the Dayton. Later, when I discovered impulse response, I realised that it’s not possible to use that mic for
measuring IR, I needed an XLR microphone with an audio interface, so bought the Dayton EMM-6. And
Behringer was so cheap on FB marketplace, I snatched it just to have a spare one.
Ok, so you have your microphone? Have the audio interface if needed? All the software downloaded? Let’s
begin to actually click buttons. We are still quite away from actual measurements, first we need to do some prep
work.
 4. Initial settings
4.1 Windows settings for microphone
Windows in some ways is a weird operating system and does weird things to input and output devices by default.
First, we need to change some things in order to make sure that our measuring rig works correctly and to save
potential troubleshooting time down the line.
Find the “Speaker” icon on the bottom right corner and right-click on it. Now open “Sound settings”.

Choose input device that you want to use, in my case I will be choosing my USB2 device (Dayton iMM-6 mic
plugged in into external USB soundcard). After that, click on “Device properties”.
Here make sure the volume is up to 100 (By default the mic volume is much lower and you will have a lower
reading).
After doing that click on “Additional device properties”.

Here in the “Advanced” tab we need to select the sample rate that we will be working with, in my case I will
choose 48kHz 24bit. In some cases, you might have an “Enhancements” tab. If you do, click on it and make
sure that all the possible enhancements are ticked off. We don’t want Windows to alter the microphone signal.
That’s it, this part is done. Off to sort out calibration.
 4.2 Microphone calibration file
Measuring microphones, as good as they are, they don’t have a perfectly linear response and every
manufacturer calibrates every single microphone. Every microphone has a unique serial number that is written
on it or on the packaging. Let’s go to google.com and enter “download <your_microphone_manifactutrer>
measuring microphone calibration file”. I end up in Dayton audio page where I need to enter the serial number
and download the .txt file.

Now we are ready to open REW for the first time. Finally, right?
First click on “Preferences” in the top right corner and in the “Soundcard” tab choose same sample rate as you
choose in Windows settings previously.

Then choose your input device.

Now click on “Cal files” tab and under “Mic cal files” browse and find your recently downloaded .txt file.

That’s it, we are ready to measure! Or are we?

 4.3 REW settings
Before starting to measure we need to make sure that the data will be gathered and displayed the way we want.
Click on “RTA” button on the top menu and then the “Wheel” icon on the newly open RTA window.
These are the settings you need to have in order to have “good” measurements.

• Mode – we will be using RTA mode and we want to measure in the highest possible resolution, in this
case 1/48th.
• Smoothing – at the measuring stage we don’t want to apply any smoothing, we need the data to be as
“raw” as possible. We will touch on smoothing later on when we will be talking about EQ.
• FFT Length – keep it at default 64k. Basically the lower this number, the faster REW will read averages,
but with less accuracy. With bigger numbers you will have much higher accuracy, but will have to wait
ages. 64k is a good middle ground.
• Window – Rectangular
• Max Overlap – 93.75%
Now one last thing to check. Click on “Appearance” and make sure the “Use bars on RTA” is unticked. We want
to see continuous line and not bars.

“Can I do the measurements now” – you ask? Almost. We’ve set up everything to measure with, but what we
will be measuring?
 4.4 Pink noise
If we are using RTA to take our measurements, we need to use pink noise.
Pink noise or 1⁄f noise is a signal or process with a frequency spectrum such that the power spectral density
(power per frequency interval) is inversely proportional to the frequency of the signal. In pink noise, each octave
interval (halving or doubling in frequency) carries an equal amount of noise energy.
Don’t worry if you didn’t understand that description, you don’t need to know that, all you need to know is where
to find that bloody noise. You might be tempted to use “Pink
noise” on YouTube, or Spotify or a “Pink noise generator” app or
a random .mp3 from the depths of the internet, but that is highly
NOT advisable. Most of those “noises” are band limited random
pink noise, we need crystal clear lab grade pink noise.
In REW top menu choose “Generator” and let’s make Noise ->
Pink periodic -> Full range.
Now you have two options how to use the pink noise and it will
depend on the setup in your car.
You can use your laptop (or external soundcard or audio
interface) and connect it to the DSP directly with something like
3.5mm to RCA cable. And then play the noise from REW
generator.
I use digital input on my DSP and signal travels from my Android
smartphone via USB. So, I need a pink noise file that I could play
from my phone. If you use a head unit or a DAP, you will need a
file as well.
Choose “Save to file” and change the “Repetitions” into a bigger
number. 440 repetitions will give you a 10-minute-long file which
I am happy with. Save the file as .waw and transfer from your
laptop into the device that you use as source. NOW we can take
measurements.
 5. Measurements
Now that we have everything ready for taking measurements we need to decide and understand what kind of
measurements we should be taking, why and how to physically take them.

5.1 Sweeps vs RTA

There are two methods to measure the frequency response sweep method and RTA.
Sweep. A logarithmic sweep signal is sent to the source, starting low and steadily increasing to a higher
frequency. A logarithmic sweep is one that takes the same time to double in frequency. It takes the same time
to go from 40 to 80Hz or 4kHz to 8kHz as from 20 to 40Hz. The mic picks up the sweep, sensing the sound that
travels directly from the source and all the sound that first bounces off the room surfaces. Some sound paths
may bounce off several room surfaces before reaching the microphone. Once the sweep completes, analysis
starts. A process called "Fast Fourier Transform" (FFT) is used to calculate the phase and amplitude of the
frequency spectrum that made up the sweep. Each doubling of the sweep length improves S/N by almost 3 dB.
RTA. A common way of viewing the spectrum of a time signal is to use a Real Time Analyser or RTA. The RTA
shows a plot of the amplitudes of the frequencies that make up the signals it is analysing. A real-time analyser
(RTA) measures and displays the frequency spectrum of an audio signal; a spectrum analyser that works in real
time.
To put it in simpler words, one goes woooop, the other goes pssshhhhh. We will be using the pssshhhhh one,
called RTA. Sweeps offer much better signal to noise ratio, but they need to be done with a stationary
microphone in a single position in space. That doesn’t really work in a car due to very small cabin space and
reflections that affect the direct sound.

5.2 Spatial averaging vs single mic position

When taking measurements, we will be using MMM – moving microphone method. This means that while we
are recording data with RTA we will need to move the microphone around our ears and head. The “moving” is
pretty simple – hold the microphone at 90-degree angle (pointing towards the roof of the car) and make small
(the size of a tennis ball) circles around your ear, make 2-3 circles and move the microphone to the other ear
and repeat. And keep going back and forth between left and right ear. Make sure you spend equal amounts of
time around each ear.
This method will give us reliable and most importantly repeatable results. If we take two measurements and they
are different from each other, we won’t have any consistency (That would happen with sweeps if the microphone
is not in exactly the same position in space for different measurements). Here are some measurements
illustrating the difference between stationary microphone and MMM.
Subwoofer

Midbass

Midrange

Tweeter
All drivers

As you can clearly see the stationary microphone and MMM are starting to deviate from around 180Hz. So, in
conclusion – measuring below ~200Hz you can use a stationary microphone fixed to the headrest, but for
measurements 200Hz and above best is to use MMM (moving microphone method).
Now we understand what kind of measurements we need to take and how to do it, so let the measuring begin!
In REW RTA mode you start the measurement by pressing the big red button on the top right corner. In the
bottom left corner of the graph you will see an increasing number of averages over time. Our goal is to measure
and let the RTA take averages until the frequency response stops changing. In my personal experience (and
many others) that happens when REW does around 80-100 averages. For subwoofers 40-50 averages might
be enough.
When you reach ~100 averages or when you notice that the frequency response stops changing stop the
measurement by pressing the red button again. Now you have to save the measurement by clicking the floppy
disk icon with “Current”. This way you will save the graph and transfer it to the main REW window.
There click on “All SPL” tab.
On the left side you will have all your measurements – please name ALL of them, as it’s very easy to get lost
and later try to remember which was left and which right, which was with 3kHz crossover and which with 3.2kHz.
Believe me, I learned that lesson the hard way…

After the test run the first measurement we will take is the background noise in the car, to filter out any
abnormalities such as aircon, transformer hum, wind, rain, grounding issues, etc.
This is an example of issue with my microphone – red trace is when I hold the microphone in my hand and green
is when I am not touching it. It’s grounding issue that will need investigating.

Same way you can spot potential errors in your measurements if you forgot to turn off something that your
hearing is accustomed to (like aircon in the garage).
Now take measurements of individual drivers – raw responses. Make sure that the measurement level is well
above the floor noise. Some say +20dB above floor noise. Some say at normal music listening levels. Some
have a number: 80-85dB (USB microphones are calibrated for SPL as well). The noise level should be loud, but
bearable, as you will be listening to it for a while.
Go to your DSP software and mute all the drivers apart for one. Bypass the EQ (if you have any), the HP and
LP filters. The only exception is the tweeter! Use a high pass filter of a lower-than-normal value for protection,
something like 1-1.5kHz. For a smaller midrange you can use 100-150Hz.

Have REW with RTA window ready. Start playing pink noise on your source and click the big red button on the
RTA window. And off you go! ~100 averages for one driver. Save measurement. Mute that driver, unmute the
other and repeat. Do that until you have all the drivers measured. I start with the left tweeter, then right tweeter,
left midrange, right midrange and so on.
Pro tip – when doing these measurements, it would be a good idea to take near field measurements of all the
drivers as well. Stick the microphone as close to the driver’s dust cap as you can without touching it and take
40-50 averages. It’s not mandatory, but might save a lot of headache later on trying to troubleshoot issues if
there will be any. At least you will be able to rule out one thing for sure. Let me show you an example of what I
mean.
In this graph you see a frequency response measurements of some midrange drivers. The person that took
these measurements couldn’t understand, why there is such a huge dip around 3kHz (the bottom graphs are
measured from drivers seat). Only after taking near field measurements (top graphs) problem became clear. It
was install related issue where the midrange driver was sunken into the door card to deep.

Anyways, done with all your first round of measurements? If done you will see something like this:

A total mess, right? Don’t worry, now the real fun begins.
 5.3 Crossover selection
This is the part where you will need to do some reading, as I or this guide won’t be able to help you to choose
your crossovers. That is your decision and it depends on your gear, environment and your music listening style.
So, strap in.

5.3.1 Purpose of a crossover

A crossover between two drivers serves a few purposes (there are many more, but we will simplify things a bit).
No 1. Protection from over excursion. You know the RMS value in watts of diver? Well, that value is only half of
the story. The RMS value is the thermal capacity of the driver, basically it means that it can handle only a
particular amount of constant power before it overheats and glue in the coil starts to melt. Excursion is the thing
that kills most of the drivers way before thermal limits. Ever played a driver in free air (without acoustic
suspension – air pillow behind it)? With just a few watts you can easily exceed Xmax and Xmech. A crossover
at the bottom end – High pass filter (HPF) – helps to protect the driver from going over Xmax. That’s why if you
want to “send it full tilt” – have high and steep crossovers. You can play your midbass drivers with no HPF but
not that loud before they start distorting. Choosing the correct crossover for the driver partially will depend on
how loud you are listening to your music. If you listen low to moderate levels – you can have your crossovers
lower compared to “sending full tilt”. A safe HPF is considered 2xFs of your midrange and tweeter and close to
Fs for the midbass. Pull out that manual and have a look at your drivers T/S parameters – Fs is somewhere
there. If your midrange has an Fs of ~150Hz and you are listening to your music loud, a safe crossover will be
300Hz or above.
No 2. Blending the drivers. The whole purpose of having 2, 3 or even 4-way front end is to let the drivers play
their intended range only. Let them do what they are designed to. And when one driver feels that it’s too much
– it passes the baton to other driver, more suitable for the job. You don’t want your midbass to play 5kHz and
up. It will certainly try, but with cone breakup, heavy cone and directivity issues it won’t be optimal, so it needs
to pass those higher frequencies to a tweeter. And it needs a Low Pass Filter to blend the drivers responses.

5.3.2 Crossover types and slopes

There are many crossover types that are characterised by the shape of the slope, they are named by people
who “discovered” them. To complicate things even further – all of them sum differently and have different phase
relationships. Relax, we are not going to go there.
A slope is a characteristic of a filter that defines how steep it is, in simpler terms – how fast the frequencies are
being attenuated. Slopes can vary from very shallow -6dB/oct to very steep like -48dB/oct. What is “oct” you
ask? It’s an octave – halving or doubling of a frequency. Take for example 200Hz. One octave down will be
100Hz, one octave up will be 400Hz. Remember this – it’s very important, as we will be talking in octaves from
now on. 20Hz to 20kHz (the whole human hearing range) is 10 octaves in total.
So, let’s take for example 80Hz 24dB crossover (let’s ignore the filter type for now). If you apply this HPF to
midbass driver that means that one octave below – at 40Hz, the frequency response will be 24dB down. In
theory.

5.3.3 Electrical vs acoustical crossovers

As everywhere in the world theory is different from practice. In real world drivers frequency response is not flat.
It has a natural roll off at both ends of the frequency response. Let’s have a look at My midrange response with
no crossovers applied. You can see that this particular driver in this particular application in this particular car
has a natural roll off at the bottom end starting at around 120Hz and at the top end at around 8kHz.
Here I overlay a 120Hz and 8000Hz 18dB/oct filters. So, you can see that the natural roll off of this driver is
18dB/oct from both ends. What happens if I apply a 24dB/oct HPF at 120Hz to this response?

18dB/oct + 24dB/oct = 42dB/oct slope. Basically, slopes add up. Now here we need to define one of the most
important concepts in crossovers that confuses a lot of people, and even more people don’t know about this –
electrical vs acoustical crossovers.

Electrical crossover is what you apply electronically in the DSP, it’s what is applied to electrical signal that goes
to your amps and speakers. Acoustical crossover is what you are measuring with RTA. And 98% of the time
they are NOT THE SAME. When tuning we care about acoustical crossovers only. Electrical crossover,
especially HPF has to be safe to prevent over excursion, that’s it.
 5.3.4 Why 24dB/oct Linkwitz-Riley?
We are not going into the details of how different filters sum and what happens to phase, all we need to know is
that 24dB/oct LR ACOUSTICAL crossovers are the “best” for our purpose, as they sum flat and don’t need phase
manipulation. In this tuning guide we will be using 24dB/oct LR ACOUSTICAL slopes everywhere.

5.3.5 Speaker “beaming”

Or on axis vs off axis. This is not that important for us to know, but still useful. This should be utilised in system
design stage, when you are choosing drivers. You would want all your drivers to play omnidirectionally – in all
directions, and not beaming. You can’t avoid that with tweeters though, as they have to be aimed on axis towards
the listener to have that top end extension. Now if you take for example a 3in midrange and you are playing it
up to let’s say 3000Hz, it can be off axis and still sound good. You don’t need to aim your midbass drivers in the
doors toward your face unless you are playing them above 4kHz.
 5.3.6 Starting typical crossover points
80-500-4000. Or 70-300-3000. Or something similar. All 24dB/oct LR. If you have absolutely no idea where to
put your crossovers but want to have a quick listen and you have a very typical 3-way front system – sub, 6.5in
midbass, 2.5-3in midrange and a tweeter, that is a good starting point.

5.3.7 Optimal crossovers

Are dictated by the cabin. Let’s have a look at the midrange response I’ve included previously, what crossover
points should we choose for this driver? For HPF we could choose 100Hz. The driver is plenty capable, it’s a 5in
that has an Fs of 41Hz. But I don’t really need it to play that low, especially from the dash. I would prefer that
my bigger 6.5in midbass drivers would play those lower frequencies.

So, let’s have a look at what my midbass is doing. As you can see – it plays very low (the driver is in infinite
baffle kicks) but above 200Hz, there are some problems. Due to the location, my midbass doesn’t want to play
above ~180Hz. So, in my case the HPF of the midrange is determined not by the midrange driver but by the
location of the midbass driver. Wow, right?

Choosing correct crossovers is all about identifying your bottle necks or week spots in the system. Maybe your
midbass in the doors plays well up to 1000Hz but the midrange has an Fs of 200Hz. Then you will cross them
somewhere around 400-500Hz. Every system is different. Take into account safety of the driver and the
frequency response that you measured previously, remember - the raw responses with no crossovers?
 5.3.8 Using individual target curves
So, you have a rough idea of what you want your crossover points to be. Next step is preparation for EQ work
and optimising the crossovers. We need to get individual drivers target curves. Do you have Jazzi’s tuning
companion downloaded (please find the link above)?
Open it and go to “3-way house curves” tab at the bottom. Now here you have to make one of the biggest
decisions of your life – choose a house curve. A house curve is the overall frequency response of the whole
system and it’s personal. Everybody likes different things. You can choose one of the presets, like Whiteledge
or JBL. I have a “Custom” one that I like. You can see the values on the right column “House curve custom”.
Enter your crossover values that you decided to use and choose LR 24dB slopes. Then export and save them
in a known location. The Excel will spit out .txt files. When you have your individual targets, drag and drop them
into REW where you have all your RAW measurements of the drivers.
Once you’ve imported the individual targets into REW you will see something like this (I have a “front sub”, that’s
why I have an extra trace). Click on “Controls”, choose “Measurement actions”. In a new open window choose
all the curves one by one and offset them to have them on the same level as your measurements (when you
import them, they will be at 0dB level). In my case ~54dB.

If you have time, patience and will, you can experiment with this by making a bunch of different curves, overlay
them in REW and see which one fits your particular divers frequency response best. For example, here I am
deciding which crossover point between 40Hz and 60Hz for my subwoofer works best. I chose 45Hz (the orange
trace).
 6. EQ
6.1 What is EQ
If you will take drivers response near field you will notice that it’s smooth with no major peaks and dips, however,
when you measure the same driver from the listening position – everything changes. You will have widely varying
frequency response. The cabin of the car or any room for that matter has physical dimensions, has reflective
surfaces and just stuff in the way. That causes the response to change. And in order for the system to sound
good we need to correct that. We need to correct the room – our car. REW is Room Equalisation Wizard. Notice,
how it’s not speaker or driver, but the ROOM.

6.2 Cut vs boost

Using EQ you have two options – to cut and to boost. For some reason everyone is scared of boosting the
signal. They think that even a little of boost will introduce distortion and clipping. Some have set “rule” of max
boost only by 3dB. Some don’t like to use EQ at all. But we will use all the tools available for us in order to reach
our goal – to make our system sound good. When it comes to boosting the main thing you need to look at is
overall boost in Helix.
This particular example is one of my midbass drivers. You can see there are 3 EQ bands with +6dB boost,
however there is so much cutting and even the channel gain is way down, so that boosting won’t clip anything.
The overall gain is -10dB (I might need to check if the gain on the amp is not increased from previous tuning
sessions).
Now let’s have a look at one of my midranges. There is a lot of EQ bands boosted, however, only a slight peak
at 1.5kHz is over the 0dB line in the overall boost. That looks like +4dB, however if you look at the channel gain
– it’s -3.25dB, so in total, my midrange is boosted by less than 1dB at a very narrow frequency range. Now, will
that clip the signal and introduce distortion? In theory yes, IF I will play 1.5kHz sine wave at full tilt. On music
that won’t have any effect.

I did some testing on Helix DSP clipping, and it does indeed clip if you go over 0dB line with the EQ boost,
however only if you play sine waves at max volume. So don’t be scared to boost, just think about what you are
doing and be smart about it. Look at the overall boost. If you have +5dB on the overall, just lower the channel
gain by 5dB and you will be ok.

6.3 How flat do you need to go?

Human hearing is quite sensitive and well-trained ears can hear a difference of 0.5-1dB while playing tones.
With music is slightly different. Are you a musician? Most likely no, and most likely you don’t have golden ears.
When EQing to target we can aim to be +/-3dB from it. We can aim +/-1dB. You can aim for whatever “flatness”
you desire. Just remember – the “flatter” you aim, the more EQ you will need and eventually you might run out
of those 31 bands available on Helix DSP. Especially for mids…
 6.4 Peaks vs Valleys and human perception
Humans hear peaks in the frequency response much more than valleys, especially high Q – narrow ones. EQ
work is quite delicate as you need to know what to EQ and what to leave alone.
For example, those peaks at 580Hz for midrange and 2.7kHz for tweeter will be very audible and we would need
to EQ them down.

That hole at 68Hz however, is a cancelation – it’s deep and narrow and it would be pointless trying to boost it.
Even if you would boost by +12dB and will get maybe 5-6dB improvement, that extra energy introduced by
boosting might damage the driver or excite panels around the driver. So, if you see a deep and narrow hole like
this – leave it alone. At lower frequencies it might be audible if a particular song happens to have main bass
note at exactly between 60 and 70Hz – you will notice lack of bass. If this trench is only on one driver, the bass
will pull to the other side. At higher frequencies, especially on tweeters, narrow and deep trenches most likely
will not be audible.
There are many more nuances about human hearing and sound perception, there are books about this stuff so,
I won’t go into much detail. You can spend some extra time in this field if you want to. Or just listen and hear for
yourself. Let’s start EQing!
When I first started to EQ, I did it manually. Make a cut with one band, then measure, make another, and
measure again. It took me 2 hours to EQ one tweeter… Then I discovered REW auto EQ function and I’m still
amazed how cool this feature is. Now I can EQ my whole 5-way system in 3-4 hours! In this part individual target
curves become very handy. Those targets will speed up the whole process and will let us avoid the guesswork.
So, let’s begin.

6.5 Auto EQ with REW

In REW open “Preferences” and find the “House curve” tab. Browse and navigate to the folder you saved your
targets from Jazzi’s tuning companion. Choose and open target you want to start working with. I will start with
my midbass.
On the left, choose the actual midbass measurement you want to work with (Left or Right) and in the main REW
window choose EQ.

Now let’s prepare our workbench. Firstly, click on the black “down” arrow to maximise the EQ window in order
to see better. In the top right corner click on “Equaliser” and choose “Generic” if using Helix DSP. If you are
using a listed DSP – choose that. As some of them have different EQ parameters. In the “Target settings” under
“Targe type” choose “Full range speaker” – this is our target curve that we imported in the preferences previously.
Play with the “Target level” to match it with your measurement. This is what we will call “setting the target line”.
After that is done click on “EQ Filters” and a new window will open. Scroll down (or just make the window bigger)
and at the bottom you will find “Crossover types”. Here is where you will finalise the crossover filter selection.
You already chose the crossover points or the crossover frequency for each driver, now we will see what actual
crossovers would be best to apply. Remember the electrical vs acoustical crossovers? Your target (the blue line)
has ideal 24dB LR slopes. Now we need to match our response to that target as close as possible by using ANY
crossovers available. Yes, any. If you need you can choose 3rd order Bessel or 7th order Butterworth. The actual
crossovers don’t matter AS LONG as they are SAFE for the driver. For example, you wouldn’t use a 1st order
Butterworth filter for a tweeter. For my midbass 70Hz LR4 and 280Hz LR4 looks to work best.

After choosing the optimal crossovers for the driver, put them into your DSP and make another set of
measurements with all the crossovers applied. With the new measurement return back to EQ window and let’s
actually EQ.
With the “target level” choose optimal target line. Now what does this mean exactly? If you choose the target too
high, you will need to boost the valleys quite a lot. In this example 37Hz, 65Hz, 100Hz, 170Hz, 220Hz and 390Hz
all needs boosting and not much cutting elsewhere. While boosting is not bad, we would like to avoid ONLY
boosting because that might be a recipe for disaster, where in the DSP our overall boost might be way above
0dB line.

Now you could go the opposite way and choose a target line that is too low. That will result in a lot of cutting and
potentially loosing your dynamic range.

Ideally you would want your target line to be at a level that might need slight boosting, but the rest is mainly
cutting.
When you decided on the target line, we need to tell REW what to do with it. In the “Filter settings” we need to
tell REW where in the frequency range to apply the EQ. Ideally you would want to match the response to target
EVERYWHERE. That means not only the top, but the slopes as well. It is recommended to EQ to target line one
octave below the HPF and one octave above the LPF. In this particular case my crossovers for the target curve
are at 100Hz and 180Hz. No, not 70Hz and 280Hz like I chose in the EQ window before! Remember – those
are electrical crossovers that help me to match the 100Hz and 180Hz acoustical crossovers. Confusing? Go
back and read everything again.
So, ideally I would need to EQ from 50Hz to 320Hz, however I will make sure that I match to the target in a wider
range. I will choose from 30Hz all the way up to 1000Hz. See that “knee” at 800Hz? I really don’t want that to
bleed into the midrange and affect the summation. I have loads of EQ band available so – why not?
Then choose “Individual max boost” – 6dB, “Overall max boost” – 5dB and “Flatness target” – 1dB. That will be
the flattest REW will go. It will use as many EQ bands as needed to reach that target. Tick both options below
that and click on “Match response to target”. And let REW to do its job. After a few seconds (that will depend
how fast your laptop is), you will see a predicted frequency response. If you are not happy with the outcome –
change the frequency range, target level or go to “EQ filters” and manually tweak some EQ bands until you are
satisfied. When done click on “Export file settings as text”.

This export function is a godsent, is saves soooooo much time, as otherwise you would have to type into the
DSP every single EQ band manually. Yes, I did that until I discovered this export function. Oh, and if you have
a Zapco DSP – good luck with typing.
After saving the .txt file (please don’t forget to name it) open Helix DSP software and right click on the channel
for the appropriate driver. Select “Load import file”. In “Files of type” choose *.txt and find that .txt file you just
saved. Click “Open” and “Ok”. Re-measure to confirm that you are happy with the result.

And that’s it – you are done! With one driver that is. Now you will have to do the whole process again for every
single driver in your system. Oh yes, my friend – tuning takes time. But good news is – the more you do it, the
faster you get. So, practice, practice and practice.
 6.6 A note about smoothing
Remember that I recommended to always measure with the highest resolution that you can? That is in 1/48 th
oct. After that, you can apply smoothing by going to “Controls” and choosing whichever smoothing you want.
This is an example of the same response with no smoothing (Green) and with 1/3rd smoothing (Brown) applied,
as you can see, they are quite different.

Now imagine the EQ work on these two different smoothing responses. Obviously, with more smoothing the
response is already much smoother and will need less EQ work to reach the target. Remember what I said about
narrow and deep dips? Take a look at 70Hz and 360Hz dips. With smoothing applied you might think that you
can boost those dips slightly and they will be flat, but with no smoothing you can see that those most likely will
be cancelations which we shouldn’t boost. So, use smoothing with caution, as smoothing hides a lot of things
and you might end up doing things that you shouldn’t be doing.
Exactly opposite is true for the tweeters.
If you would try to EQ every peak and dip in the tweeter response, it would be pointless. You would use many
EQ bands and the effect by using them would be inaudible.

Now here’s the thing. Many people say (and it’s backed up by scientific research), that our hearing is closest to
1/3rd octave smoothing. We might hear more granularly in the lower octaves but for high frequencies it’s 1/3rd. In
REW there in an option to use “Psychoacoustic” smoothing, which applies a gradually increasing smoothing.
What smoothing you want to work with is up to you. Many use 1/3rd for everything. Many use 1/12th for subs,
1/6th for midbass and 1/3rd for midrange and tweeters (Psychoacoustic). I use 1/48th for everything (yes, I know
it’s pointless). The more smoothing you apply, the less EQ bands you will need to use. Yeah, told you it’s a lot
of stuff to consider.
 7. Level matching drivers
You did all the EQ work, the Left and right drivers have a verry similar response, now we need to make sure that
left and right are playing at exactly the same level (volume). There is an easy trick to level match the individual
drivers. You will use this for your midbass, midrange and tweeters. Let’s have a look at my midranges. My right
midrange is slightly louder than the left one. I’m in the UK, so it’s my diver side that is louder, for Americans most
likely it will be opposite. Now you can try to match the responses by playing with level, but for more accuracy do
this.

Go to controls and apply highest possible smoothing which is 1/1 – full octave. Then choose “Measurement
actions”. Find the top trace in the drop-down menu and play with “SPL offset” until the responses are matched.
In my case I had to bring the right mid down by 3.2dB.

Go to Helix DSP software and reduce the channel gain by -3.2dB. Re-measure to confirm. And that’s it – your
mids are level matched. Repeat for midbass and tweeters.
 8. Checking for clipping and
compensating gains if needed
After all the EQ work is done, it’s a good idea to double check for possible clipping and review the gain structure.
If for example In Helix DSP you see that the overall boost is too high for a particular driver, maybe it would be
worth revisiting the EQ part and choose a lower target line to avoid boosting too much?
Maybe after all the EQ cuts the driver is not loud enough? Go to the amp and increase the gain.
Now maybe you have noticed, that we were working only with individual drivers one by one and not looking at
the system as a whole. If you would try to listen to the whole system now it will sound pretty bad… Why? Time
alignment is the thing that pulls everything together. If you would measure for example both midrange drivers
playing together, instead of a nice summed response you would get something like this:

What the hell is happening at 730Hz and 1.3kHz? At those frequencies the drivers are out of phase due to being
not in time. That is – time alignment is not set correctly. So that is our next step – set the time alignment using
impulse response.
 9. Time alignment
9.1 Why time align?
Probably you noticed that in a car you have speakers all over the place and you are sitting offset to the side of
the car. The distance from your ears to the drivers vary quite a lot. What we need to do is to make sure that
soundwaves reach our ears from all the drivers at exactly the same time. If you think about an ideal music
listening experience, you probably imagine a comfy chair and two speakers in front of you. Now at home we can
make sure that sound from both speakers reaches us at the same time by putting them equal distances from
the listening position, but in a car it’s impossible, hence we need to use time delay. We need to delay the signal
from the closest driver and let the further driver to “catch up”.
Up front bass. Everyone is craving that “Up Front bass” that is soooo difficult to achieve, however it’s easy when
you understand what is “Up Front Bass”. It’s simply time alignment. When you have the sub time aligned to the
rest of the system your bass will be “up front” and “on the dash”.

9.2 Time alignment using tape measure

For basic “good enough” results you can use a simple tape measure method. I will confess something, when I
installed my first head unit with TA function, I didn’t know anything about time alignment. So, I measured the
distances to the drivers and put those numbers in the HU TA window. Needless to say, the results were not that
great…
What you need to do is measure all the distances from the middle of your head to the voice coil of each driver.
Not to the grill or the dust cap. You need to estimate the distance a bit deeper into the driver. Then find the
furthest driver and subtract that distance from all the other drivers distances. Then you will have a distance
number that needs to be converted into time using the speed of sound. If that seems a bit complicated, you can
use a handy delay calculator in Jazzi’s tuning companion (there is a separate tab at the bottom). Just enter the
physical distances to the drivers and the spreadsheet will spit out the delays in ms that you need to enter into
the DSP. Is that accurate? Some say it is, however for low frequencies time is not everything – phase is. For the
next steps we will be using a two-channel audio interface, an XLR microphone and a loopback cable. You CAN’T
do these measurements with a USB microphone such as UMIK-1, you need an XLR mic.
 9.3 Time alignment using Impulse Response
9.3.1 What is Impulse response?
In signal processing and control theory, the impulse response, or impulse response function (IRF), of a dynamic
system is its output when presented with a brief input signal, called an impulse. More generally, an impulse
response is the reaction of any dynamic system in response to some external change. In both cases, the impulse
response describes the reaction of the system as a function of time (or possibly as a function of some other
independent variable that parameterizes the dynamic behaviour of the system).
The impulse response is in essence a recording of what it would sound like in the room if you played an extremely
loud, extremely short click - something like the crack of a pistol shot. The reason for measuring the impulse
response (by more subtle means than firing a gun in the room) is that it completely characterises the behaviour
of the system consisting of the speaker(s) that were measured and the room they are in, at the point where the
measurement microphone is placed.
In other words, and in our case it’s a measurement that allows us to “see” when the cone of a driver is starting
to move and how much later it happens after the signal is played. So, it helps us to measure delay.

9.3.2 Microphone position

During impulse response measurements we will be using stationary mic. You need to figure out a way to keep
the microphone in the same position for the duration of the time alignment process. The microphone should be
“inside your head”, that is right between your ears where you usually listen to your system. You can use a mic
stand, some sticky tape or a DIY solution.

9.3.3 Loopback
This little cable probably raises the most questions when we talk about time alignment with a loopback. Most
people are not DIYers and don’t have even a soldering iron and they ask if it’s possible to buy the loopback
cable. You kind of can, but you will need some adaptors. I made my cable using an old RCA cable, some speaker
wire and a pair of 6.35mm solderable jacks from eBay. You need this cable to give HolmIMPULSE a reference
signal to which it will be comparing the measurement made with the microphone. This way, the program can tell
how late the signal from the microphone is compared to the one playing from the computer.
Loopback cable connects one of the audio interface outputs to one of the inputs AND at the same time goes to
the DSP via RCA. You can use a single RCA plug, I just had this dual cable lying around. It’s long so that I could
reach the DSP which is in the back. The XLR microphone will be connected to the other free (left) input.
If you definitely can’t make it yourself probably the easiest way is to get a male 6.35mm mono Y splitter, female
6.35mm to RCA adaptor and a mono RCA cable.

The RCA will be going directly to your DSP analog input, so you will need to do some routing in the Helix I/O
configuration window.

9.3.4 Phantom power

This is pretty simple, but might be forgotten by accident – using an XLR microphone don’t forget to enable
phantom power in the audio interface.
 9.3.5 Using HolmIMPULSE
Open HolmIMPULSE and go to “Device and Signal” tab. There choose your soundcard that you will be using for
these measurements. In my case, I will be using Focusrite Scarlett 2i2. The input channel is where your XLR
microphone is connected. The output is where you connected the loopback cable that is going to the DSP. The
“PCM-Amplitude” is your “volume knob”. You can start at -10dBFS or lower. Double check the “Sample Rate”
(to match your laptop) and “Starting frequency”. Everything else, you can leave default. You can import the
microphone calibration file, but it’s not really necessary.

Now in the options menu choose the impulse axis to be displayed as time.
Let’s go through some basics. We will need to measure “Phase”, so “Amplitude” and “Distortion” is not important
for us.
Name the empty slots from the drop-down menu to the corresponding drivers you will measure.
Set the Highpass and Lowpass accordingly to what you are measuring. Similar to EQ – one octave below
crossover and one octave above. In my case subwoofer is crossed at 45Hz, so just to have some extra, I will
measure up to 200Hz. Midbass drivers I will measure from 50Hz to 400Hz and so on.
Untick “Measurement signal” – sometimes it pops up on the graph after a measurement, we don’t need it.
Now you should start my measuring the driver that is the furthest from the listening position. In most cars that
will be the subwoofer in the back. However, in a short cabin truck furthest driver might be passenger side tweeter
or mid. Or maybe even midbass. For me it’s the sub.
Go to your DSP and mute all the channels except for the subwoofer.
Click “Measure” and enjoy the sweep. You might need to play with the channel gains on the audio interface to
have a strong enough signal as output and enough signal going into the mic.
After the sweep finishes you will see something like this. If you click and draw a rectangle around an are in the
graph, it will zoom into that rectangle. If you want to zoom out, just right-click on that graph. Now what we need
to do is to offset all the measurements. Go to the measurement “Options” and play with the “Offset (samples)”.

I have no idea what units those are, but for me 4600 works fine. What you need to offset is the point on the
impulse response where the impulse is starting to rise. That point should be at 0ms. Roughly. That exact 0ms is
not really important, as we need to know the relative delay between the drivers, not absolute values. Found that
offset? Have that number in mind.
Go to “Data analysis” tab and choose “Time zero locked” and under “Locked offset” enter your number from
before. That will offset all future measurements. Re-measure to confirm the offset is working.

I am measuring the front subwoofer, as I have one in my system. If you don’t have it – go straight to the midbass
drivers. In the DSP mute the rear sub and unmute one of the midbass drivers. Don’t forget to change the sweep
frequency range, I changed it to 50-400Hz. Then click “Measure” on the slot that was previously named “Left
Bass” (or similar). I have a measurement like this (the blue line is the new midbass measurement). As you can
see the impulse is not starting at 0ms, but rather far to the left. If you hoover your mouse over the point where
the impulse is starting to rise – at the bottom you will see the exact value in ms.
Go to Helix “Time” settings and add that delay of 6.78ms to the Left midbass driver.

Go back to HolmIMPULSE and re-measure the driver. Now pay attention to the phase graph above. Those
phase lines should match as close as possible for the drivers to be “in time” and “in phase”. Tweak the delay in
the DSP and re-measure until you have as close match as you can. It’s not going to be perfect so don’t worry
too much. If the phase is not clear to see, check if the “Gating” is set high enough. If not, drag it to the right. For
low frequencies you need to “see” a “longer” response, for tweeters even 2-3ms gating is enough.
If you have something like this where you have the impulse starting at the same time, but phase graphs are in
between each other, just switch the polarity of the driver in the DSP, measure again and they will match. In this
case the Sub and Midbass are 180 degrees out of phase and would cancel each other out if played together.

Now repeat the procedure for the other side midbass driver and match both of them. If they for some reason
don’t match, that will most likely be a problem with the slopes and EQ. If the slopes of drivers are not identical –
the phase won’t match. That’s why it’s very important to do quality EQ work.
And you work your way up. Match one of the midranges to the midbass. Pay attention to the crossover region,
between the drivers.

Due to reflections and other stuff, midranges are always the most problematic drivers. Here you can see that
the impulse response is matched, phase is ok up to 1.5kHz, but after that it’s a mess… That’s where all pass
filters might help, but that is advanced level tuning that we won’t be touching that. You can read about it and
play with all pass filters by yourself.
One note about phase matching – phase is important below ~1kHz. Above that – time becomes the main factor.
So, when time aligning tweeters to midrange drivers, we won’t be looking at the phase at all – we will be trying
to catch the beginning of the impulse. And that is the tricky part. Where does the impulse begin? Where the pre-
ringing ends?

Here you can see the tweeters are matched in time, but not in phase. And that’s ok, because we don’t hear the
phase that high up in frequency and if we move the mic even half an inch to the side – those graphs will change.
See how the beginning of the impulse for the tweeters “drifted” to 0.65ms and is not at 0ms anymore? Again,
that is ok. We have sorted the interaction between drivers, that is the most important part of TA.
 10. Checking for summing
Now finally after the time alignment is done, we can do some final checks and tweaks. Firstly, we need to level
match the pairs of drivers and the sub. We level matched the driver pairs between themselves to make sure that
they play at the same volume. Now we need to balance the whole system. After TA the driver pairs should sum
nicely, if they still don’t sum properly, you will need to do some troubleshooting.
So, measure both tweeters, both mids and both midbass drivers together.
Now it’s not really level match, it’s more match to the individual targets that we have from Jazzi’s tuning
companion, because not all drivers will have to play at the same level. Subwoofer will be highest in level. If you
choose any other house curve, your tweeters might be lower in level than the rest of the drivers. Use exactly the
same technique as level matching left and right drivers. Apply smoothing and play with the “SPL offset” until you
match the frequency responses of the drivers to the target curve. Then in the DSP link the driver pairs and adjust
the channel gain accordingly.
To link drivers in Helix DSP, just click on the red squares above driver names.

In most cases your midbass drivers will “dictate” the level of the overall house curve, as most of the time they
are the “weakest link” in the system output wise. Midbass drivers require a lot of power, they are not as sensitive
and most likely mounted in a door that has some loses due to vibration of the sheet metal. If you have a rising
slope from ~200Hz down to 30Hz, individual target curves will require the midbass to be higher in level compared
to midranges and tweeters, so you will need ether to bring mids and tweets down or increase the gain on the
midbass amp IF you have any headroom.
In my particular case the midbass drivers should be the lowest level in the system due to the front subwoofer
taking most of the beating in the midbass area. So, in my system the “weakest link” surprisingly are the tweeters.
They are quite old and not as sensitive. I had to bring everything down to match the tweeter target curve. I could
have increased the gains on my tweeter amp, however I want to keep the floor noise of the amp to the minimum
especially because it’s the highest frequencies, and to be honest, I have plenty enough output.
After the final level matching you should have something like this.

After that just enable all the drivers in the DSP and see if they sum. Just make another measurements with ~100
averages of the whole system playing. If everything is done correctly, they should sum perfectly and you should
have the centre image between you drivers on top of the dash. You can always play with EQ, and with the levels,
as chasing perfection can be a never-ending game. Now I would recommend to relax, put some music you like
and listed until your battery dies or you fall asleep. Listen to the system for a few days and come back to tweak
it if you’ll find something that you are not 100% happy with. And there will be something, I guarantee…

The end. Congratulations in finishing this whole process and till having your sanity intact!
Information sources used in making this guide (just to name a few):

• REW manual - https://www.roomeqwizard.com/help/help_en-GB/html/gettingstarted.html

• Audiotec-Fischer Sound Tuning Magazine -
https://www.google.com/url?sa=i&rct=j&q=&esrc=s&source=web&cd=&ved=0CAMQw7AJahcKEwjw4s
zO4-z7AhUAAAAAHQAAAAAQAw&url=https%3A%2F%2Ffanyv88.com%3A443%2Fhttps%2Fwww.audiotec-
fischer.de%2Fmedia%2Fpdf%2Fd3%2F47%2F70%2FATF_Sound_Tuning_Magazine-
DSP_Setup_Guide_Vol1_web.pdf&psig=AOvVaw0ak8B_Uxsyxdabee8ZE01u&ust=167068348640890
6
• Audiotec-Fischer DSP PC-Tool Knowledge Base - https://www.audiotec-fischer.de/en/knowledge-
base/DSP-PC-Tool/
• Mini DSP manual - https://www.minidsp.com/applications
• Audiofrog tuning guide -
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKE
wiT4Pm65Oz7AhUPEcAKHaOvC48QFnoECA0QAQ&url=https%3A%2F%2Ffanyv88.com%3A443%2Fhttps%2Ftestgear.audiofrog.com%
2Fwp-content%2Fuploads%2F2018%2F03%2FA-Straightforward-Stereo-Tuning-Process-and-Some-
Notes-About-Why-it-Works.pdf&usg=AOvVaw2uPIOysPGQwhV-kBn4YUJE
• JL Audio TUN training materials - https://www.youtube.com/watch?v=bxl51yZFHII&list=PLvu-
Zyg6h_QMUgphg9EBVQCSTxTIQmzWi
• Nathan Lively YT channel - https://www.youtube.com/@nathanlively
• Kyle Ragsdale YT channel – https://www.youtube.com/watch?v=KEFwrH2L3Aw&list=PLYVlZcJp-
V5Igh6-wR1hKOj3WbZ9ZH9Ei
• Audio Science Review YT channel -
https://www.youtube.com/watch?v=TwGd0aMn1wE&list=LL&index=29&t=21s
• Bennett Prescott YT channel - https://www.youtube.com/@bennettprescott
• DAmoreEngineering University Series -
https://www.youtube.com/watch?v=jc3D4A9i1oI&list=PLNMbsRIxVU1IWx89WeTUgAUTNhqgjJlew
• PSsound YT channel - https://www.youtube.com/watch?v=sCEXh0Sv5TA&list=PLemyAERC-
b21SNeLJhWnI3dsY_VMBmKKI
• ChapmanConstDesign - https://www.youtube.com/watch?v=kMslXK3MI9E&list=LL&index=30&t=11s
• Erin's Audio Corner YT channel - https://www.youtube.com/watch?v=afNZGe_LazE&list=LL&index=35
• Floyd Toole online lecture - https://www.youtube.com/watch?v=afNZGe_LazE&list=LL&index=35
• The Old Fashioned Podcast -
https://www.google.co.uk/url?sa=t&rct=j&q=&esrc=s&source=web&cd=&cad=rja&uact=8&ved=2ahUKE
wjAzISDqe37AhXIQ0EAHcq3BqwQFnoECDgQAQ&url=https%3A%2F%2Ffanyv88.com%3A443%2Fhttps%2Fopen.spotify.com%2Fshow
%2F3dkrWQmiOXrsrZUsFjo2TD&usg=AOvVaw3kNZkYml7M6A_6p0upooLq
And many many other sources…