Sound Sleuth 7.1.4 Mixing Rig

If you haven't experienced immersive audio on a 7.1.4 speaker setup or larger, you're missing out on something amazing. Setting one up can be an expensive endeavor, and many recording studios have yet to invest in it. I set out to find a low-cost, yet high-quality, solution. Here are my findings. If you already have an audio interface that can output 12 or more channels, you can build this setup for about $2,500 USD.

This is technically a DIY solution because it involves assembling multiple parts, either from kits or components I've designed. The results are top-notch and suitable for any recording studio to mix immersive content, such as Dolby Atmos.

This journey began when I built an Ambisonic microphone and started experimenting with immersive audio. I discovered various ways to decode it into binaural and other stereo methodologies. I was fascinated by the possibilities, but something always felt not fully immersive. To delve deeper, I reached out, emailed, networked, and chatted with multiple experts in the field. Everyone I met was supportive and encouraging. I also joined the Audio Engineering Society (AES), which I highly recommend.

At the tail end of Covid, I attended my first AES meeting in The Hague, Netherlands. There, I finally experienced Dolby Atmos and other immersive audio formats on multiple speaker playback systems. It was amazing. I finally understood the true essence of immersive audio. Until you have heard a 7.1.4 system, or greater, you haven't truly experienced immersive audio. I stand by that statement.

Later, I attended an AES conference in the UK focused on immersive audio at the University of Huddersfield, chaired by Professor Hyunkook Lee. This solidified my belief in the necessity of all the speakers for a genuinely immersive experience. It also provided an opportunity to ask questions, absorb knowledge, and learn from some of the world's leading experts on the subject.

As a microphone builder, I realized I couldn't advance my microphone designs immersively without a proper playback system. Hence, the Sound Sleuth 7.1.4 Mixing Rig was born!

I have a spreadsheet with all of the parts used for the build. There are a few specific things called out in the steps as your system may vary. Specifically how you mount the speakers etc. Here are links to the miscellaneous things:

Drill Bit: https://www.amazon.com/gp/product/B00AHVTO2W/?th=1

Cable Clamp: https://www.amazon.com/mankk-Mounting-Fastener-Management-M-116/dp/B0BY26BJGY/

Silicon 14 gauge wire: https://www.amazon.com/BNTECHGO-Silicone-Flexible-Strands-Stranded/dp/B017TFR6SM/

Rack Wheels: https://www.amazon.com/gp/product/B0B94NXLRB/?th=1

Rack Handles: https://www.amazon.com/gp/product/B09ZXWWVT9/

Rubber bumpers: https://www.amazon.com/gp/product/B08KDM9QY3/

Speaker Stands: https://www.amazon.com/gp/product/B076X9C7H1/

Standards for immersive audio:

https://www.itu.int/dms_pub/itu-r/opb/rep/R-REP-BS.2159-9-2022-PDF-E.pdf

RECOMMENDATION ITU-R BS.2051-3* - Advanced sound system for programme production

Plugins:

Sparta (Open Source) https://leomccormack.github.io/sparta-site/

IEM (Open Source) https://plugins.iem.at/

Fiedler Audio https://fiedler-audio.com/

Reaper: https://www.reaper.fm/

Immersive audio playback systems were first used with the Disney movie *Fantasia* in 1940. This was a three-channel system with "Left," "Center," and "Right" channels, designed to create the effect of sounds moving around the audience. In the 1970s, Quadraphonic audio with four channels emerged, but it flopped commercially. Next came what is commonly called Surround Sound or 5.1, featuring five speakers and one subwoofer for Low Frequency Effects (LFE). Essentially, 5.1 is Quadraphonic with a center channel, driven by the film industry to allow dialogue to come from the center (where the screen is) and other sounds to be placed around the audience.

I had a surround sound setup in my media room back in the day and loved it—especially for watching movies. Some audio releases on DVD, including concerts, were mixed this way, but as an audio-only format, it never really took off. 

Today, technology has improved and become more affordable, making it easier to record and playback multiple audio channels. The current bottlenecks are: Proprietary standards, user confusion, and the need to maintain backward compatibility with existing systems.

The Basics: 

What is 7.1.4? What do all those numbers mean, and what exactly is Dolby Atmos?

First, let's break down the numbers: they represent the number of audio channels and their configuration. Think of Mono as "1" and Stereo as "2". Surround sound, or 5.1, consists of five speakers and one subwoofer. In a 7.1.4 setup, there are seven speakers at ear level, one subwoofer, and four additional speakers for height. This is where the magic happens. The extra height speakers, along with the speakers arranged around you in a circle, create an immersive experience that you have to hear to fully appreciate. It is truly amazing.

So, how are these speakers set up? How are they connected, and how can you get started? Let's begin with the classic stereo setup: left and right main speakers. These are actually the foundation and the starting point. Interestingly, there is an expectation that the immersive mix can "collapse" down into stereo, much like how we used to worry about mono compatibility with stereo. These are speakers one and two, left and right. Next, we add a center channel, which is the third speaker. Then we add the LFE (Low Frequency Effects) or subwoofer, making it four speakers.

If we just added two rear channels, we would have Surround Sound or 5.1. To move to 7.1.4, we add the remaining speakers as follows:

Side Left and Side Right speakers: These are positioned to the left and right sides of the listening area, making them speakers five and six. They add depth and a sense of space to the audio experience.

Rear Left and Rear Right speakers: These are positioned behind the listening area and are designated as speakers seven and eight. They help create the surround sound effect, placing sounds behind the listener.

Front Height Left and Front Height Right speakers: These are mounted above the front left and right speakers, creating a vertical dimension in the audio. These are speakers nine and ten.

Rear Height Left and Rear Height Right speakers: These are mounted above the rear left and right speakers, completing the height dimension and making them speakers eleven and twelve.

When all these speakers are combined, you get a 7.1.4 setup: seven speakers at ear level, one subwoofer, and four height speakers. 12 Audio Channels in total.

Dolby Atmos takes advantage of this setup by allowing sound to move freely around the listener, creating a three-dimensional audio experience. Sounds can be placed and moved precisely anywhere in the room, including above the listener, making it feel as though you are truly immersed in the audio environment.

This setup allows you to experience immersive audio as it was meant to be heard, bringing movies, music, and other content to life in a way that traditional stereo setups or binaural rendering does not match.

7.1.4 Channel Order:

1. Left
2. Right
3. Center
4. LFE or Subwoofer
5. Left Side Surround (side/surround)
6. Right Side Surround (side/surround)
7. Left Rear (rear/back)
8. Right Rear (rear/back)
9. Left Front Height (height front)
10. Right Front Height (height front)
11. Left Height Rear (height rear)
12. Right Height Rear (height rear)

Understanding “Height” in Speaker Placement

When we talk about "Height" in speaker placement, there are a couple of considerations that depend on ceiling height. The goal is to position the main and surround speakers near ear level, with the height speakers positioned at twice that height and facing the mixing "sweet spot." Dolby provides extensive literature and even a design tool to assist with these setups. It's important to note that Dolby specifies different setups for consumer listening versus the mixing environment, which can be a bit confusing.

Key Factors in Speaker Setup:

1. Main Speakers: Ensure they are of high quality.
2. Surround Speakers: They should be sonically equal and full range.
3. Height Speakers: Also sonically equal and full range.
4. System Tuning: The overall system should be tuned.
5. LFE (Low-Frequency Effects): Should be matched and tuned to the overall system.

Reference Standards:

Industry-standard systems, such as those from Genelec, are used by the world's best studios. These systems can cost between $35K to well over $100K due to their advanced features like Dante and built-in DSP. However, from a DIY perspective, we can create a setup that works well, sounds amazing, and is much more affordable.

My goal was to create a fully functional 7.1.4 setup for mixing and demoing immersive audio. Let's start with speakers. I chose passive speakers, which require multi-channel amplification and an active subwoofer. I use Reaper as my DAW, which is highly configurable for immersive audio and, in my opinion, the gold standard for this purpose. Although DAWs are primarily designed for stereo, we can adapt them for immersive audio.

The Speakers:

I already had a great pair of main speakers that I built myself. As a DIY enthusiast, I considered designing my own surround speakers. I started by looking at drivers on Parts Express, my usual source. I found the C-Notes kit, which looked great on paper and was reasonably priced. After building and testing a pair, I took them to a friend's studio in the DFW area. I found they sounded really good, holding up well in an A/B comparison and were only “bested” by a pair of Adam A77x speakers. Encouraged, I ordered five more pairs.

The Amplifier: 

This one I knew I would have to build. I started with a multichannel amp board that has 6 channels and I used two of them for 12 channels total. These boards use the TDA7498 chip set and are Class D amplifiers. I quickly learned that Class D amps are not the option to go with due to self generated EMI. They work great when close to the speaker they are powering as in most powered speakers. One six channel boards by itself worked great. When multiple ones are in a system with long speaker runs I got a very weird hiss and almost a fan noise type of sound. So I started sleuthing and troubleshooting. The data sheet for the TDA mentions that if you use more than one, the clocks need to be syncronized. One master driving the rest. Now I know why. The final nail in the coffin was when I put Ferrite chokes on all the speaker cables leaving the amp case. That lowered the noise but didn't eliminate it. So I went back to an old school solution: The LM3886. I used these before and knew they would work. I wish I had started there as it would have saved me about 6 weeks of time and some coin. I found a pre laid out PCB on PCBWAY and ordered 15. They come in multiples of 5. I needed to add two things for robustness and was good to go.

The Subwoofer:

I designed the subwoofer using two Dayton Audio 10” drivers in an isobaric enclosure, a design first proposed by Henry F. Olson in the 1950s. There are several pros and cons to this design, but my primary advantage is that it halves the box volume, making the subwoofer more compact. It is built with ¾” 13-ply Baltic Birch. Using the Thiele and Small parameters for the driver I selected, and a system Q of 0.707 for the box volume, resulted in a smooth roll-off below resonance. The subwoofer is powered by a 300W plate amplifier from Dayton Audio. Wiring the two woofers in parallel reduces the impedance by half, allowing more power from the amplifier, which is rated for 300W into 4 ohms while the drivers are 8 ohms each. This setup produces a nice low end that is easy to integrate into the overall system.

I also built an 8RU Rack Case to hold the Audio Interface, the amplifier, power supplies and a few additional items. Details for each of these is in the following steps.

The build:

Starting with the crossovers. I got an assembly line going and built them in pairs. Check each PCB visually when complete and clean off the flux. You may need a little higher wattage iron then small PCB’s as there are wide traces on the board for the current carrying ability of the crossover. I used the supplied wire for the speaker connections and crimp on connectors (see parts list) to connect to the drivers. Note the Tweeter has two terminals of different sizes but isn’t labeled with a red dot for polarity. The larger terminal is Positive! I used separate red and black wires 16” long for the incoming wires. I did that so I could push them out of the Speakon connector hole to solder on the Speakon connector. 

Prep the box for use with a Speakon connector by drilling a 15/16” hole in the back panel. See the parts list for the drill bit I recommend. I centered the hole on the back panel, allowing for consistency if I mounted the speaker in its side or normally, standing up.

Assemble the boxes with wood glue and clamps per Parts Express instructions.

Finishing the boxes:

This is a bit up to you. I settled upon gray “hammered” finish paint. MDF needs to be prepped in order to get the paint to look good. There are many YouTube videos on this. Some of mine came out better than others but all look decent. My takeaways are:

1. Finish sand them with 220 grit 
2. Clean off all the dust
3. Seal the box with shellac. I used two coats
4. Sand lightly again
5. Seal the edges with body filler and sand again
6. Paint them! I did multiple coats. 
7. Push the Crossover PCB into the box through the woofer hole. It only fits one way and is a bit tight. 
8. Pull the two wires for the incoming audio out through the Speakon connector hole. Solder them to the connector. Note there are labels for “+” and “-” Pay attention to these!
9. Pull out the woofer and tweeter wires. Note the woofer uses two same size connectors and the tweeter two different, large to positive. 
10. Connect the woofer. You may have to pinch the connector lugs to make a tight connection. Mine were a bit loose. Same with the tweeter
11. Mount the woofer. Drill 1/16” pilot holes for the screws! The MDF may split if you don't.
12. Mount the tweeter. Drill 1/16” pilot holes for this as well. Tighten evenly ensuring that it is mounted flush with the front of the speaker. 
13. I used 4 rubber bumpers on each speaker

The 15/16" Drill Bit: https://www.amazon.com/gp/product/B00AHVTO2W/?th=1

Rubber bumpers: https://www.amazon.com/gp/product/B08KDM9QY3/

The Amplifier 

We are using the LM3886,

If you are unfamiliar with this chip, it is a power operational amplifier specifically designed to be used as an audio amplifier. It originally found its way into an amplifier called a gaincard by 47 laboratories. It used about $100 in parts, sold for $3000, and got rave reviews from the audiophile community. It was also so simple that the audio community started building DIY versions called “Gainclones”. I was one of those people and I can attest to how well they worked and ease of build. Let's look at the schematic and see how it works and then what is critical to making it work successfully. This is your classic Operational Amplifier gain stage. The input resistor and feedback resistor (1KΩ and 20KΩ) set the overall gain of the amp to 20, or about 26dB. This is plenty for what we are doing. Note the two capacitors at the supply rails: CS. These are our first “critical components” for proper operation. They need to be close to the LM3886 and they need to be large. We are using 1000uF caps that are also bypassed with small .1uF caps. Next up are two features that ensure the amp does not oscillate or “ring” at high frequencies. The first is a Zobel network. It is simply a small resistor and capacitor in series across the amplifier output. We are using a 3 Ω resistor and .1uF capacitor. It really helps when there is no load on the amplifier. In my case, I am building 12 channels and will use 11 of them 90% of the time so this is important. Finally, due to potentially long speaker cables, we are also including “L” and “R” from the data sheet. A small inductance in parallel with a resistor to decouple capacitance of longer speaker cable runs. We actually are winding the coil with 10 turns of magnet wire around a 10Ω resistor. Both of these are added off the board due to its design but they ensure the whole build is bullet proof. 

There is also one other component value change I made. The LM3886 supports a mute pin and the data sheet shows using a 100uF and 1K resistor (Rm and Cm) This gives a time constant of .1 seconds. The way I am powering the amp up there is a chance that one power supply rail comes up before the other – which could cause an amplifier pop in the speakers. I checked the as-built schematic for the Gainclone boards I had originally built from a company called Chipamp, which no longer is around. They used 10K for Rm which gives 1 second. I upped that to 20K. 

I didn't design the board we are using or I would have included the zobel parts and the output RL network. This also means that the 1K resistor I replaced with a 20K is labeled 1K on the board. After my time sunk into the Class D amps, and having everything else ready to go, I was looking for the fastest way to get the entire project up and running. I use PCBWay for my PCB builds and they have a shared project section where I found these: 

https://www.pcbway.com/project/shareproject/LM3886_60W_HI_FI_AUDIO_PCB_BOARD.html

Other than the Zobel and “R-L” parts, they had everything I was looking for including the ability to connect the supply rails between boards. I ordered 15 of the boards and had them built with 2 Oz copper. Meaning the traces are thicker copper, better for conductivity and power handling on the board. Then I put a BOM together for Mouser and ordered all the parts. I already had two power supplies capable of supplying 36V each so there was my +/-36V supply for all the boards. I built one board and tested it with my bench power supply and one of the Dayton Audio C-Note speakers. Worked great. Then I built the rest of the boards and mounted them to heat sinks and built two “modules” consisting of 3 heatsinks with 2 boards per heat sink. I took all the XLR and Speakon connectors out of the original chassis and mounted them onto single RU Panels with 12 XLR holes (Redco is amazing) and tested them. Not only did they work, they were super quiet, clean and sounded great. They are so quiet that when I turned them on initially, I was afraid they didn't work! If there is enough interest and I have time (so many projects in my head!) I will lay out a 12 channel version that includes everything. 

The build:

I go into detail for the PCB and wiring in the build video here:

The build:

To turn this all into a functional pro level system we need more than just the amp circuit boards. We need to think about the power supplies, turning it on and off, and electrical safety. Let's start with the power supplies. Back in the day building an amplifier required a huge transformer in the power supply. Big ones, full of copper, that weigh a lot, and cost a lot. A 500W version we would use for this weighs 9 Lbs and costs $128 on Digikey. Conversely a 36V switching supply rated for 340W costs less than $40. Two of those and we are all set. We need a dual supply of +36V and -36V. To make this, we connect the “+” of one to the “-” of the other and that becomes our circuit ground point. I made a decision to have the power supplies on all the time and switch the DC to the amplifiers when they are used. I used an old school toggle switch for the retro vibe, and two 60 Amp relays, one for the +36V and one for -36V. The relays are 12V so I also needed a voltage converter. Those are readily available these days. The one I selected is a buck converter and is adjustable. You have to set it for 12VDC. Connect it to +24VDC or more then adjust to 12VDC. 

PCB Assembly

The board is 2Oz copper and the ground plane connections will pull away a lot of heat making soldering difficult. You need to raise the soldering iron temperature and be patient. 

1. Insert the (2) 1KΩ resistors and solder
2. Insert the (2) 20KΩ resistors and solder (These are in place of the board labeled 1K and 22KΩ. See the Picture)

Warning: The Electrolytic Capacitors are polarized and must be installed correctly or they will fail catastrophically 

1. Insert the (2) 1000uF caps and solder
2. Insert the (3) 100uF caps and solder
3. Insert the (2) .1uF capacitors and solder
4. Insert the LM3886 IC and solder
5. Prep the Zobel network by connecting the 3Ω resistor and a .1uF capacitor in series – see the pictures
6. Build the RL network
7. Scrape the insulation off of ¼” of 1ft of 19 gauge magnet wire.
8. Solder that close to the body of the 10Ω on one side
9. Wrap the wire around the body of the resistor for 10 turns. (9-11 works)
10. Cut the excess wire so that you can connect the other end of the coil to the resistor.
11. Scrape ¼” of the insulation of the magnet wire
12. Connect the magnet wire to the resistor lead near the body. See the photos
13. Solder one end of the RL network to the OUT connection point on the PCB

Build the other 11 boards. ;-)

Heat Sink Preparation

Take two of the boards, hold them onto the Heatsink and mark the spots to drill and tap M3 holes. The LM3886 has to be flush with the flat surface of the heat sink for proper heat transfer. WHen you hold the PCB to the heatsink, look at that edge of the PCB, it extends just a bit more than the LM3886 and you need to take that into account when you mark the hole.

Drill the hole with a 2.5mm drill then tap with an M3 tap. A drop of cutting oil helps here. Clean the heatsink afterwards to ensure no oil remains. 

Cut two pieces of aluminum angle 12” long. We are going to mount three heats sinks to each. You will need to drill holes on each end to allow them to be screwed down on the bottom of the amp case. On the other side we will need two holes for each of the heat sinks. You want the heat sink to not touch the bottom of the amp case ¼” is a good amount of clearance. I am not giving exact measurements for you as you can use ¾” up to 1-½” angle based on what you can source. Both Lowes and Home Depot carry this. See the photos for how I did it. Note that I already had 6 heat sinks as my parts bin is extensive. 

Once you have the mounting holes sorted out we want to add the speaker wires and the input wiring prior to mounting the PCB’s to the heat sinks. 

PCB Preparation

Input connections:

1. Cut about 2-3 feet of the Mogami shielded wire – You need to go from the amp to the front panel and be able to be dressed neatly in. We will trim the excess when the XLR end is connected. 
2. Strip one end back about ½”
3. Twist the shield together  
4. Strip about ⅛” from the end of the red wire
5. Cut the White wire flush with the shield end – You should just see the red wire protruding
6. Tin the shield and the red wire
7. Tin the holes for the input connections on the PCB
8. Solder the shield to ground
9. Solder the red to input

Output connections:

1. Cut 2-3 feet of red and black 18 gauge wire 
2. Twist the two wires together (You can use zip cord style wire that is already connected together) – this helps with dressing all the cables and with some EMI protection
3. Strip the ¼” from one end of the black and red wire
4. Tin the wires
5. Solder the red to the RL filter that is floating from the PCB
6. Solder the black wire to the ground connection on the PCB

Amp Module Prep

1. Mount three of the modules to one 12 inch section of angle aluminum using M3 screws

Power Supply Preparation

1. Check the 110/220 switch on the side of the power supply to ensure it is set for 110 US. If you are in a 220V country, 220.
2. Cut 10”-12” off the end of the power cord
3. Carefully strip back the outer insulation to remove the three inner wires giving us 10”-12” of black, white, and green wires.
4. Strip the outer insulation of the remaining power cord
5. Strip ½” of the insulation from all the power supply wires
6. On one of the supplies connect the short wires:
7. Black to Line “L”
8. White to Neutral “N”
9. Green to Ground “G” 
10. On the second power supply connect the power cord ends and the loose end of the wires coming from the first supply
11. Black to Line “L”
12. White to Neutral “N”
13. Green to Ground “G” 
14. Place the power supplies on a workbench or flat surface and plug the power cord in.
15. Measure the voltage on each with a DMM and adjust the little potentiometer on the left for 36VDC.
16. Unplug the Power supplies
17. Connect the Negative(-) of one supply to the Positive(+) of the other using two 18 gauge wires or one 16 gauge wire. This becomes our DC ground for the Amplifiers.
18. Prep and attach an 18” 14-16 gauge red wire to the Positive(+) of the “+” Power supply. 
19. Prep and attach an 18” 14-16 gauge black wire to the Negative(-) of the “-” Power supply. 
20. Use a sharpie or other permanent marker and label the power supplies Positive and Negative for reference. 

Wire the control relays 

1. Prep and Connect two 12”-18” 22 gauge wires to the Buck Converter as follows
2. Red to V+ in
3. Green to V- in
4. Connect the V- out to the Relay coils as shown in the connection diagram
5. Connect the V+ out to the toggle switch using an 18-24” 22 gauge wire
6. Connect a 18-24” wire to the other side of the toggle switch and the other coil terminals on the relay 
7. Strip a ¼ inch of the insulation from the red 14-16 gauge wire coming off the power supply
8. Solder that to one of the relay Contact Terminal
9. Connect a 12” 14-16 gauge red wire to the other Contact Terminal of the relay
10. Strip ½” from the other end of the 14-16 gauge red wire. 
11. Connect this to the +36VDC Wago connector
12. Repeat this for the other relay except use 14-16 gauge black wire and connect it to the -36VDC Wago connector.
13. Plug the power supply in and ensure the relays pick up and drop out when you turn the toggle switch on and off. 

Assemble the full amplifier

1. Glue three 5 terminal Wago connectors to the bottom of the amp chassis. See the photos. I use E6000 glue. Let this dry
2. Label the case bottom + 36, -36, and Ground near each Wago connector. 
3. Mount the Power supplies to the bottom of the rack using angle brackets. Attach the angle bracket with short M3 screws to the Power Supplies (there are threaded holes for this on the Power Supply) 
4. Mount the bracket and supply to the case with small wood screws. Drilling a 1/16” pilot hole really helps.
5. Connect the power supplies common point to the Wago connectors as follows Green (or any color not red or black) to “Ground”
6. Glue the relays to the bottom of the case with E6000 glue
7. Mount the Buck Convertor with either double sided tape or with E6000 glue 
8. Let the glue dry 
9. Plug the power supply in, turn the toggle switch on, and measure the voltages at the Wago connectors to ensure you have the polarity correct. 

Warning: Reverse powering the amplifiers will cause the electrolytic capacitors to rupture and potentially explode

1. Mount the amplifier board assemblies to the bottom of the case on the opposite side from the power supplies. See the photos. Use short wood screws and drill a pilot hole
2. Dress the wires toward the back of the rack

Prepare the input connectors 

1. Mount the 12 female XLR connectors in one of the single RU panels. Use 4-40 screws and lock nuts.

Note: The connectors have a silverish tab to press in to unlock the cable when plugged in. This is the top of the panel ;-)

1. Wire the amplifier inputs to the XLR connectors as follows
2. Pin 1 Shield
3. Pin 2 Red
4. Pin 3 White (I am pre wiring this for future upgrades lol)
5. Connect Pin 1 and 3 at the XLR – this gives us our balanced to unbalanced connection
6. Cut the wires to length and dress them loosely with velcro as you go

Prepare the output connectors

1. Wire the speaker leads to the Speakon connectors as follows

Note: there is a notch at the top of the Speakon connector. Ensure they all face the same direction when installing 

1. Mount the 12 Speakon connectors to the other single RU panel. Use 4-40 screws, and lock nuts. 
2. Solder the speaker wires to the Speakon connectors as follows
3. Red to “1+”
4. Black to “1-”
5. Cut the wires to length as you go so that they can all be neatly dressed in.
6. Make sure the speaker connection matches with the input connection as you go.

Mount the two I/O panels to the two plywood side rails on the back of the rack. The XLR’s go on top and the Speakon below.

Test the amplifiers 

There are a couple ways to test these. You can do one or two channels at a time with a pair of the speakers. This can be done with any stereo source you can connect via an XLR and a pair of the speakers. You will need to ensure that the output of your source is adjustable as there is no level adjustment on the amps themselves. As this was my third amplifier design, I already had the speakers setup and placed. I connected my Audio Interface and tested that way

Finish the amplifier 

Once everything is tested, dress in the cables with tie wraps, mount the toggle switch on the front panel and drill holes in the front panel that lets you screw it to the front rack rails

On the back mount the ½” plywood piece that goes below the I/O panels. I clamped the power cord with a cable clamp to the base of the case.

Congrats, this was the hard part!

The Subwoofer

I designed the subwoofer using two Dayton Audio 10” drivers in an isobaric enclosure, a design first proposed by Henry F. Olson in the 1950s. There are several pros and cons to this design, but my primary advantage is that it halves the box volume, making the subwoofer more compact. It is built with ¾” 13-ply Baltic Birch. Using the Thiele and Small parameters for the driver I selected and a system Q of 0.707 for the box volume results in a smooth roll-off below resonance. The subwoofer is powered by a 300W plate amplifier from Dayton Audio. Wiring the two woofers in parallel reduces the impedance by half, allowing more power from the amplifier, which is rated for 300W into 4 ohms while the drivers are 8 ohms each. This setup produces a nice low end that is easy to integrate into the overall system. Construction uses butt joints with wood glue. The bottom panel is not glued and is screwed on during final assembly. 

Some tips for the build:

1. Drill holes in the lower driver plate to bring the wires through using a ⅛" drill bit, one for each wire.
2. Wire the drivers in parallel in phase, as they face the same direction.
3. Stain the first few inches of the inside bottom piece before assembly; I did not think this through initially. ;-)
4. Use rubber bumpers on both the bottom of the subwoofer and the side opposite the drivers, allowing the sub to face up if needed depending on the room.

This is a really good subwoofer and can be built standalone for adding LFE to any setup.

The Rack Case

I built an 8RU rack case specifically for this project using half a sheet of ¾” Baltic Birch. The dimensions are shown in the diagram. Note that the sides are ⅛” longer than the exact dimensions of 8 Rack Units to ensure the rails fit properly. I finished the case and the subwoofer with Ebony stain and semi-gloss polyurethane. In addition to the audio interface and amplifier, I included an 8-channel mic pre to ADAT interface as I already owned it.

At the start of this project I intended on building a 12 channel amplifier in a 2RU case. That didn't work as you will see in the amplifier section. My completed project has the amp modules and power supplies mounted to the bottom of the case. So I added a birch veneered ½” plywood shelf and ¼” front panel to the case to enclose the amps. I also replaced the I/O panel from the 2RU case I originally intended to use with two single RU rack panels that both have 12 Neutrik XLR cutouts. These became the I/O panels, one with female XLR’s for the inputs and one with Speakon connectors for the outputs. The build is straightforward, just a box, and takes less than half a sheet of Baltic Birch. We will use the second half for the subwoofer. See the pictures.

As we are building the amp into the rack case, there is a bit of crossover with the case and the amp build. To allow that we need to add four ¾” birch plywood pieces that are 2” X 7” in the corners. About 2 inches in from the back and right behind the front rack rails. This lets us put a shelf in above the amps and enclose them. There is enough scrap from the build to provide this. 

Rack Wheels: https://www.amazon.com/gp/product/B0B94NXLRB/?th=1

Rack Handles: https://www.amazon.com/gp/product/B09ZXWWVT9/

Speaker Connections

I wanted the setup to be portable for local demonstrations and tours. My mixing room is shared space, so it is not feasible to have the setup permanently installed. The solution was to make the side and rear surrounds quickly detachable. These are set up on speaker stands and connected with speaker cables using Speakon connectors.

I chose two kinds of wire:

• SOOJ heavy rubberized cable, which lays flat and is very robust, used for the surrounds, mains, and center channel. 
• Traditional 16-gauge speaker wire for the permanently installed height speakers.

The speakers themselves can be easily taken down for transport when needed. The wire and Speakon connectors cost around $400. It is well worth it for the simplicity of setup and use. There was no way I wanted to futz around with 12 pairs of banana jacks or pushing wire through screw down terminals – Not happening!

The Audio Interface:

As part of this whole endeavour, I upgraded my audio interface. I put a lot of thought into this and I bought a Focusrite Red Line 16. This provides balanced line level outputs on DB25 connectors. This lets me use two DB25 to XLR breakout cables. Which all go in the back of the rack. The output for the subwoofer plugs into XLR output #4 and connects with an XLR to RCA cable – See Rane note #110 for how to wire it. Yes, we are actually not using one of the twelve audio channels of the amplifier. The other thing the Focusrite provides is the ability to use one of the front panel knobs as a monitor level adjustment. And, link multiple channels to it. Think of it as a 12 channel volume control. Having this ability is a critical piece of an immersive audio setup. There are other solutions including standalone monitor controllers. These run anywhere from $1000 to over $6K for the Grace Designs one. There is also a just released interface specifically for this type of setup, the Audient ORIA. I have read some mixed reviews of it and it is not drastically cheaper than the Focustrite, which is now the brains of my studio.

Breakout cables https://www.amazon.com/gp/product/B0010CI52M/

Making all this come together and function as a system requires a bit of planning. Most obvious is physical placement of the speakers. Then we have Speaker Management,  which is the signal processing needed for each channel. Finally, we need to set Reaper up to support multiple channels of audio. This one is interesting as there are both the output channel requirements and track channel requirements. Just what does an immersive audio track look like? We shall see!

I have a video for the setup here:

Speaker Placement

There are a couple ways to do this and even a couple “standard” positions. I encourage you to google “7.1.4 speaker placement” You will get multiple results, all with variations of where to place them. Movie theaters vs home listening. Even in the studio. This to me, translates as there is no perfect setup and I look at these as guidelines. In the studio you want a sweet spot to mix to. In the home or theater, you need multiple people to be in the sweet spot. Question of the day: How “exact” is your stereo setup? My thoughts here are to get the speakers placed as best as you can and let's start listening! My room setup evolved from my original mixing desk with Left and Right speakers. My height front speakers are mounted wider than my main speakers while the rear height are symmetrical on the back wall. See the photos. I have seen the heights closer together and directly above the mix location. I have it on high authority that several grammy winning mixer/producers, who shall remain nameless, have ceiling mounted 8 inch coaxial speakers as height speakers. One with an 8 foot ceiling and adding in just a tiny amount of delay to make them further away. Having them is more important than what they are. Just make sure all four are equivalent. After positioning all the speakers and connecting them I did an initial test to ensure they were all working. I am using a Mac. Go to Audio Midi setup and set your speaker setup to 7.1.4. Use the built in test buttons for each speaker to make sure they are where you think they are. Initially I had my rear left/right height speakers crossed. Yet one more example of why Speakon connectors are so worth it!

There is a standard called out for this. Interestingly it calls for all the speakers to mounted the same distance from the listener or time aligned to 100uSec. Hmmm that is 3.4cm assuming speed of sound is 340M/sec. I'm not meeting that requirement, and neither are you. Which is totally OK. Ask yourself this: Do you go into the studio and take a tape measure to the stereo monitors and the mixing chair? As we shall hear, it is not as critical as one might think.

Speaker Management and DAW setup

Speaker Management

Speaker management involves ensuring all speakers are level-matched and potentially applying some EQ if needed. In my case, my two existing stereo mains have different sensitivity levels compared to the Dayton Audio C-Notes, as does the subwoofer. We can use these tools to tune the room as well, but that is beyond the scope of this document. And, as I quickly learned not really necessary. I'm expecting blowback on that statement but I stand by it.

Before we can properly manage the speakers, we need to set up Reaper for 7.1.4. I'll combine the steps since Reaper handles speaker management. Once set up, we'll create a few new templates for various ways of mixing immersive audio. Follow these steps in a new session:

Setting Up the Master Channel in Reaper

1. Open routing for the Master channel.
2. Set the Master channel to 12 tracks.
3. At the bottom for hardware send, set it to multi-channel all twelve channels.
4. Rename the hardware outputs for easier reference later. Go to “Settings” -> "Options" -> "Preferences" -> "Audio" -> "Channel Naming" and then choose the "Edit names" button for OUTs.
5. Add plugins to the Monitor effects to allow level matching and EQ if necessary.

Adding Monitor Plugins

1. Go to View, then Monitoring Effects.
2. Add 12 ReaEQs, each set to one of the 12 channels.
3. In each ReaEQ, set up the I/O matrix so that EQ one is only on channel one, EQ two on channel two, etc.
4. Label each EQ for the speaker it affects. Initially, delete all but one band for each EQ. EQ is only applied to the LFE channel.
5. Use the Plugin Pin Connector and I/O settings to ensure “pass through unmapped audio channels” is turned on. This allows the other 11 channels to pass through without processing.
6. Copy and paste the setup for the remaining channels, adjusting each to apply only to its specific channel.

Matching Levels

1. Play white noise through each speaker one at a time and use a measurement mic at your listening position.
2. Adjust the ReaEQ plugin for the Dayton C-notes, which are about 4 dB less sensitive than the main speakers.
3. Measure the subwoofer and apply about 10dB of additional gain. Use a single low pass filter set to about 200Hz for the LFE.

This could take an entire book! Along with this article, I have created a complementary video. Both serve as entry points and are not comprehensive. Immersive audio has been around for a while, but the ability to listen to and mix it is becoming more accessible. The following outlines what I have learned by talking to some of the best in the world, learning, and actually doing. I encourage you to experiment. I find myself rethinking how I record to utilize this format. Tom Dowd famously said, “Always record in stereo,” which I have always done. Now, I need to think, “Always record for immersive.”

Three Main Methods for Mixing Immersive Audio:

1. Adapting the DAW to 12 Channels:
2. Mix as usual but pan sounds around the entire soundscape. This method allows for experimentation and creative mixes but poses challenges for content delivery outside the studio.
3. Using Ambisonics:
4. Record natively in ambisonics or convert existing tracks using plugins. This method provides an immersive listening experience and can be rendered using open-source plugins. Recordings made with an ambisonic microphone can recreate the sound environment vividly. My AmbiAlice took on a whole new life with this setup! Ambisonics are tailored for Atmos beds.
5. Using Dolby Atmos:
6. This proprietary format requires special plugins and is supported by major streaming platforms. Understanding Atmos involves grasping its use of audio beds and objects.

Understanding Atmos Beds

An Atmos bed track maps out all speaker channels in your DAW setup. For example, a 7.1.4 setup has twelve channels. If played back on a different setup (e.g., 9.1.6), there might be silence on channels that don't exist in the original bed, depending on how the rendering engine is configured. You can specify how to handle these channels, leaving some final decisions to the system. If you have LFE content, it has to come through on a Bed.

 This is verbatim from Dolby’s website:

1. Objects have no access natively to the Low-Frequency Effects channel. In most circumstances, this is not a consideration, as the content that exists in the LFE channel should also exist in the main mix. (That is, do not rely on the LFE to be the “bass channel.”)
2. Bed channels map differently to different overhead speaker configurations. Overheads in an x.y.4 configuration will create a phantom center of the x.y.2 component, whereas in an x.y.6 overhead configuration, it will use the point source speakers. If you have a concern about a source in the overhead beds, try switching it to an object so you can have more control over how it renders.
3. Depending upon the position and size metadata applied to an object, objects and bed channels can be sonically identical. For instance, an object placed in the left front with size set to zero will be identical to placing the audio in the Left channel bed.
4. A Dolby Atmos input bed can be 2.0, 3.0, 5.0, 5.1, 7.0, 7.1, 7.0.2, or 7.1.2. Distribution of the bed in the Renderer is defined by the number of speakers in a particular room. For example, if a room is configured with no center speaker for playback purposes, 3.0 and greater bed configurations will try to phantom image the content in the bed Center channel to the available left and right positions. This is further accentuated in the overhead domain. Overheads in a x.y.4 configuration will create a phantom center (front/back center) of the x.y.2 component, whereas in an x.y.6 overhead configuration, it will use the point source speakers. If you have a concern about a source in the overhead beds, try switching it to an object so you can have more control over how it renders.

Atmos Rendering and Object-Based Audio

Confused yet? All of these decisions are made by the Dolby renderer engine during playback. There are settings that determine how this happens. Beds are ideal for static elements that aren’t intended to move or for sounds that you’ve recorded or mixed exactly as you want them. The real magic of Atmos comes with objects. One of my favorite statements from Dolby is, “If you have a concern about a sound, try switching it to an object.” Which is interesting as you won’t know how it will behave until it plays back on a different system with more or fewer speakers.

Atmos Objects

Atmos objects are audio tracks with metadata linked to them. This metadata includes positional data (X, Y, and Z coordinates) indicating where the sound should be placed. Upon playback, the local Dolby renderer maps that position to the actual room configuration. 

Benefits of Atmos Objects:

Positional Accuracy – If you move an Atmos object up and to the left behind you, the positional metadata will ensure it’s placed correctly during playback, regardless of the playback system’s configuration.

Adaptability – An Atmos system will try to place your sound where you intended based on the object’s metadata. For example, a mix created in a 7.1.4 setup will play back correctly on a system with more or fewer speakers. The final Atmos decoder will place each object where it should go.

Challenges with ADM Files

The metadata is exported as part of an Audio Definition Model (ADM) file. These can be quite large, as they contain all the audio and metadata for each object. An ADM file can handle up to 128 channels of audio. As of summer 2024, the ADM file format and its practical use remain the biggest hurdles. ADM files are what get exported from our DAW via the Dolby Renderer, but they aren’t designed to be portable or directly used by an end user. They can be shared via large file-sharing services or portable drives, but this is not their intended use case.

Doing all of this with Reaper:

This is the basics of how to do this with a focus on what is different from standard mixing. 

Mixing to 7.1.4 Using just what comes with Reaper

Normal mixing

Instead of sending things left or right, we have 12 possible speakers and locations. Just what does that Pan Pot do now? I mean really....? Reaper comes with a Plugin called ReaSuround Pan that lets us access all 12 speakers. And, has great features for linking and automating parameters. There some great YouTube videos on it

1. The Master Track needs to be set up for 12 Track Channels. This is done in the routing panel
2. Each individual Track needs to be setup for 12 channels as well
3. Add the ReaSuroundPan VST to each track. 

Ambisonics:

1. The Master Track needs to be set up for 12 Track Channels 
2. We need a couple plugins specific to Ambisonics. Sparta and IEM
3. For any audio that is going to be encoded into Ambisonics we need:
4. A Track with enough track channels for the Ambisonic order. 1st is four, 2nd is 9, 3rd is 16. In general # track channels = (Order+1) Squared.
5. Un-check send to Master. These are getting routed elsewhere and then finally decoded.
6. Setup one channel that has an Ambisonic Decoder on it. This is the one that will go to the Master Bus.

Atmos using Feidler Audio’s Atmos Composer. 

1. This has two plug-ins: a “Beam” and the “Composer”; each track gets a Beam and the Master gets the Composer. Note that Fiedler Audio Atmos plugins bypass the normal track routing, even the level setting on the track. You can have plugins before the Beam Plug in. For example, you can add an EQ, compressor etc. This is because most DAWs are so stereo workflow focused that they decided to take this path and ensure that it works. Even rendering the ADM file is done outside the normal Reaper file rendering.

So if you do the bottom method, you won't actually render what you are thinking. You need to have all the tracks use an Atmos Beam plugin and go to the Atmos Composer. Play around with it and you will get used to the workflow pretty quickly. We are barely scratching the surface of this.

7.1.4 sound (or higher) is amazing. I am super happy with how this all came together. Having this will totally help me design my next generation of immersive audio microphones. Here are the essential takeaways:

1. This is the only way to truly hear immersive audio. Binaural rendering is nowhere near capable of reproducing what multiple speakers can. It has a place but just like 3D movies in their current state, it isn’t ready for prime time. Plus everyone hears the same thing!
2. Speaker Placement is not as critical as one might think. Yes, you do need the space to get all the speakers placed around you, follow the guidelines. The “sweet spot” is bigger than you expect. And if you think about it, have you measured exactly where your current two monitor speakers are? Do you ensure you are sitting at the perfect equilateral triangle for placement?
3. Speaker management was not as complex as I thought it would be. I didn’t EQ any of the channels other than a low pass filter for the subwoofer on the LFE channel. I did need to match levels and that was pretty simple to accomplish. You do need a means of adjusting the output level other than the Master fader on your DAW. 
4. The Dayton Audio C-notes are an excellent low cost speaker with really good dispersion that lends itself to use in an immersive setup. 

Did I succeed in my quest? Absolutely! Everyone who has heard this is amazed.