• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Speaker placement and subwoofer integration for a small living room using MiniDSP SHD + REW

xovaan

Member
Joined
Oct 4, 2021
Messages
60
Likes
21
Hey everyone!

First, I want to give a big thanks to everyone for helping me get this far-- there's an overwhelming amount of info online for integrating subs into a 2.2 system, both in general knowledge and program interface navigation for testing. The below is what I've learned through the wonderful help of several members, includng CumSum, 3dbinCanada, and Chromatischism. Apolgies if I left someone out-- thank you too!

Because using the MiniDSP and a UMIK-1 is fairly common, there are some guides online to help. However, for a layman like me who is just starting to grasp the overwhelming landscape of audio, often I find myself questioning edge cases and not being able to dig deeply enough to find exactly what I'm looking for. While I have read a lot of material, the documentation for REW is huge. Hopefully the answers in this post help lay groundwork for people to get their system in a place they like, and in addition learn to ask the questions needed once they're comfortable with the context they acquire :)

As I receive help in getting my room in order, hopefully this thread will act as a quick guide to help people set up their own system as I learn what I'm doing through the help of others and seeing my system get to a place I'm happy with :)

Goal:
- Put speakers and couch in an optimal spot
- Integrate two subwoofers into the aforementioned two channel setup
- For both above, using REW and the UMIK-1

Room:
- Rectangular: 16.5'x9.5'x8.5' LxWxH, with a 4" divet at the far wall

Equipment:
- MiniDSP SHD + UMIK-1 microphone + REW for testing
- Two 6.5" MAOP 11 drivers in Pensil 11 cabinets as speaker towers
- Two Rythmik L12 subwoofers
- SMSL DA-9 amplifier
- A decent amount of room treatments, with the side panels on curtain rails to adjust as necessary

Pictures of room (and divet on far wall) (WIP):
IMG_6098.jpgIMG_6106.jpgIMG_6107.jpgIMG_6108.jpgIMG_6109.jpg


Steps taken:
- I used a room calculator and my ears to find the best spots for speakers and couch. This involves measuring my room's dimensions and placing things in base-rate known good starting points.
- In one of the room shots above, you can see my couch at 6 feet from the rear wall and the others is in a nearfield position. Each speaker driver center is about 15" from the side walls, flush with the side traps. Not exactly rule of primes, however my room is kind of narrow and I cannot go narrower with the speakers due to my projector screen

Where I'm at with REW and the miniDSP:
- Place the microphone in the listening position at ear height, with the microphone pointed toward where I'll be looking while listening (straight ahead)
- Configure the MiniDSP SHD matrix to have the speakers and subwofoers on the correct channels (where you plug them in the back of the SHD)
1634312086614.png

- My UMIK-1 came with a calibration file-- I uploaded it into REW for my tests
- I loaded up REW, and using REW's pink noise generator, I adjusted speaker and subwoofer volume to be a good level for the tests. For speakers, I'll select "full range", for subwoofers, I selected "sub cal" for the time being:
1634312246874.png
1634312467279.png

For this test, I'm aiming for ~80dB for speakers and subs when tested individually. For when subs and speakers are tested sub+sub, speaker+speaker, and all four at once, I turned down the volume in this pink noise window to aim for ~80dB when the time comes. The -13.00 dBFS is arbitrary as far as I understand it-- you want your amplifier and SHD to read at a -XX.XX dB on the display that gives you the ability to crank it up to a volume you enjoy after tests are done. When I put my amplifier to 50, the SHD to -14, and REW's generator to -15, I get about 80dB :)

- Now I perform a measurement test. For this, I put down purple 3m tape (delicate surface tape-- avoid other types of masking tape if you can!) in a matrix to mark room ratios and take note of the "best" places I've found for speakers and subs.
- The way I've found decent spots for my speakers and subs is through attempting two methods thus far: sweeping sine wave measurements while moving the speakers and subwoofers, and something called the Mic Measurement Method shown here:
- The MMM above has you hold your mic and pointing it at the height and direction you will be looking while listening while a pink noise from the screenshot above is being generated, and making slow circular motions while recording the average frequency through the REW menus outlined in that video. This is MUCH faster than sweeping waves, however try both and see what works best for you! (the end of the video shows comparisons to other methods too!)

Measurement window:
1634312858905.png

I muted my left and right speakers so I am only performing this measurement for my subwoofers (and I did the same for my speakers' tests too):
1634313340381.png


I then checked the levels to make sure they are the volume I wanted, selected "output", and ran a test. Here is what I've done to position my subs. I have a really bad null and no matter what I do, I can't seem to fix it :( however having two subs and everyone's help means I can at least have them help each other for most of their ranges :)

1634313257624.png


For my speakers my measurements are wild, too:

1634313631004.png



Next, I need to time align the speaker and subs to one another. This ensures they're in phase and they don't cause any weird peaks and nulls when they fire together. This is where I got pretty confused, however with CumSum's help I was able to work through it :)

To time align, I selected a speaker that I will sync my other speakers and subs to. If I do this correctly, my other speakers will accurately show how much distance between them and my reference speaker there is. Here is a quick guide that I will outline below: https://www.minidsp.com/applications/auto-eq-with-rew/speaker-time-alignment

First, to set my reference speaker (in this example I will sync everything to my left speaker), in my channel matrix I will only run the left speaker:

1634313975618.png


And in my measurement window, I put my options to be the following:
1634314027806.png

1634314232783.png

1634314286537.png

"Ref output" is my reference speaker. This will remain "L" throughout my testing, with "Output" changing between L sub, R sub, and R speaker. When I run my subs and right speaker, Output will remain "R" and I will run these speakers and subs through my channel matrix's right channels. I do this because I need the chirp from the reference speaker to start as the measurement points for the others. Important to note that I muted the channels so only a single sub and speaker were running during the measurements.

1634314732496.png


I ran the test relatively loudly, per CumSum's suggestion, >80dB, in order to get things to register correctly. I heard a high pitched chirp followed by my sine sweep. If this registers correctly, I will see some notes in my left side measurement window (with my graph tabs) that will have distances like this:

1634314548086.png


My left reference speaker is 0.0867ms, which is below 1ms and works for matching my other speakers. The distances above are actually spot on from where my speakers are in relation to the reference speaker when I use a tape measure to measure the mic to each.

For the above, I wanted to match the timing delay / lag from my other speakers to my furthest speaker/sub, which L Sub. When you lag the speaker timing, it will match the frequency output, otherwise you'll get weird peaks and nulls. In the above, my subwoofer is 0.9088ms from my left speaker, so I did some math to zero them all in. I suck at math so bear with me and if anyone could check my math that would be amazing:

1. I added L Sub's delay to every other channel and make L Sub's timing in that chart 0ms since it's furthest away
2. My reference speaker is now .9955ms (.0867+.9088)
3. Now, with .9088 added to each channel, I added and subtracted the right speaker and sub to match my reference speaker:

Left Speaker: .0867 + .9088 = .9955ms, my new reference point
Right Speaker: -.0796+.9088-.9955 = .1663ms to add to reach .9955, so -.0796+.9088+.1663 = 1.1547ms total
Left Sub: 0ms
Right Sub: -.2559+.9088-.9955 = .3426ms to reach .9955, so 1.2514ms total

Final measurements I need to add to the SHD outputs:
L Speaker: .9955ms
R Speaker: 1.1547ms
L Sub: 0ms
R Sub: 1.2514ms


1634316033999.png


Now I changed "Timing" in the measurement window to "no timing reference" and put my channel matrix back to the regular outputs and my measurement sweep from 0 to 300, then ran my left speaker, right speaker, and to see if anything is wonky. If I am timed correctly, the sum of my L+R speaker measurements should be roughly the highest points of my left and right speaker measurements throughout:

1634313749349.png


I then applied a high pass filter to my speakers at the 80hz / 24dB per octave mark, and did the same with a low pass filter for my subwoofers:
1634316355475.png

1634316414538.png


And when I ran everything together, everything looks to be in order:
1634316568717.png



Where I am at now and what I need help with:

Things sound pretty good after running these timing delays and speaker/sub positions through Dirac Live for my final EQ, however:

1. I'm not sure if I'm missing any crucial steps above?
2. I'm not sure what to do about these wonky and quite large nulls, especially given my space restrictions. I've tried a lot of different couch and speaker placements and tried moving my panels around and nothing helps these very large nulls
 
Last edited:

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,653
Likes
5,491
Location
Brussels, Belgium
Why not let Dirac determine the optimal delay for all the outputs? Check my guide here:
 
OP
xovaan

xovaan

Member
Joined
Oct 4, 2021
Messages
60
Likes
21
Also how did you determine that 80Hz is the best crossover frequency for you?
For those tests, I did 80hz, however for my current nearfield (closer?) setup, I did 60hz for the speakers and 100hz for the subwoofers which seems to work the best. However I believe someone recommended that I still need to fiddle with the timing a bit on the subs to see if I can fill the nulls

Why not let Dirac determine the optimal delay for all the outputs? Check my guide here:
Oh wow, very cool, thank you! I'll give this a shot as well. I assume this is done after I move my subs around to get good frequency response measurements from each?
 
Last edited:

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,653
Likes
5,491
Location
Brussels, Belgium
Oh wow, very cool, thank you! I'll give this a shot as well. I assume this is done after I move my subs around to get good frequency response measurements from each?

I would put the listening position roughly 38% away from the front wall if you have a (regular) rectangle shaped room, and if you have one listening position yeah moving the subwoofers around a bit first is ofcourse optimal.

If you can, put one subwoofer 25% off the ground and the second 75% off the ground. As usually the floor-ceiling bounce is the worst. if you can't do that, put the subwoofers on the left edge and right edge of the listening position.

Basically you want to use the two subwoofers to minimize the seat to seat variation by placing them within the axis that your listening positions extends within. Quite a mouthful, hopefully this page from Floyd Toole's book should help.

1634320286475.png
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,653
Likes
5,491
Location
Brussels, Belgium
For those tests, I did 80hz, however for my current nearfield (closer?) setup, I did 60hz for the speakers and 100hz for the subwoofers which seems to work the best. However I believe someone recommended that I still need to fiddle with the timing a bit on the subs to see if I can fill the nulls
What you want to do is determine the acoustical low frequency 'crossover' of your speaker (it's called LF-Slope or LF roll-off).

with that sort of information you can 'stack' an electric/digital crossover with the SHD on top of the speaker's acoustic crossover and integrate your subwoofer in a way that generates a flat an-echoic response.

to determine the accoustic LF-Slope of your speaker you need to do near-field bass measurements (incredibly easy and simple). Guide below, make sure to seal any ports in the speaker first as we don't want the port's 'dirty' bass in our final system. you bought the subwoofers for a reason :). near-field bass measurements are also an incredibly easy and effictive way to gain match your subwoofers to your speakers, if you don't know what that means let me know.


once you have the anechoic bass response of your speaker, use REW's new speaker driver feature to determine the exact LF roll-off of your speaker.

1634321083771.png


For example, my speakers in the living room have a Butter-worth 12db/oct accoustic low frequency roll-off @108Hz

index.php


So what i did was stack a digital BW 12dB/oct @ 108 Hz on top using the SHD, resulting in a Link-witz rilery (LR) accoustic LF roll-off of 24db/oct @ 108 Hz.

So all i did was put a 'matching' digital low-pass of LR 24db/oct @ 108Hz on the SHD on my subwoofers.

The end result is this:

1634321353763.png


After you have a flat anechoic response, you run the Dirac calibration. I'm sorry for bombarding you with so much information but idk when my subwoofer integration guide will be finished so i'm trying to help you in the best way i can.
 
OP
xovaan

xovaan

Member
Joined
Oct 4, 2021
Messages
60
Likes
21
I would put the listening position roughly 38% away from the front wall if you have a (regular) rectangle shaped, and if you have one listening position yeah moving the subwoofers around a bit first is ofcourse optimal.

If you can, put one subwoofer 25% off the ground and the second 75% off the ground. As usually the floor-ceiling bounce is the worst. if you can't do that, put the subwoofers on the left edge and right edge of the listening position.

Basically you want to use the two subwoofers to minimize the seat to seat variation by placing them within the axis that your listening positions extends within. Quite a mouthful, hopefully this page from Floyd Toole's book should help.

View attachment 159235
Not a mouthful, really appreciate the insight--

38% is ~6'4"-- currently my couch is at 7'6", about 45%. If I move it closer then I'm about 3 feet from my 100" projector screen, hence why my other position was 38% from the rear wall as an alternative to test. Here's my setup at 7'6" for reference:

IMG_6110.jpg

For 75/25 sub placement, which sub should I do this for? I have spare speaker stands that are 24" high (23.5% of my 8.5' ceiling) and I can combine them both to support a single sub. I can also place a speaker on the console (28", so 27.4%), however when I did so the front right sub didn't measure much differently than when it was located in its current front right position on the ground
 
OP
xovaan

xovaan

Member
Joined
Oct 4, 2021
Messages
60
Likes
21
What you want to do is determine the acoustical low frequency 'crossover' of your speaker (it's called LF-Slope or LF roll-off).

with that sort of information you can 'stack' an electric/digital crossover with the SHD on top of the speaker's acoustic crossover and integrate your subwoofer in a way that generates a flat an-echoic response.

to determine the accoustic LF-Slope of your speaker you need to do near-field bass measurements (incredibly easy and simple). Guide below, make sure to seal any ports in the speaker first as we don't want the port's 'dirty' bass in our final system. you bought the subwoofers for a reason :). near-field bass measurements are also an incredibly easy and effictive way to gain match your subwoofers to your speakers, if you don't know what that means let me know.


once you have the anechoic bass response of your speaker, use REW's new speaker driver feature to determine the exact LF roll-off of your speaker.

View attachment 159237

For example, my speakers in the living room have a Butter-worth 12db/oct accoustic low frequency roll-off @108Hz

index.php


So what i did was stack a digital BW 12dB/oct @ 108 Hz on top using the SHD, resulting in a Link-witz rilery (LR) accoustic LF roll-off of 24db/oct @ 108 Hz.

So all i did was put a 'matching' digital low-pass of LR 24db/oct @ 108Hz on the SHD on my subwoofers.

The end result is this:

View attachment 159238

After you have a flat anechoic response, you run the Dirac calibration. I'm sorry for bombarding you with so much information but idk when my subwoofer integration guide will be finished so i'm trying to help you in the best way i can.
Never apologize, this is all really interesting and helpful and I am going to try everything today that you're outlining :)

Reading through that guide now and will be back with questions soon!
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,653
Likes
5,491
Location
Brussels, Belgium
Not a mouthful, really appreciate the insight--

38% is ~6'4"-- currently my couch is at 7'6", about 45%. If I move it closer then I'm about 3 feet from my 100" projector screen, hence why my other position was 38% from the rear wall as an alternative to test. Here's my setup at 7'6" for reference:

38% is usually a starting point in studios to minimize the influence of the room axial modes that extends between the front wall and the rear wall. Ofcourse at some point if you're watching movies in the room you have to pick a compromise, studios don't have to worry about projector screens ;).

45% is okay, 50% is bad because you would be dead ass in the centre of the room mode (check the book page I screenshoted above).

What you need to keep in mind is that there are 3 axes in the room, X , Y and Z. each one of these axes generates a room mode and messes up your bass response a bit.

38% away from the front wall would minimize Y issues, putting the subwoofers on the edges of X or 25%/75% of X would solve X. leaving Z unresolved. HOWEVER, all the people in the room should (hopefully) be of similar height when they sit down in their listening positions so the Z axis mode should be dealt with EQ/Dirac as there is little seat to seat variation.
For 75/25 sub placement, which sub should I do this for? I have spare speaker stands that are 24" high (23.5% of my 8.5' ceiling) and I can combine them both to support a single sub. I can also place a speaker on the console (28", so 27.4%), however when I did so the front right sub didn't measure much differently than when it was located in its current front right position on the ground

I thought you only had two subs? so one would be 75 and the second would be 25. I would forget about dealing with the Z axis for now and try to put the subwoofers across the X axis or the Y axis for now. Whichever one (Y or X) that you think is the most problematic (highest seat to seat variation).
 

Tom C

Addicted to Fun and Learning
Forum Donor
Joined
Jun 16, 2019
Messages
854
Likes
746
Location
USA
Great many thanks to the OP and all contributors. I started following this thread when it was part of the Rhythmik L12 thread. Great idea to break it out.
Two questions:
- For initial placement of the first sub, I’ve read about the technique of placing the sub at the listening position, then taking measurements with the mike placed at different candidate locations in the room. When you find what works best, that’s where the sub goes. Saves having to move a 100 lb. sub around to different trial and error locations. Does anyone have experience with this, and can support or refute its validity? Can the method be extrapolated to placing subs #2, #3, etc.?
- I haven’t seen any discussions of Single Source to Sink (2 subs) or Controlled Acoustic Bass System (4 subs). Has anyone tried these at home, and are they worthwhile?
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,653
Likes
5,491
Location
Brussels, Belgium
- For initial placement of the first sub, I’ve read about the technique of placing the sub at the listening position, then taking measurements with the mike placed at different candidate locations in the room. When you find what works best, that’s where the sub goes. Saves having to move a 100 lb. sub around to different trial and error locations. Does anyone have experience with this, and can support or refute its validity?

This works, but good luck convincing someone to do the crawl :D, as you want yourself to be the 'brains' on the compter to interpret the Frequnecy response as it changes.

But honestly, with one subwoofer, i would just place it where it would work aesthetically in the room (close to a wall) and just let EQ handle the rest.
 

samysound

Active Member
Joined
May 3, 2020
Messages
222
Likes
151
Location
USA
Hey everyone!

First, I want to give a big thanks to everyone for helping me get this far-- there's an overwhelming amount of info online for integrating subs into a 2.2 system, both in general knowledge and program interface navigation for testing. The below is what I've learned through the wonderful help of several members, includng CumSum, 3dbinCanada, and Chromatischism. Apolgies if I left someone out-- thank you too!

Because using the MiniDSP and a UMIK-1 is fairly common, there are some guides online to help. However, for a layman like me who is just starting to grasp the overwhelming landscape of audio, often I find myself questioning edge cases and not being able to dig deeply enough to find exactly what I'm looking for. While I have read a lot of material, the documentation for REW is huge. Hopefully the answers in this post help lay groundwork for people to get their system in a place they like, and in addition learn to ask the questions needed once they're comfortable with the context they acquire :)

As I receive help in getting my room in order, hopefully this thread will act as a quick guide to help people set up their own system as I learn what I'm doing through the help of others and seeing my system get to a place I'm happy with :)

Goal:
- Put speakers and couch in an optimal spot
- Integrate two subwoofers into the aforementioned two channel setup
- For both above, using REW and the UMIK-1

Room:
- Rectangular: 16.5'x9.5'x8.5' LxWxH, with a 4" divet at the far wall

Equipment:
- MiniDSP SHD + UMIK-1 microphone + REW for testing
- Two 6.5" MAOP 11 drivers in Pensil 11 cabinets as speaker towers
- Two Rythmik L12 subwoofers
- SMSL DA-9 amplifier
- A decent amount of room treatments, with the side panels on curtain rails to adjust as necessary

Pictures of room (and divet on far wall) (WIP):
View attachment 159194View attachment 159195View attachment 159196View attachment 159197View attachment 159198


Steps taken:
- I used a room calculator and my ears to find the best spots for speakers and couch. This involves measuring my room's dimensions and placing things in base-rate known good starting points.
- In one of the room shots above, you can see my couch at 6 feet from the rear wall and the others is in a nearfield position. Each speaker driver center is about 15" from the side walls, flush with the side traps. Not exactly rule of primes, however my room is kind of narrow and I cannot go narrower with the speakers due to my projector screen

Where I'm at with REW and the miniDSP:
- Place the microphone in the listening position at ear height, with the microphone pointed toward where I'll be looking while listening (straight ahead)
- Configure the MiniDSP SHD matrix to have the speakers and subwofoers on the correct channels (where you plug them in the back of the SHD)
View attachment 159204
- My UMIK-1 came with a calibration file-- I uploaded it into REW for my tests
- I loaded up REW, and using REW's pink noise generator, I adjusted speaker and subwoofer volume to be a good level for the tests. For speakers, I'll select "full range", for subwoofers, I selected "sub cal" for the time being:
View attachment 159205View attachment 159206
For this test, I'm aiming for ~80dB for speakers and subs when tested individually. For when subs and speakers are tested sub+sub, speaker+speaker, and all four at once, I turned down the volume in this pink noise window to aim for ~80dB when the time comes. The -13.00 dBFS is arbitrary as far as I understand it-- you want your amplifier and SHD to read at a -XX.XX dB on the display that gives you the ability to crank it up to a volume you enjoy after tests are done. When I put my amplifier to 50, the SHD to -14, and REW's generator to -15, I get about 80dB :)

- Now I perform a measurement test. For this, I put down purple 3m tape (delicate surface tape-- avoid other types of masking tape if you can!) in a matrix to mark room ratios and take note of the "best" places I've found for speakers and subs.
- The way I've found decent spots for my speakers and subs is through attempting two methods thus far: sweeping sine wave measurements while moving the speakers and subwoofers, and something called the Mic Measurement Method shown here:
- The MMM above has you hold your mic and pointing it at the height and direction you will be looking while listening while a pink noise from the screenshot above is being generated, and making slow circular motions while recording the average frequency through the REW menus outlined in that video. This is MUCH faster than sweeping waves, however try both and see what works best for you! (the end of the video shows comparisons to other methods too!)

Measurement window:
View attachment 159207
I muted my left and right speakers so I am only performing this measurement for my subwoofers (and I did the same for my speakers' tests too):
View attachment 159209

I then checked the levels to make sure they are the volume I wanted, selected "output", and ran a test. Here is what I've done to position my subs. I have a really bad null and no matter what I do, I can't seem to fix it :( however having two subs and everyone's help means I can at least have them help each other for most of their ranges :)

View attachment 159208

For my speakers my measurements are wild, too:

View attachment 159210


Next, I need to time align the speaker and subs to one another. This ensures they're in phase and they don't cause any weird peaks and nulls when they fire together. This is where I got pretty confused, however with CumSum's help I was able to work through it :)

To time align, I selected a speaker that I will sync my other speakers and subs to. If I do this correctly, my other speakers will accurately show how much distance between them and my reference speaker there is. Here is a quick guide that I will outline below: https://www.minidsp.com/applications/auto-eq-with-rew/speaker-time-alignment

First, to set my reference speaker (in this example I will sync everything to my left speaker), in my channel matrix I will only run the left speaker:

View attachment 159212

And in my measurement window, I put my options to be the following:
View attachment 159213
View attachment 159214
View attachment 159215
"Ref output" is my reference speaker. This will remain "L" throughout my testing, with "Output" changing between L sub, R sub, and R speaker. When I run my subs and right speaker, Output will remain "R" and I will run these speakers and subs through my channel matrix's right channels. I do this because I need the chirp from the reference speaker to start as the measurement points for the others. Important to note that I muted the channels so only a single sub and speaker were running during the measurements.

View attachment 159217

I ran the test relatively loudly, per CumSum's suggestion, >80dB, in order to get things to register correctly. I heard a high pitched chirp followed by my sine sweep. If this registers correctly, I will see some notes in my left side measurement window (with my graph tabs) that will have distances like this:

View attachment 159216

My left reference speaker is 0.0867ms, which is below 1ms and works for matching my other speakers. The distances above are actually spot on from where my speakers are in relation to the reference speaker when I use a tape measure to measure the mic to each.

For the above, I wanted to match the timing delay / lag from my other speakers to my furthest speaker/sub, which L Sub. When you lag the speaker timing, it will match the frequency output, otherwise you'll get weird peaks and nulls. In the above, my subwoofer is 0.9088ms from my left speaker, so I did some math to zero them all in. I suck at math so bear with me and if anyone could check my math that would be amazing:

1. I added L Sub's delay to every other channel and make L Sub's timing in that chart 0ms since it's furthest away
2. My reference speaker is now .9955ms (.0867+.9088)
3. Now, with .9088 added to each channel, I added and subtracted the right speaker and sub to match my reference speaker:

Left Speaker: .0867 + .9088 = .9955ms, my new reference point
Right Speaker: -.0796+.9088-.9955 = .1663ms to add to reach .9955, so -.0796+.9088+.1663 = 1.1547ms total
Left Sub: 0ms
Right Sub: -.2559+.9088-.9955 = .3426ms to reach .9955, so 1.2514ms total

Final measurements I need to add to the SHD outputs:
L Speaker: .9955ms
R Speaker: 1.1547ms
L Sub: 0ms
R Sub: 1.2514ms


View attachment 159220

Now I changed "Timing" in the measurement window to "no timing reference" and put my channel matrix back to the regular outputs and my measurement sweep from 0 to 300, then ran my left speaker, right speaker, and to see if anything is wonky. If I am timed correctly, the sum of my L+R speaker measurements should be roughly the highest points of my left and right speaker measurements throughout:

View attachment 159211

I then applied a high pass filter to my speakers at the 80hz / 24dB per octave mark, and did the same with a low pass filter for my subwoofers:
View attachment 159221
View attachment 159222

And when I ran everything together, everything looks to be in order:
View attachment 159223


Where I am at now and what I need help with:

Things sound pretty good after running these timing delays and speaker/sub positions through Dirac Live for my final EQ, however:

1. I'm not sure if I'm missing any crucial steps above?
2. I'm not sure what to do about these wonky and quite large nulls, especially given my space restrictions. I've tried a lot of different couch and speaker placements and tried moving my panels around and nothing helps these very large nulls
Hi, thanks for organizing this info into a guide. Did you try a configuration with both subs on for both channels? I wonder if you you might get smoother bass response across your listening area with both subs on for both channels.
 
OP
xovaan

xovaan

Member
Joined
Oct 4, 2021
Messages
60
Likes
21
In that anechoic guide, here are my results for my aligned sum after placing the mic flush with the port, .25" from the woofer, taking measurements with the same untouched volume settings for each, adjusting so the port's 15hz is flush with the woofer's, and then doing all the other stuff and getting the blue anechoic line:

1634326858600.png


And when I look at the driver tool in your settings, it looks like my rolloff is 40hz, which if I recall correctly is roughly in-line with what DIYaudio's creators of this cabinet and driver combination said is the bass extension surprisingly:

1634327576001.png


Hopefully I did this correctly-- because the rolloff is pretty mild, would that mean I set my crossover to 30hz at 12dB/octave? Not entirely certain what you mean when you say:

"So what i did was stack a digital BW 12dB/oct @ 108 Hz on top using the SHD, resulting in a Link-witz rilery (LR) accoustic LF roll-off of 24db/oct @ 108 Hz.

So all i did was put a 'matching' digital low-pass of LR 24db/oct @ 108Hz on the SHD on my subwoofers."
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,653
Likes
5,491
Location
Brussels, Belgium
In that anechoic guide, here are my results for my aligned sum after placing the mic flush with the port, .25" from the woofer, taking measurements with the same untouched volume settings for each, adjusting so the port's 15hz is flush with the woofer's, and then doing all the other stuff and getting the blue anechoic line:

View attachment 159249

And when I look at the driver tool in your settings, it looks like my rolloff is 40hz, which if I recall correctly is roughly in-line with what DIYaudio's creators of this cabinet and driver combination said is the bass extension surprisingly:

View attachment 159253

Hopefully I did this correctly-- because the rolloff is pretty mild, would that mean I set my crossover to 30hz at 12dB/octave? Not entirely certain what you mean when you say:

"So what i did was stack a digital BW 12dB/oct @ 108 Hz on top using the SHD, resulting in a Link-witz rilery (LR) accoustic LF roll-off of 24db/oct @ 108 Hz.

So all i did was put a 'matching' digital low-pass of LR 24db/oct @ 108Hz on the SHD on my subwoofers."

it's going to be way more simpler if you just put few socks in the ports of the speakers and just have faith in the subwoofers doing what you purchased them to do.

The driver's measurement looks a little off. how far was the microphone from the dust cap of the woofer?
 
OP
xovaan

xovaan

Member
Joined
Oct 4, 2021
Messages
60
Likes
21
38% is usually a starting point in studios to minimize the influence of the room axial modes that extends between the front wall and the rear wall. Ofcourse at some point if you're watching movies in the room you have to pick a compromise, studios don't have to worry about projector screens ;).

45% is okay, 50% is bad because you would be dead ass in the centre of the room mode (check the book page I screenshoted above).

What you need to keep in mind is that there are 3 axes in the room, X , Y and Z. each one of these axes generates a room mode and messes up your bass response a bit.

38% away from the front wall would minimize Y issues, putting the subwoofers on the edges of X or 25%/75% of X would solve X. leaving Z unresolved. HOWEVER, all the people in the room should (hopefully) be of similar height when they sit down in their listening positions so the Z axis mode should be dealt with EQ/Dirac as there is little seat to seat variation.


I thought you only had two subs? so one would be 75 and the second would be 25. I would forget about dealing with the Z axis for now and try to put the subwoofers across the X axis or the Y axis for now. Whichever one (Y or X) that you think is the most problematic (highest seat to seat variation).
Oops, should have clarified: I have two subs, one against the wall in that picture and the other stucked by the right speaker. I don't have much XY room to work with, however lifting one sub off the ground to 25% is doable. I don't really know how I'd get the other sub to 75% though haha

I suppose a good compromise is 38% from the back wall? This would mean I could possibly squeeze the subs into a different XY position with the extra front space
it's going to be way more simpler if you just put few socks in the ports of the speakers and just have faith in the subwoofers doing what you purchased them to do.

The driver's measurement looks a little off. how far was the microphone from the dust cap of the woofer?
Probably 0.5", with the gain reduced via the measurement tool. Here's with it the foam nearly touching (which is about 0.25" without the mic foam, in line with what the guide recommends):

1634328828745.png
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,653
Likes
5,491
Location
Brussels, Belgium
(which is about 0.25" without the mic foam, in line with what the guide recommends):

View attachment 159256
looks much better! btw you can remove the foam, it's there incase you're measuring outside so gentle wind and dust wouldn't bother you.

for the new measurement, try to figure out LF-Slope of the driver (ignore the port measurement for now) but don't forget the baffle step adjustment step!
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,653
Likes
5,491
Location
Brussels, Belgium
I suppose a good compromise is 38% from the back wall? This would mean I could possibly squeeze the subs into a different XY position with the extra front space
just keep the listening position where it is now then, i think it's a better compromise to stay closer to the speakers (maximize direct sound to reflected sound ratio) than whatever improvement that might occur in the sub-bass response.
Oops, should have clarified: I have two subs, one against the wall in that picture and the other stucked by the right speaker. I don't have much XY room to work with, however lifting one sub off the ground to 25% is doable. I don't really know how I'd get the other sub to 75% though haha

mull it over for a while then :) I feel like i have bombarded you with a lot to do already.
 
OP
xovaan

xovaan

Member
Joined
Oct 4, 2021
Messages
60
Likes
21
Update: Had a movie night with my friends and watching movies up close is actually pretty radical so I moved my setup to around 6.5' from the front wall (as close to 38% as I could get).

Been playing around with phase alignment / timing and found that 100hz/12dB subwoofer crossover and 80hz/24dB speaker crossover produces this:

Left speaker + left sub:
1634424839980.png


Right speaker + right sub:
1634424863625.png


Left speaker, both subs:

1634424888417.png


Right speaker, both subs:

1634424912190.png


You can probably see in the pic that I have a black hole around 60hz that no phase alignment fixes when playing the right sub and speaker at the same time. From what I've read, frequencies under 80hz are directionless, so the two subs working together would help fill this?

Also, in my channel matrix, this is my current routing and I was wondering if it's correct:

1634425200786.png


Or should my matrix look like this?

1634425226188.png
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
5,653
Likes
5,491
Location
Brussels, Belgium
Update: Had a movie night with my friends and watching movies up close is actually pretty radical so I moved my setup to around 6.5' from the front wall (as close to 38% as I could get).

Been playing around with phase alignment / timing and found that 100hz/12dB subwoofer crossover and 80hz/24dB speaker crossover produces this:

Left speaker + left sub:
View attachment 159497

Right speaker + right sub:
View attachment 159498

Left speaker, both subs:

View attachment 159499

Right speaker, both subs:

View attachment 159500

You can probably see in the pic that I have a black hole around 60hz that no phase alignment fixes when playing the right sub and speaker at the same time. From what I've read, frequencies under 80hz are directionless, so the two subs working together would help fill this?

Also, in my channel matrix, this is my current routing and I was wondering if it's correct:

View attachment 159501

Or should my matrix look like this?

View attachment 159502

Well running the subs in mono should reduce seat to seat variation. Indeed sub-bass is difficult to localise so there is little point in running it in stereo like you’re doing now.

Any luck on figuring out the low frequency slope of your speaker? You should really have matching high pass and low pass filter frequencies btw. Otherwise the response will not be flat.
 
OP
xovaan

xovaan

Member
Joined
Oct 4, 2021
Messages
60
Likes
21
For the low frequency slope, after doing those tests, I found that it was about 70hz, however running various timings with the measurement tool through REW (and matching them with sweeping tests) I found that I got wild results if I deviated from my 100hz sub/80hz speaker crossover.

I made a matrix of combinations of frequencies and octave rolloffs and did sweeps at 70, 80, 100, and 120 for the subs and speakers for LR12 and LR24 and found that (after time adjustments) the screenshots above produced the flattest results. Not sure if I'm doing something incorrectly though?

Left channel:

1634428943342.png


Right channel:

1634428974477.png
 
Last edited:
Top Bottom