• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Sub crossover frequency vs. room nodes

sejarzo

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
961
Likes
1,066
Curious if anyone has faced this same issue and has any advice...

I have a basement room that's unfortunately exactly 7 feet high and exactly 21 feet deep, with a post nearly in the middle that puts the only reasonable listening position dead center. Those dimensions create a very strong mode at 80-81 Hz at the listening position.

Before I implement Dirac with a minidsp DDRC-24, does having a strong mode at 80 Hz suggest that I would get a better end result by crossing over at either about 70 or 90 Hz? My mains will run flat to just about 60 Hz.
 
Last edited:

Tom C

Major Contributor
Joined
Jun 16, 2019
Messages
1,501
Likes
1,370
Location
Wisconsin, USA
How many subs? Do you have flexibility in where you can place the subs? And the placement of the mains, where will they be?
 
OP
sejarzo

sejarzo

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
961
Likes
1,066
How many subs? Do you have flexibility in where you can place the subs? And the placement of the mains, where will they be?

One Hsu VTF-3, no real flexibility has to go in left front corner due to location of doors to closets, stairs coming down, and door to workshop. Has a 4" thick x 2 ft wide x 6 ft high rigid fiberglass trap behind it. Mains are Paradigm Studio 40v3, 20 inches off the 17-ish wide wall, each 66" off the side walls, 78" apart. I wish I could bring them out more but then they are too close to the sofa.

That post in the center really complicates matters in addition to the the length being exactly 3x the height. The resulting 80-81 Hz mode is massive.
 
Last edited:

antcollinet

Master Contributor
Joined
Sep 4, 2021
Messages
7,408
Likes
12,291
Location
UK/Cheshire
Far from an expert here - but would it work to allow your mains to go down to about 70Hz, and allow the sub to go up to 90.

You've then got 3 drivers going at the 80Hz mode with the possibility of filling it in?
 
OP
sejarzo

sejarzo

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
961
Likes
1,066
Far from an expert here - but would it work to allow your mains to go down to about 70Hz, and allow the sub to go up to 90.

You've then got 3 drivers going at the 80Hz mode with the possibility of filling it in?

Interesting. Unsure if I can low pass the sub at 90 and high pass the mains at 70. I bought the DDRC-24 a few years ago, before having to move my elderly parents into senior living, soon followed by moving my father in law in with us...and thus being forced to use the space for storage until recently.

The room was originally set up almost 20 years ago as shown. It had horrible slap echo until I used a lot of absorption, but with the right main close to the side wall, and the left main 13 feet or so from the wall on that side, I could never get anything to sound right. So now I am just trying to see if turning it all 90 degrees to the right works any better for the next couple of years before we downsize. Ergo...I don't want to spend much money on a second sub, etc.

HT and Music Room C.jpg
 
Last edited:
D

Deleted member 48726

Guest
Well, have you tried it crossed at 80 Hz and cutting the peak in Dirac?
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,834
Likes
16,496
Location
Monument, CO
Since sub and main speakers are essentially co-located and you can't move the listening position it is unlikely EQ will fix the problem of a deep null. It comes from physical dimensions. The usual solutions are to either move the listening position or add a sub (or more) placed to compensate ("drive") the null. Looking at the picture, could you move the sub to the left end of the couch to see if that helps?
 
OP
sejarzo

sejarzo

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
961
Likes
1,066
Well, have you tried it crossed at 80 Hz and cutting the peak in Dirac?

Haven't gotten there yet, still have to download the new firmware, etc. Also just got the room cleaned out of the old junk and furniture moved.

(LOL I did get rid of that Hitachi projection TV a while back. Took me an entire Sunday afternoon to dismantle it into small enough pieces for me to get it up the stairs and out to the curb. I was at work when it was delivered and had no idea how the guys got it down here. That was the cheapest decent 60" TV available at that time...$4000 and 130 lb, I think.)

To clarify, my question was intended to be more general--is there any rule of thumb based on others' experiences that suggests crossing over at a huge node a bad idea, good idea, or essentially immaterial if using room correction?
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,834
Likes
16,496
Location
Monument, CO
To clarify, my question was intended to be more general--is there any rule of thumb based on others' experiences that suggests crossing over at a huge node a bad idea, good idea, or essentially immaterial if using room correction?
At the crossover frequency phase is changing rapidly so your room correction scheme must be able to handle that (it has to work harder). Using just delay to compensate probably will not work. To compensate a room mode, I would set the crossover at least an octave (factor of two) above or below the modal frequency and see what I can do with the delay adjustment on my sub (or in the processor). That again requires the sub not be in the same place as the speakers since the idea is to introduce a signal to compensate for the mode (null; peaks are smaller and EQ can handle them). At 80 Hz the wavelength is about 14 feet so having the sub near, i.e. within a foot or two, of the main speakers means they will essentially excite room modes the same way (does not matter if it's the subs or the mains, the end result is the same). If there are boundaries that will change things a bit, but placing a sub in a corner means it excites all the modes in the room. If it is the corner near the main speakers, then where you set the crossover might not matter much.

Todd Welti's paper discusses multiple subs but you can see how room modes get activated: https://www.harman.com/documents/multsubs_0.pdf

HTH - Don
 
OP
sejarzo

sejarzo

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
961
Likes
1,066
Since sub and main speakers are essentially co-located and you can't move the listening position it is unlikely EQ will fix the problem of a deep null. It comes from physical dimensions. The usual solutions are to either move the listening position or add a sub (or more) placed to compensate ("drive") the null. Looking at the picture, could you move the sub to the left end of the couch to see if that helps?

Yeah, I am an engineer and totally understand the physics. It's not a null that is the big problem, it's a mode at the listening position that causes an 80 Hz tone to boom like crazy, which I have to assume is due to it being a mode for both the length and height dimensions.

Everything has rotated 90 degrees to the right from the picture. The TV and center channel speaker are gone. The sub is now sitting in front of that bass trap, the left main is in front of the 4 foot absorber to the right of the bass trap, the right main is where the silk tree is in that picture. The couch is on the other side of the post. The only possible sub locations are in that corner, or where it is shown in the picture. There's really no option to change the listening position.

As I noted above, this is more a general question about how best to set a crossover frequency with respect to known modes.
 
OP
sejarzo

sejarzo

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
961
Likes
1,066
And sorry, I just realized that I have been mistyping "mode" as "node" in previously replies. DOH.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,834
Likes
16,496
Location
Monument, CO
Yeah, I am an engineer and totally understand the physics. It's not a null that is the big problem, it's a node at the listening position that causes an 80 Hz tone to boom like crazy, which I have to assume is due to it being a node for both the length and height dimensions.

Everything has rotated 90 degrees to the right from the picture. The TV and center channel speaker are gone. The sub is now sitting in front of that bass trap, the left main is in front of the 4 foot absorber to the right of the bass trap, the right main is where the silk tree is in that picture. The couch is on the other side of the post. The only possible sub locations are in that corner, or where it is shown in the picture. There's really no option to change the listening position.

As I noted above, this is more a general question about how best to set a crossover frequency with respect to known nodes.
Sorry, people mix "node" and antinode" frequently so I usually just go with "mode" and explicitly say peak or null. Plus there is a fundamental null at the center of a closed rectangular room.

A peak can generally only be 3~6 dB louder than baseline so using EQ, either manually or from the room correction program, you should be able to reduce the volume at the listening position. Where the crossover sits is somewhat immaterial except the DSP will have to handle both the sub and mains. You can certainly try adjusting the crossover to see if it helps, but I would expect the room correction program to handle it for you.
 
OP
sejarzo

sejarzo

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
961
Likes
1,066
To compensate a room mode, I would set the crossover at least an octave (factor of two) above or below the modal frequency and see what I can do with the delay adjustment on my sub (or in the processor). ...

That would put the crossover at 40 Hz, which seems unreasonably low, or 160 Hz, which is too high and would result in localization, right? That's at least the theoretical quandary, eh?
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,834
Likes
16,496
Location
Monument, CO
That would put the crossover at 40 Hz, which seems unreasonably low, or 160 Hz, which is too high and would result in localization, right? That's at least the theoretical quandary, eh?
It is a (my) rule of thumb, along with my desire to cross over an octave above the mains -3 dB point if possible. Since the sub is near the mains, you could go a little higher and probably not notice. How high, I couldn't say off-hand, but you could try 100 or 120 Hz and see if it helps.

But see my previous post; room correction should be able to reduce both mains and sub appropriately.
 
OP
sejarzo

sejarzo

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
961
Likes
1,066
Thanks for all the info, Don!

Had I known all I have learned since before we finished the room, I would have had that post removed and replaced with two spread apart so as not to constrain the listening position and never configured it as in the photo. LOL it's more than a little late for that now.

I ended up doing 99+% of my listening on headphones after moving the sub around to 3 different locations when the room was in the original configuration and never being happy with all the problems. I tried a TacT system back in the day, along with two different AVRs with different versions of Audyssey, none of which ever provided much satisfaction.
 

Willem

Major Contributor
Joined
Jan 8, 2019
Messages
3,654
Likes
5,275
Remember a second sub does not have to be identical, or even large.
 
OP
sejarzo

sejarzo

Addicted to Fun and Learning
Forum Donor
Joined
Feb 27, 2018
Messages
961
Likes
1,066
Remember a second sub does not have to be identical, or even large.

Agreed. I have kept my eye on the local craigslist for something that I can get cheap for experimentation and then sell for what I paid for it.
 
Top Bottom