• 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!

Manually time-aligning subwoofer(s) to mains - how to

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
While I am trying to align the sub to my fronts, I followed the steps from the first post: Measured my speakers+sub, one by one, with XO turned off, and did a 30-150Hz sweep. Now I am trying to figure out if it's all correct, by looking at the graphs, and to me it seems that all speakers are aligned properly, but please do check!

But by looking at the post right above from @thorvat , I tried using the "Align phase at cursor" in the Alignment tool. And when comparing Sub to FL vs Sub to FR, or comparing FL to FR, the alignment tool sees a 5-6 meter diff? That my FL is 5-6 meter closer to MLP that FR? Using a tapemeasure, I know they are exactly 215cm from the MLP.

Have I done something wrong, or are there something wrong with my understanding?
MDAT: XO Off - 30-150 - 20211104.mdat

Took a quick look...

I rather prefer to see the full-range sweep with no crossovers and EQ.

The sub is likely closer to one side than the other -- so probably not located at midpoint.

Looks like a ~4ms negative offset (delay to the mains) would align the LCR to the sub better -- if you don't want to apply xo filters.

The phase slope of the surrounds are different, though, from the rest.
 

thorvat

Senior Member
Joined
Aug 9, 2021
Messages
323
Likes
386
While I am trying to align the sub to my fronts, I followed the steps from the first post: Measured my speakers+sub, one by one, with XO turned off, and did a 30-150Hz sweep. Now I am trying to figure out if it's all correct, by looking at the graphs, and to me it seems that all speakers are aligned properly, but please do check!

But by looking at the post right above from @thorvat , I tried using the "Align phase at cursor" in the Alignment tool. And when comparing Sub to FL vs Sub to FR, or comparing FL to FR, the alignment tool sees a 5-6 meter diff? That my FL is 5-6 meter closer to MLP that FR? Using a tapemeasure, I know they are exactly 215cm from the MLP.

Have I done something wrong, or are there something wrong with my understanding?
MDAT: XO Off - 30-150 - 20211104.mdat

This indeed is looking strange. Phase seems to be not far from being aligned with L and R in the 70-80Hz region yet impulse and step response are off. How exactly did you measure? Which speaker did you use for playing a timing reference signal?
 

thorvat

Senior Member
Joined
Aug 9, 2021
Messages
323
Likes
386
I rather prefer to see the full-range sweep with no crossovers and EQ.

Nope - if you want to compare IR and step of mains vs sub you need a ranged sweep. In most cases 2Hz-200Hz sweep with 2M size would do the job. (yep, not from 20Hz, but from 2Hz).
 

hmt

Senior Member
Joined
Apr 11, 2020
Messages
397
Likes
479
First you shuld measure with XO on. Your sub is also too loud by about 10 db. I would adjust it to the same level. Also A measurement of L and R simultaneously would be helpful because this is the scenario you want to optimize for.
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
Nope - if you want to compare IR and step of mains vs sub you need a ranged sweep. In most cases 2Hz-200Hz sweep with 2M size would do the job. (yep, not from 20Hz, but from 2Hz).

I prefer to do more than that…

I want to be able to simulate the application of different HPF and LPF outcomes by convolving the full IR sweep. But that is me.
 

thorvat

Senior Member
Joined
Aug 9, 2021
Messages
323
Likes
386
First you shuld measure with XO on. Your sub is also too loud by about 10 db. I would adjust it to the same level. Also A measurement of L and R simultaneously would be helpful because this is the scenario you want to optimize for.

While everything you said is true, and has to be done, level adjustment won't help with phase adjustment.
 

thorvat

Senior Member
Joined
Aug 9, 2021
Messages
323
Likes
386
I want to be able to simulate the application of different HPF and LPF outcomes by convolving the full IR sweep. But that is me.

That should be done after sub(s) integration. For sub integration ranged sweep is the way to go.
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
That should be done after sub(s) integration. For sub integration ranged sweep is the way to go.

I don’t see why not either way. I personally prefer a full-range sweep and, no, you aren’t really losing anything if you choose not to apply a range limit — in fact, we are gaining more information.
 

thorvat

Senior Member
Joined
Aug 9, 2021
Messages
323
Likes
386
I don’t see why not either way. I personally prefer a full-range sweep and, no, you aren’t really losing anything if you choose not to apply a range limit — in fact, we are gaining more information.

You can't compare IRs and step responses of sub and mains when doing full range sweep as main's much wider response would alter their IR and Step too much to be compared with limited range of the subs.

You can however still match phases without even looking at IR and step response as I described in my previous post.

Regarding your method of using full range sweeps "to simulate the application of different HPF and LPF outcomes" - if it is about detereming the optimal XO point between subs and main I would rather recommend measuring and comparing distortion via RTA of sub and mains in the 40-100Hz region as that will serve as a better indicator of the LF limit where mains start to struggle.
 

hmt

Senior Member
Joined
Apr 11, 2020
Messages
397
Likes
479
While everything you said is true, and has to be done, level adjustment won't help with phase adjustment.
I know but I meant it just to make it easier to interpret the graph and to judge how good the summation bn Subs and Mains works.

edit:

Application of HP and LP can be done via the EQ module in REW but I do not find any way to copy that prediction. Did a quick approach by first summing L and R and saving the aligned sum of it. Then I summed the aligned LR sum with the sub and used 80Hz as point for aligning the phase (phase slope was a bit worse) slope. Got 5 ft delay for the mains.
 
Last edited:

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
You can't compare IRs and step responses of sub and mains when doing full range sweep as main's much wider response would alter their IR and Step too much to be compared with limited range of the subs.

There is a function within REW to apply a 1/1 or 1/3 filter at the center frequency of your choice -- if you really want to match the shape and timing of the IR and/or step via said limited range. Again, you aren't losing any information with a full-range sweep at all.

Application of HP and LP can be done via the EQ module in REW but I do not find any way to copy that prediction. Did a quick approach by first summing L and R and saving the aligned sum of it. Then I summed the aligned LR sum with the sub and used 80Hz as point for aligning the phase (phase slope was a bit worse) slope. Got 5 ft delay for the mains.

I use rePhase to generate all the filters and import them into REW for convolution. Then we use the alignment tool... But I know everybody has their own preferred method.
 

fluid

Addicted to Fun and Learning
Joined
Apr 19, 2021
Messages
691
Likes
1,196
Application of HP and LP can be done via the EQ module in REW but I do not find any way to copy that prediction. Did a quick approach by first summing L and R and saving the aligned sum of it. Then I summed the aligned LR sum with the sub and used 80Hz as point for aligning the phase (phase slope was a bit worse) slope. Got 5 ft delay for the mains.
In REW you can convolve responses together by using the A*B option in the All SPL tab, because it is a multiplication there will need to be an adjustment to the level afterwards to return it to a sensible value. Usually around -83dB. You can save a filters impulse response and convolve it with your original measurement if this is what you were trying to do.
 

SitronNO

New Member
Joined
Nov 6, 2021
Messages
4
Likes
0
This indeed is looking strange. Phase seems to be not far from being aligned with L and R in the 70-80Hz region yet impulse and step response are off. How exactly did you measure? Which speaker did you use for playing a timing reference signal?
1. Placed the mic (UMIK-1) at MLP, pointing upwards.
2. Turned the volume on my AV-receiver (Anthem MRX 740) up to -25dB, so SPL showed about 78dB
3. In my AV-receiver, I turned off XO for all channels (speakers). I did not, however, turn off my ARC (Anthem Room Correction)
4. I checked that the distance was correct according to my measurements with a laser measurement tool
5. I ran REW, took a measure with:
- Type: SPL
- Range: 30-150Hz
- Length: 1M, repetition: 1
- Timing: Use acoustic timing reference
- Output: Front Left
- Timing Ref: Front Left
6. Repeted step 5, but with Output: Front Right (then Center, Sub, Surr Left, Surr Right)
7. Saved the .mdat file that was included.
 

SitronNO

New Member
Joined
Nov 6, 2021
Messages
4
Likes
0
There is a function within REW to apply a 1/1 or 1/3 filter at the center frequency of your choice -- if you really want to match the shape and timing of the IR and/or step via said limited range. Again, you aren't losing any information with a full-range sweep at all.
Can you please explain to a newbie how and where to do this step, and what I am supposed to look for? (Or give a link to someone who does :))
 

SitronNO

New Member
Joined
Nov 6, 2021
Messages
4
Likes
0
Your sub is also too loud by about 10 db
I measured the sub alone (in other words the LFE-channel), and that will always get a 10dB boost. Ref: https://hometheaterhifi.com/editorial/the-misunderstood-01-lfe-channel-in-51-digital-surround-sound/

For this reason, home Dolby Digital equipment is pre-set to play LFE data 10 dB higher than a main channel (or 10dB higher than the bass from a main channel). It is only necessary to set the subwoofer relative to a main channel and the LFE level will be correct.

When I measure the sub with the rest of the speakers (and XO turned on), the sub is at the same level as the speakers for all channels, expect the LFE-channel.
 

hmt

Senior Member
Joined
Apr 11, 2020
Messages
397
Likes
479
I would measure L+R combined but with the L+R speakers switched off or disconnected. This way you would also have the XO filter measured. Measuring the sub channel alone would not only result in a +10 db boost but also in an application of an LPF at 120Hz unless disabled in the AVR.
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
Can you please explain to a newbie how and where to do this step, and what I am supposed to look for? (Or give a link to someone who does :))

I do not exactly like this method since the phase response in a real room isn't always flat esp. around the xo -- it could be very, very mangled, indeed -- depending on which center frequency you choose to filter, the optimal delay will be biased towards it. That's also why I'd rather compare the complete IR.

Anyway, you go to the "Filtered IR" tab view in REW. Click on the graph controls (sprocket wheel button). Select and enable a filter for each sub and monitor. Open the overlays window and go to either the impulse or step tab view. Select the traces you want to "align".

1636304609551.gif


To apply a delay or timing offset: go back to the main window's SPL & Phase tab view, and in the graph controls select "offset t=0" to adjust based on the overlay view.

You can do as @hmt suggested, which is to use the L+R sum or vector avg and apply the xo filters you like prior doing this process.
 

ernestcarl

Major Contributor
Joined
Sep 4, 2019
Messages
3,106
Likes
2,313
Location
Canada
Just to note, the timing delay offset suggested by REW (in your measurements without any xo filters) is not actually all that incorrect:


1636310479966.png



1636310284955.gif

*FDW applied to improve graph visualization


At best, you could say that it's sort of a compromise.


To apply frequency dependent window:
1636310373680.png
 

hmt

Senior Member
Joined
Apr 11, 2020
Messages
397
Likes
479
The problem with the non flat and often different phase response is one reason I use the alignment tool in REW. Concerning the picture: isn't REW aiming for the first impulse that seems to be inverted?
 
Top Bottom