Optimizing Subwoofer positioning

[ErLan]

Hello ASR members,

My Journey continue... this time optimizing subwoofer positioning...

In order to identify the best position for my subwoofer I took systemic measurement of the sub response at the listening position. The measurement were in steps of 30cm and covered the range of the possible locations along the front wall.
Looking at the results I tried to identify the point where BOTH Frequency Response and Time response (as reflects in spectrogram) are close to ideal - which will allow me later on a flat integration with my mains.

Apparently (as expected?) my non-treated room demonstrated an inherent conflict, where, for places where frequency response was not-bad (hence, close to flat line till ~30-35Hz) the time response was messy (peaks are sparse all over the place with no consistent delays) and vice versa <see attached>

I will appreciate ANY perspective on the preferred position, and how do you see next steps to relief the identified pain point.

p.s. #1: The measurements were taken through LFE channel with and a high LPF (500z); practically I am intending to cross the sub at ~100-130Hz.
p.s. #2: My sub is Wharfedale WH-D10 (sealed, 10")

 

[Keith_W]

Option 1: Do the subwoofer crawl

Option 2: Put your subs where you like (within reason) and use DSP.

BTW there was no attachment with your post.

 

[sweetchaos]

Are you measuring frequency response with ReW and a measurement microphone (like MiniDSP Umik-1)?

 

[ErLan]

Sorry, apparently one file was too big. Files are now attached.
@sweetchaos, yes. measuring with REW and UMmik-1; DSP is done via. EQAPO
@Keith_W, subcrawl seems to me like a frequency response biased method, especially when modes emphasis and magnify, what is your experience with that?

 

[computer-audiophile]

Option 1: Do the subwoofer crawl

LOL

"Crawl around the listening room and listen for when the sub seems to sound the best. (IE. Listen for depth, loudness, tightness and definition)"

 

[FrantzM]

Hi

Welcome

I wouldn't worry about the time response, yet ( perhaps ever). A few questions:
Could you increase the level so that the reference would be 75 dB?
What sub is this? Just in case.
Would you be willing to use some EQ ?
What mains do you have?

Peace.

 

[FrantzM]

LOL

"Crawl around the listening room and listen for when the sub seems to sound the best. (IE. Listen for depth, loudness, tightness and definition)"

He is measuring it, and seem willing to do so..
He's already using the better approach.


Peace.

 

[fpitas]

When I was a boy, we had to do the subwoofer crawl. And we liked it!

 

[ErLan]

@FrantzM, Interesting

I have a powered Wharefdale WH-D10 sub (sealed box, 10"), and my mains are LS50.

I set the power on sub to 12:00, therefore still have some juice to take it to 75dB.

EQ is surely one of the tools I am using (EQAPO)

My main challange to cross sub and main, originates from that the is that room modes modes messes up the phase makes it impossible (?) To generate an overlap range without any dip due to phase cancellations.

You mentioned I should not be worrying on that, can you elaborate?

 

[MAB]

LOL

"Crawl around the listening room and listen for when the sub seems to sound the best. (IE. Listen for depth, loudness, tightness and definition)"

It's really incredibly undignified!
But actually quite effective.
Think of it as an in-room role-reversal of the listener and the speaker.
Hopefully it makes sense! It is one of the few good alternatives to using a mic, since it takes advantage of the large influence of the room on bass frequency response.
It is also very useful in placing your mains to get their bass response tamed in a room, if you don't have a microphone.

 

[fpitas]

But actually quite effective.

Yes, it is. If I was stuck without DSP, or wanted to simplify, that would be the approach I'd use.

 

[ErLan]

Well, though crawling it is a fun family experience on my end as well. The purpose of the delta 30cm measurements that I took is to strengthen amd optimize the decision based on qualitative data (ASR, after all ).

 

[FrantzM]

Hi

The sub is ... small but you'll have to live with it.
suppose you will leave the subs in the front plane, the volume where the front speakers are.
Put the microphone at the listening positon, the MLP for short. Then conduct a series of measurements of the sub at different location, trying to find the smoothesst FR from the sub. Do not worry about the issue of phase, yet... Just find the best, smoothest most extended ( between say 30 to 150 Hz) frequency response. Mains should be off. Sub only and yes, you can push it a bit past 12... I wouldn't feed anything below 20 Hz to the sub though, so limit the sweep range to 25~200 Hz...

Then

Post the measurements here

Peace.

 

[ErLan]

Thanks @FrantzM ,
Did you get a chance to look at the measurements I attached (see post #4 in this thread)?

Exlcuding measurement 1 at file "... 1of2",
I believe they meet most of the acquisition points mentioned (main gap is the recommanded range where the ones that I took are covering wider span 10-500Hz, while your recommendation 25-200Hz, but I believe resolution is still good).

 

[Keith_W]

Oops, I should have linked to a better article. That article made me look stupid! Never mind, damage is done, mea culpa. It is still conceptually the same - place the subwoofer on your sofa, and then go around the room measuring sweeps at different points. Choose a point where there are the fewest peaks and nulls, bearing in mind it is more important to avoid nulls than peaks.

If you are going to cross your sub high (you stated 100-130Hz), bear in mind that bass becomes directional above 80Hz, so you must position your sub in front of you. Even then, remember that harmonic distortion products are higher frequency and thus more directional, so even 80Hz might be audibly directional if your sub distorts. So: choosing a high crossover point limits where you can position your sub.

As for timing, it is rare that an optimal position for your sub in terms of frequency response will also be the optimal position for timing. Unless you are willing to use DSP to correct timing, there is nothing you can do about it.

 

[ErLan]

Thanks @Keith_W.
This approach resonates with what I did (instead of moving the mic, I moved the sub..).
The comment on the high crossing and its influence on position makes sense and indeed nerrows the possible region to 2-3 places (in the attached files, 260cm-290cm that is from both sides of LP thqt is ~275cm from both left and right walls).
Those points also seems good from frequency response, neverthless, time response is somewhat impaired (especially in the designated crossover region). Any idea what can be done there to improve (DSP is definitly an option).

 

[Keith_W]

Thanks @Keith_W.
This approach resonates with what I did (instead of moving the mic, I moved the sub..).
The comment on the high crossing and its influence on position makes sense and indeed nerrows the possible region to 2-3 places (in the attached files, 260cm-290cm that is from both sides of LP thqt is ~275cm from both left and right walls).
Those points also seems good from frequency response, neverthless, time response is somewhat impaired (especially in the designated crossover region). Any idea what can be done there to improve (DSP is definitly an option).

I will put it to you this way - for every 1m further away your sub is, you will add 2.9ms of delay to your speaker. This is the so-called "time of flight" delay.

However, subs do not only have time of flight delay. There may be additional delays from propagation, especially if the sub has built-in DSP. These can add anywhere from 5-30ms of delay. Even if the sub is sitting on your lap, you will not get rid of this delay.

I recommend that you position your sub so there are as few nulls as possible. They probably need to be in front of you if you want a high crossover point. Alternatively, choose a lower crossover point and you can have more flexibility where you place them. Peaks can be lopped off with DSP. Delays can be dealt with by measuring the delay, and delaying your main speakers by the corresponding amount.