MartinC

Subwoofer Integration with a miniDSP 2x4 HD

Recommended Posts

Posted (edited)

Just before Easter I took a punt on a BK Monolith subwoofer that was being sold cheaply locally to me. I'm very glad I did :)

Had I simply relied on the adjustments on the sub to integrate it with my main speakers I don't think I'd be using it for anything more than TV/movies though, but with the aid of my miniDSP 2x4 HD I've now got it working well for music too. The important functions of the miniDSP are that it provides:

  1. A proper crossover between the main speakers and sub (the mains are no longer run full-range)
  2. Time delay adjustment to get good phase matching over the crossover region
  3. Equalisation to help manage room modes (which I'm applying only to the sub signal, not the main speakers)

To be able to use these capabilities I'm running my miniDSP as both a DAC and digital pre-amp, with outputs being fed directly into my power amp for the main speakers, as well as a mono output going to the (powered) sub.

My main speakers are Meadowlark Shearwater HotRods which have a specified output down to 35 Hz (-3 dB) so they are not what I think most would consider bass-light. The upshot of adding the sub has though been a significant improvement, giving a much more solid, powerful bass response that is still punchy and tuneful rather than just being one, big boomy mess (as bad systems with subs easily can be in my experience). This gels with what's going on at higher frequencies to be, well, just better! Oh, and of course it's good for TV/movies too :D.

For those of a nervous disposition I should probably warn you that I'm now going to post some graphs. Quite a few in fact. So here's your chance to escape this thread without hopefully too much distress if you wish...

Still here? OK. Here are two graphs showing first the left channel response and then the right, with the grey curves being just the Shearwater's on their on and the coloured curves being what I've achieved with the Monolith added (and with EQ applied to the sub only).

982025962_LeftChannel.jpg.2a4573678a2bca880004ceceba7cea0f.jpg

1847811818_RightChannel.jpg.5a3c7074bd415153f3ff6b3f82470fb0.jpg

The crossover from mains to sub is at 120 Hz, using 48 dB/octave Linkwitz-Riley filters. The comparisons are slightly unfair to my main speakers in that I optimised their position with the crossover to the sub in mind (i.e. not worrying what happened below about 80 Hz) but I couldn't make things much better than shown. The crossover frequency is higher than I'd originally intended but was primarily chosen as this was a region of good phase agreement between the left and right mains*, allowing for good matching to the sub and therefore a reasonably well-behaved crossover. As you'd expect I was concerned that 120 Hz might be too high, such that the sub would be obviously localisable but I don't fine that it is. A significant factor in this will be the sharp crossover roll-off. As examples, both the sub's own crossover control (which I bypass) and that applied by my blu-ray player, roll-off much more slowly, so whilst I could nominally set them to a more typical 80 Hz crossover frequency they would both be outputting audible signal to much higher frequencies than with my current setup.

* For the steady-state response. First arrivals signals from the main speakers are of course in-phase across the whole bass region, but what we hear at these frequencies is dominated by the steady-state reached once reflected signals are added in. 

The EQ applied to the sub has no net boosts applied, and I took care that the filters themselves did not introduce significant time domain ringing (as judged by filter RT60s in REW being < 350 ms). I could of course apply EQ to my main speakers run on their own to make the comparisons above look rather less extreme but the results would not be as good as with the sub. This is because having an entirely separate gain control on the sub means I've been able to sacrifice a bit of its headroom to achieve a flatter response. My system still plays louder than I could comfortably listen to without sub output compression occurring though.

These graphs are at a position where the centre of my head would be. Yes there is variation in performance away from this location but the result is still far better than what I'd get with the main speakers alone (the response of which also varies with position of course). 

Finally, for the few people who might be interested, here are waterfall plot comparisons for the right channel:

149034589_RightMainwaterfall.jpg.71d78def14a8a5677c07f67fc0baf86a.jpg

249630967_Rightcombowaterfall.jpg.5239ff8f0bae9aeaad91940cbbb757b4.jpg

 

What I'm working on right now is experimenting with adding some acoustic panels to my room which both changes the required EQ and affects the decay times, but I thought I'd post what I'd achieved under 'normal' conditions first.

Edited by MartinC
  • Like 2
  • Upvote 1

Share this post


Link to post
Share on other sites
Super Dealer

Try a little less EQ on the subs, ‘most’ people prefer a gently down sloping target curve, 

Keith

Share this post


Link to post
Share on other sites
Posted (edited)
34 minutes ago, PuritéAudio said:

Try a little less EQ on the subs, ‘most’ people prefer a gently down sloping target curve, 

Keith

Thanks. Yes, I'm aware of that, and there is actually a slight downward slope over the sub region but it's not particularly obvious given the large range needed to show the initial measurements. I believe 1 dB per octave is a typical slope preferred and I've currently gone with around a 2 dB drop from 20 to 120 Hz. One option I might explore long term would be a slightly more elevated bass response for movies than for music (which I could easily switch between on the miniDSP).

I think most of the target curve preference stuff comes from a target curve across the whole frequency range rather than just the limited range I'm applying EQ to of course.

Edited by MartinC

Share this post


Link to post
Share on other sites

I'll just add that I created the downward slope to the bass response manually whilst optimising the EQ. REW does have an option to run its optimisation to a desired slope (or any target curve you want) but I've not found it particularly successful.

  • Upvote 1

Share this post


Link to post
Share on other sites
Moderator
 
 
1
10 minutes ago, MartinC said:

I'll just add that I created the downward slope to the bass response manually whilst optimising the EQ. REW does have an option to run its optimisation to a desired slope (or any target curve you want) but I've not found it particularly successful.

I would agree. The word "clunky" springs to mind. In comparison, the Dirac tools are much easier to use to both create a target curve and apply it to the target. 

Share this post


Link to post
Share on other sites
4 minutes ago, Tony_J said:

I would agree. The word "clunky" springs to mind. In comparison, the Dirac tools are much easier to use to both create a target curve and apply it to the target. 

I'm sure you're right. What I was particularly referring to though was that when the optimisation is run to a slope, the results are generally a poorer match to what was being aimed for than when trying to EQ flat. Or at least this has been my experience.

  • Upvote 1

Share this post


Link to post
Share on other sites
On 22/05/2019 at 08:10, MartinC said:

Just before Easter I took a punt on a BK Monolith subwoofer that was being sold cheaply locally to me. I'm very glad I did :)

Had I simply relied on the adjustments on the sub to integrate it with my main speakers I don't think I'd be using it for anything more than TV/movies though, but with the aid of my miniDSP 2x4 HD I've now got it working well for music too. The important functions of the miniDSP are that it provides:

  1. A proper crossover between the main speakers and sub (the mains are no longer run full-range)
  2. Time delay adjustment to get good phase matching over the crossover region
  3. Equalisation to help manage room modes (which I'm applying only to the sub signal, not the main speakers)

To be able to use these capabilities I'm running my miniDSP as both a DAC and digital pre-amp, with outputs being fed directly into my power amp for the main speakers, as well as a mono output going to the (powered) sub.

My main speakers are Meadowlark Shearwater HotRods which have a specified output down to 35 Hz (-3 dB) so they are not what I think most would consider bass-light. The upshot of adding the sub has though been a significant improvement, giving a much more solid, powerful bass response that is still punchy and tuneful rather than just being one, big boomy mess (as bad systems with subs easily can be in my experience). This gels with what's going on at higher frequencies to be, well, just better! Oh, and of course it's good for TV/movies too :D.

For those of a nervous disposition I should probably warn you that I'm now going to post some graphs. Quite a few in fact. So here's your chance to escape this thread without hopefully too much distress if you wish...

Still here? OK. Here are two graphs showing first the left channel response and then the right, with the grey curves being just the Shearwater's on their on and the coloured curves being what I've achieved with the Monolith added (and with EQ applied to the sub only).

982025962_LeftChannel.jpg.2a4573678a2bca880004ceceba7cea0f.jpg

1847811818_RightChannel.jpg.5a3c7074bd415153f3ff6b3f82470fb0.jpg

The crossover from mains to sub is at 120 Hz, using 48 dB/octave Linkwitz-Riley filters. The comparisons are slightly unfair to my main speakers in that I optimised their position with the crossover to the sub in mind (i.e. not worrying what happened below about 80 Hz) but I couldn't make things much better than shown. The crossover frequency is higher than I'd originally intended but was primarily chosen as this was a region of good phase agreement between the left and right mains*, allowing for good matching to the sub and therefore a reasonably well-behaved crossover. As you'd expect I was concerned that 120 Hz might be too high, such that the sub would be obviously localisable but I don't fine that it is. A significant factor in this will be the sharp crossover roll-off. As examples, both the sub's own crossover control (which I bypass) and that applied by my blu-ray player, roll-off much more slowly, so whilst I could nominally set them to a more typical 80 Hz crossover frequency they would both be outputting audible signal to much higher frequencies than with my current setup.

Interesting post! I am looking forward to doing a similar setup with a miniDSP and a couple of subs in my system. I was interested to see from the graphs that the peaks in the main speakers bass response between 30-50 Hz have been tamed even though you were only eq'ing the subwoofer. That is just what I'm hoping will happen with my system which has a peak between 40-50 Hz that I'd like to tame in a similar manner. Everything I've read says the effect is improved with at least two subwoofers too.

Share this post


Link to post
Share on other sites
40 minutes ago, rdale said:

Interesting post! I am looking forward to doing a similar setup with a miniDSP and a couple of subs in my system. I was interested to see from the graphs that the peaks in the main speakers bass response between 30-50 Hz have been tamed even though you were only eq'ing the subwoofer. That is just what I'm hoping will happen with my system which has a peak between 40-50 Hz that I'd like to tame in a similar manner. Everything I've read says the effect is improved with at least two subwoofers too.

The miniDSP is crucially applying crossovers to both the mains (high-pass) and sub (low-pass) at 120 Hz. Since there is now essentially no output from the main speakers at 36 Hz  and 48 Hz they do not excite the room modes at these frequencies, and the sub has EQ applied to stop similar peaks being present in its output. If I didn't apply a crossover like this then I would still have significant low frequency peaks generated by the main speakers showing in the overall response.

Share this post


Link to post
Share on other sites
Posted (edited)

I'll just add that as well as preventing room mode issues, applying a crossover like I have means my power amp and main speaker mid/bass drivers are no longer handling the lowest bass content at all, which at least theoretically should offer other benefits. 

Edited by MartinC
  • Upvote 1

Share this post


Link to post
Share on other sites
Moderator

You can achieve a similar result without a crossover. My first experiments with DSP were with a (borrowed) DSPeaker Antimode Dual Core unit feeding a pair of Lowthers which had a BK sub connected across the high level outputs. We deliberately set the sub volume too high; the DSP then compensated for both the room modes and the overblown bass, which had the effect of rolling off the Lowther bass output in the region where the sub overlapped. Worked very well indeed.

Share this post


Link to post
Share on other sites
Posted (edited)
7 minutes ago, Tony_J said:

You can achieve a similar result without a crossover. My first experiments with DSP were with a (borrowed) DSPeaker Antimode Dual Core unit feeding a pair of Lowthers which had a BK sub connected across the high level outputs. We deliberately set the sub volume too high; the DSP then compensated for both the room modes and the overblown bass, which had the effect of rolling off the Lowther bass output in the region where the sub overlapped. Worked very well indeed.

Was that applying EQ to the main speakers as well as the sub?

Edited by MartinC

Share this post


Link to post
Share on other sites
Moderator
Posted (edited)
9 minutes ago, MartinC said:

Was that applying EQ to the main speakers as well as the sub?

Yes - the DSP fed the amp for the main speakers and the speaker output  also fed the sub amp's high level inputs. So when we ran the DSP correction sweeps the DSP was basing the correction on the combined output of the mains and the sub.

Edited by Tony_J
  • Like 1

Share this post


Link to post
Share on other sites
Just now, Tony_J said:

Yes - the DSP fed the amp for the main speakers and the speaker output fed the sub amp high level inputs. So when we ran the DSP correction sweeps the DSP was basing the correction on the combined output of the mains and the sub.

I thought so. That would obviously prevent the main speaker peaks being a problem and I could do the same, although I do think there are reasons that the crossover approach is better. 

Either way though we're talking about some way of reducing the output from the main speakers at the frequencies of room modes, whereas what I believe @rdale has in mind is applying no adjustment at all to the signal sent to the mains speakers, just the sub(s). The idea instead would be to rely on cancellation between the sub and main speaker signals to achieve a flatter response, which is possible but more complicated, and runs into issues where bass frequency content in music is not panned dead-centre (which is isn't universally).

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.