Hi everyone,
Following my first (somewhat messy) post full of questions, I’ve now refined my project and would love to get feedback from people with experience in this area.
To recap, my goal is to build a subwoofer to complement my home studio setup, which already includes the excellent Neumann KH120 A monitors.
Here’s the design brief, keeping in mind a limited budget:
Based on my WinISD simulations:
With 100 W (150WRMS recommended for the driver), Xmax is exceeded below 30 Hz. Adding a 25 Hz 4th-order LR high-pass filter helps protect the driver, but significantly increases group delay (>35 ms). So I opted for a 60 W amplifier to limit excursion while keeping group delay in check.
Based on my calculations, I need around 8 W (given the 85 dB @ 2.83V sensitivity) to hit 94 dB SPL, and since the driver is rated at 150 W RMS, 60 W seems like a safe and reasonable choice. I chose this amplifier, which also has a built-in DSP (https://www.soundimports.eu/en/sure-electronics-aa-ja31211.html)
What do you think about this setup?
Am I on the right track? Are there any mistakes in my reasoning? Do you have any suggestions or improvements?
Thanks in advance for your answers !
Following my first (somewhat messy) post full of questions, I’ve now refined my project and would love to get feedback from people with experience in this area.
To recap, my goal is to build a subwoofer to complement my home studio setup, which already includes the excellent Neumann KH120 A monitors.
Here’s the design brief, keeping in mind a limited budget:
- Crossover at 80 Hz between monitors and sub
- As flat a frequency response as possible from 30 Hz to 80 Hz
- Good transient response and low group delay
- Target level around 85 dB SPL, with 94 dB SPL max for fun
- Compact enclosure
Based on my WinISD simulations:
- Theoretical Qtc of 0.710 in 24.3 L, with an Fc of 39 Hz
- From what I understand, a Qtc near 0.707 means Fc ≈ F3, so I expect a drop of about -3 to -6 dB between 30–40 Hz, which seems acceptable, of course it's not perfect but for limited budget it's ok I think
- Group delay stays below 25 ms between 20–100 Hz, which I read is within acceptable limits for low frequencies
With 100 W (150WRMS recommended for the driver), Xmax is exceeded below 30 Hz. Adding a 25 Hz 4th-order LR high-pass filter helps protect the driver, but significantly increases group delay (>35 ms). So I opted for a 60 W amplifier to limit excursion while keeping group delay in check.
Based on my calculations, I need around 8 W (given the 85 dB @ 2.83V sensitivity) to hit 94 dB SPL, and since the driver is rated at 150 W RMS, 60 W seems like a safe and reasonable choice. I chose this amplifier, which also has a built-in DSP (https://www.soundimports.eu/en/sure-electronics-aa-ja31211.html)
What do you think about this setup?
Am I on the right track? Are there any mistakes in my reasoning? Do you have any suggestions or improvements?
Thanks in advance for your answers !
Last edited:
Group delay will stay where you set it. You mentioned a DSP, so you can filter the response, meaning amplitude & phase, to anything you like. Which in turn determines group delay, which is just another expression of phase.
Ok, what type of filter would you recommend to better control the phase shift?
Would a linear-phase filter be a better option?
As I mentioned earlier, I tested a 4th-order LR high-pass filter at 25 Hz, and I read that whether it's implemented digitally or analogically, the impact is essentially the same.
It does significantly increase group delay — which, to be honest, I don’t fully understand from a physical point of view (I’ll need to dig a bit deeper into that).
I wasn’t originally planning to use a high-pass filter — is it really necessary if the driver doesn't reproduce those lower frequencies anyway?
Last edited:
To have an idea of what your studio mixes sound like at low frequencies, low frequency response is required.
There is plenty of signal, both useful and not, well below 30Hz.
I'd consider 20Hz response a minimum requirement for studio monitoring this century.
https://www.scan.co.uk/PDFs/Products/kh120measurements.pdf
A pair of your KH120A should do 106dB at one meter with 3%THD at 80Hz (~81dB at 30Hz).
Considering how much less sensitive our hearing is at low frequencies, it makes sense to use a subwoofer with at least as much clean output potential as your studio monitors.
At Xmax, a driver typically has 10% THD.
If you spec your subwoofer for near Xmax, it will have at least that much distortion, which makes subjective mix decisions more difficult- distortion adds harmonics, increasing the perception of bass.
A single 8" with only 10.8mm Xmax would be ~10% distortion at 30Hz at 95dB, more than 10dB less output than your KH120A at 3%.
At 30Hz, a 5dB difference in level sounds twice or half as loud, at -10 dB 1/4 as loud as your monitors.
That said, you should measure how much boundary gain your room provides, there is ~18dB potential variance.
I'd suggest planning for a dynamic balance between the subwoofers and top cabinets, ~106dB in room at 20Hz, with the subs not exceeding Xmax.
Other than reducing the potential for the driver(s) suffering mechanical or thermal damage, a HP filter is not necessary.
If you provide enough power, you may hear signs of distress before the coil melts or the suspension breaks.
With no HP filter, the subwoofer cone(s) can give you a visual indication of LF content you may not hear 😉 .
Art
Attachments
Got a list of recordings with significant & useful MUSICAL content below 30Hz?
I used to have such a list and it was VERY short. Much more stuff if you include film stuff like dinosaur footsteps and/or Arnie S blowing up the Universe.
"Significant & useful" is very subjective.
Point is, unless you just plan to high pass everything you mix, you won't be able to decide what low frequency content is significant or useful without it having proper dynamic representation to judge it by.
Billie Eilish has drone tones from 30Hz down to under 5 Hz on "You should see me in a Crown".
The "Raising Sand" studio album by Alison Krauss and Robert Plant comes to mind as an older example of having a fair amount of musical content I enjoy below 30Hz.
Art
Point is, unless you just plan to high pass everything you mix, you won't be able to decide what low frequency content is significant or useful without it having proper dynamic representation to judge it by.
Billie Eilish has drone tones from 30Hz down to under 5 Hz on "You should see me in a Crown".
Mayhem13,
If your 2x6.5" near field sub is about 1/2 meter from your head, should be able to do around 93dB at 20Hz there.
Although "most engineers" may HP an entire recording as high as 25Hz, there are also many who do not.
After hearing quite a bit of LF on Billie Eilish's Grammy performance last week, checked into some of her tunes.
https://www.audiosciencereview.com/...-go-by-billie-eilish-insane-bass.39762/page-3
The flat VLF (very low frequency) "drone tone" extending below 40Hz on "You should see me in...
If your 2x6.5" near field sub is about 1/2 meter from your head, should be able to do around 93dB at 20Hz there.
Although "most engineers" may HP an entire recording as high as 25Hz, there are also many who do not.
After hearing quite a bit of LF on Billie Eilish's Grammy performance last week, checked into some of her tunes.
https://www.audiosciencereview.com/...-go-by-billie-eilish-insane-bass.39762/page-3
The flat VLF (very low frequency) "drone tone" extending below 40Hz on "You should see me in...
The "Raising Sand" studio album by Alison Krauss and Robert Plant comes to mind as an older example of having a fair amount of musical content I enjoy below 30Hz.
Art
For now, I’ve been working only with my pair of neumann monitors. While my mixes aren’t amazing, they’re good enough to send to a mixing engineer afterwards. I would’ve liked to reach down to 20 Hz, but I imagine that would require a larger driver and more power and my budget is limited.
I hear your point, but I’ll probably stick with 30 Hz for now (maybe 20 Hz in a future project).
Regarding output level, I know my monitor speakers can reach up to 106 dB SPL, but I’ve set them to 94 dB SPL using the switches in the back of the monitors, since I usually sit about 1 meter away. 99% of the time, I aim for 85 dB SPL, so I don’t really need my subwoofer to go beyond 94 dB SPL either. Also, I almost always listen to music while seated at my desk even for casual listening which is part of why I designed it this way.
Update: I'm using a 2nd-order Butterworth high-pass filter at 25 Hz to limit group delay under 25 ms, and it significantly reduces cone excursion (see image below). I'm powering the sub with a 60 W amplifier even with 100 W input, the excursion curve remains a parabola centered around 30 Hz, with Xmax capped at 9 mm. So I don't expect harmonic distortion to be a problem.
In any case, thanks a lot for your input — it really helps me. I’m excited to move on to the practical side of things for this first project!
That's a very steep high-pass filter, which is −6dB at 25Hz, and −3dB at 31.5Hz. The combined roll-off rate is 6th-order. You'd get most of the desired displacement-reducing effect using a 2nd-order high-pass filter, or even a 1st-order filter set to 20Hz or less.
Note that, as you reduce the low-frequency cut-off point of a closed-box enclosure, the group delay increases. However, the change that occurs going from a 40Hz cut-off to a 30Hz cut-off is not all that large for frequencies above 30Hz. In fact, the group delay is reduced above 36Hz.
Here is a VituixCAD simulation of the GRS SW8-4HE 4-ohm driver in a 24-litre closed-box enclosure. A 4th-order Linwkwitz–Riley low-pass filter has been included, set to 69Hz to achieve a filtered acoustic response that is −6dB at 80Hz.
Maybe it would be worth considering a 10-inch driver? Apart from reducing the displacement requirements, the larger drivers are often a little bit more sensitive than a typical 8-inch "subwoofer" driver.
Switching to the GRS 10SW-4HE in the same 24-litre enclosure produces the following response. Notice how both the input power and driver excursion requirements are both significantly reduced. This will provide some worthwhile headroom for transient peaks at low frequencies.
If the enclosure volume is increased to 36-litres, then we can lower the −3dB cut-off point to 30.8Hz from the previous 34.1Hz. At 36-litres, this enclosure is still very compact for a subwoofer.
And if we add a 3300µF series capacitor to the 36-litre subwoofer, we get the following response. Here the −3dB cut-off point has been lowered to 27.1Hz, and the driver excursion is reduced below 20Hz or so.
What about Pink Floyd's "Dark Side Of The Moon"? And more recently there is the song "Put on" by Little Jeezy. All of these have musical content below 30Hz. Unfortunately, the requirement for low-frequency extension in subwoofers, and the music systems that they are attached to, for music applications seems to be a contentious topic.
Last edited:
Thanks for bringing our attention to that album, which won a few Grammy Awards back when it was released.
Below is a frequency analysis of the track "Fortune Teller" from that particular album. There is significant low-frequency energy at 23.6Hz, judging by the peak at that frequency.
There sometimes IS useful musical content at 20-35Hz - but always at significant lower level as the main bass drum (around 60Hz). So a closed speaker which is EQed to be flat to low frequencies has NORMALLY no problem with music cause the content it needs to play down there is 20dB lower (-40dB at the sample from Billie - looks more like noise to be honest).
It changes with movies ... there you have full power down to 20Hz and need 2-3x the drivers.
So no - for music you don't need a high pass for a closed subwoofer! Keep it's natural great behaviour.
For live/PA that's something complete different ... just last weekend I got a "great" mic drop again ...
p.s.: wow, that Fortune Teller track is a huge exception. You will find a few of these but there are very little tracks out there who behave like this.
It changes with movies ... there you have full power down to 20Hz and need 2-3x the drivers.
So no - for music you don't need a high pass for a closed subwoofer! Keep it's natural great behaviour.
For live/PA that's something complete different ... just last weekend I got a "great" mic drop again ...
p.s.: wow, that Fortune Teller track is a huge exception. You will find a few of these but there are very little tracks out there who behave like this.
Reinforcing one of weltersys' points, annoying and unintended low frequency content winds up in recordings quite a bit. This often seems like someone along the chain didn't have a system capable of playing down to 20 Hz (or wherever the noise happens to be). With more people doing their own editing for YouTube, etc., unintended sounds in recordings annoy me more these days than I recall in the past - wind noise, thumps, bumps, low-frequency traffic noise, etc.
On the other end of the scale, there's a guy with a good health-related channel whose mic is too close to his mouth, and I can hear him snap his teeth together periodically - drives me bonkers.
That album also has some strangeness in the very low frequencies that I don't think was intentional. Headphones, speakers . . . something needs to have extended low frequency performance if you're mixing/editing and want to ensure that you don't have issues down there.
This spot in Breathe is one of them
My mixes seemed fine using near field monitors with a -3dB point ~57 Hz.
I'd use headphones to check for any wild stuff down lower.
2006 my musical partner got a new keyboard and started laying down bass and drum tracks.
The two octaves below the near field monitors then became a problem, VLF is hard to judge with phones.
Being in a hurry to finish the backing tracks, I auditioned a number of "studio subwoofers", then purchased a Mackie HRS 120 (one 12" woofer and one 12" passive radiator), free field response -2dB at 20Hz. The extended clean VLF of the HRS 120 was instantly obvious when listening in the 10,000 square foot (900 square meter) when compared to the other subs that rolled off in the 30-35Hz range.
Using that sub (and a footswitch mute) with the studio monitors I could hear what I was missing, and confidently know how the mix would sound with or without the VLF.
Unless you are sending the individual tracks to be re-mixed, VLF "mistakes" can't be corrected properly.
The attenuator switches on the back of your monitors set the output level achieved from one monitor with an input of 0dBu, they don't limit the output to that level.
A pair receiving the same signal will put out approximately +6dB more SPL.
Depending on your meter ballistics and the duration of the peak signals, "0dBu" could be as much as 20dB lower than peak/impulse SPL.
Go for as much displacement (Sd x excursion) as your budget allows now, you can add a second sub (and amp, if required) for more headroom later.
For studio monitoring, stereo subs are also useful for making L/R pan decisions.
The sonic advantages due to expanded placement options for a pair of subwoofers far outweigh the reduction in cabinet vibrations that horizontally opposed drivers of an equal enclosed volume offer.
Art
Last edited:
I don't think in 1973 when Pink Floyd's "Dark side of the moon" was originally recorded on tape, and mastered for vinyl that there was any intention of it being remastered by different engineers for digital and vinyl release 50 years later in 2023.
Much less being streamed with YouTube's codec in 2025:
Anyway, none of it sounds quite like it did in 1973 when my hearing was still in pristine condition, and most of the speakers I heard it on were junk by comparison to the OPs 😉 .
Art
Thank you for sharing your experience with using the Mackie HRS120 subwoofer. I looked it up and it does seem to be a very capable performer. Its published frequency response curve is shown below (−3dB at 19Hz):
Last edited: