That is good to hear, thank you. Yes it is the room which is my main concern too, in my case it will vary and can't be predicted. This is mainly why I'm trying to make a compact version that can be used very near field, as well as in-room if/when circumstances are conducive. From what you say, I'll probably have to look at delay for both situations, it seems.
My (relatively) smaller design won't individually have huge output, but another reason for wanting compactness is that if/when it gets used for the whole room I happen to favour the multiple distributed approach, so there would be more than one which then need to be small enough to fit in many different places. Which is a restriction, and why I'm basically going for relatively small sealed designs with a big amp and EQ/DSP, but small is going to be most appropriate for my situation anyway - in almost any predictable use-case. If I had the opportunity, I would love the efficiency of bigger designs and am envious of some people's circumstances where they can have large and even infinate baffle subs, as I could too in the past.
But actually, for mainly music in quite small rooms, I think my opposed 12" design will be enough especially if there are a few. Certainly initially, in the near-field there'd be very little excursion, and/or I could EQ to a lower frequency at lower SPL setting for organ/electronic music or explosions and T-Rex moments in movies.
I would think carefully before building this exact system for nearfield use, where all the output will come from the slot and the exit will be located near your ear(s). In that case you will be MUCH more sensitive to the slot resonances. These are not 100% EQ-able, because any air movement can excite the 15dB-tall resonance even if there is no frequency component at the resonance frequency.
Because the exit will be close to your ears and your hearing is more sensitive to higher frequencies than lower ones in this band, you should think carefully about how it will sound.
When the system is far/farther away in the room, these noises are or will be much less of a problem.
Thank you, Charlie, that is a very important point. I haven't thought enough about that, i.e. that the EQ might only be a partial solution.
There are several things I can do to significantly reduce the peak acoustically, so whilst I'd been a bit carefree of those, I'll now pay more attention and alter the design a bit. It does seem like relatively minor changes can very significantly reduce the peak, so there might be cause for optimism. Especially as my intentions for the slot are just to (basically) bring two drivers together, rather than to do clever things wrt response etc.
Perhaps this is partly a compromise from wanting dual use out of these; a pure nearfeld design would likely choose a different path. I did look at the bandpass option that I think you suggested, but front chamber sizes conflicted a bit with requirements for compactness. So if simulations and prototypes for the slot don't work out then probably I'd go with simple sealed arrangement, and just be inventive about getting it close to the ears/head.
In fact.. that might be the better option anyway, and not necessarily mutually exclusive. Between what Juhazi has done and what GM has mentioned, I'd expect that a pair of sealed subs could fairly easily be configured into a dual-opposing design. It would just take serious attention to how well they're mechanically coupled.
There are several things I can do to significantly reduce the peak acoustically, so whilst I'd been a bit carefree of those, I'll now pay more attention and alter the design a bit. It does seem like relatively minor changes can very significantly reduce the peak, so there might be cause for optimism. Especially as my intentions for the slot are just to (basically) bring two drivers together, rather than to do clever things wrt response etc.
Perhaps this is partly a compromise from wanting dual use out of these; a pure nearfeld design would likely choose a different path. I did look at the bandpass option that I think you suggested, but front chamber sizes conflicted a bit with requirements for compactness. So if simulations and prototypes for the slot don't work out then probably I'd go with simple sealed arrangement, and just be inventive about getting it close to the ears/head.
In fact.. that might be the better option anyway, and not necessarily mutually exclusive. Between what Juhazi has done and what GM has mentioned, I'd expect that a pair of sealed subs could fairly easily be configured into a dual-opposing design. It would just take serious attention to how well they're mechanically coupled.
Another thought, is that if my preference for future room subs would require multiples (distributed) anyway.. would there be any reason not to simply build one very-nearfield sub dedicated to every chair? The number of subs/boxes might be slightly greater but the number or size of drivers and amps would likely be similar or even lower.
I'm just going to have to build something and give it a real-world test. To find if I'd be completely happy with nearfield alone, whether opposing drivers are particularly beneficial in that situation, and so on.
EDIT: very nearfield subs are reasonably common in enthusiast setups, but typically used for film/movie situations and often firing directly at/into the back of chairs. Upholstry materials (like many others) are apparently more transparent to lower frequencies, so no doubt it works - presumably with some slot effect at such close quarters. But as my aim is music and fidelity with appropriately modest SPL, I'm looking at more unimpeded path up at ear height.
I'm just going to have to build something and give it a real-world test. To find if I'd be completely happy with nearfield alone, whether opposing drivers are particularly beneficial in that situation, and so on.
EDIT: very nearfield subs are reasonably common in enthusiast setups, but typically used for film/movie situations and often firing directly at/into the back of chairs. Upholstry materials (like many others) are apparently more transparent to lower frequencies, so no doubt it works - presumably with some slot effect at such close quarters. But as my aim is music and fidelity with appropriately modest SPL, I'm looking at more unimpeded path up at ear height.
Last edited:
Why not simply just make a tunnel instead of U slot? Or connect two boxes with rods? Distributed means more cables to hide...
A nearfield sub most likely sounds best no matter how it is made.
A nearfield sub most likely sounds best no matter how it is made.
Yes, there are many variations possible; I might even have seen the connecting rods idea posted before, now that you mention it. Perhaps the best shape I've modelled (from an acoustic perspective) is a shallow V-slot, which could be developed further and I think would be fine for my purposes with some attention to detail.
Of course this thread is all related to the idea that I want opposed force-cancelling drivers (and one front-facing outlet), in a small box. This seems a good idea given suspended floors with other people living in close quarters and I think it would be the ideal from that perspective.
But it has been pointed out to me that the BandPass design, my previous intention, still has mileage; probably I moved on from it too early. it is bigger per driver so dual opposed configurations might not be workable in my situation, which led me to look at slots instead. However, due to acoustic gain, the cone excursion needed in a bandpass can be less for low frequencies, e.g. around half for similar SPL, which would reduce the need/benefit of having force cancelling drivers to begin with. With the right design, there might also still be the option of coupling two together for force cancelling when circumstances permit/demand, with attention to the engineering.
So many very good options! 🙂
Of course this thread is all related to the idea that I want opposed force-cancelling drivers (and one front-facing outlet), in a small box. This seems a good idea given suspended floors with other people living in close quarters and I think it would be the ideal from that perspective.
But it has been pointed out to me that the BandPass design, my previous intention, still has mileage; probably I moved on from it too early. it is bigger per driver so dual opposed configurations might not be workable in my situation, which led me to look at slots instead. However, due to acoustic gain, the cone excursion needed in a bandpass can be less for low frequencies, e.g. around half for similar SPL, which would reduce the need/benefit of having force cancelling drivers to begin with. With the right design, there might also still be the option of coupling two together for force cancelling when circumstances permit/demand, with attention to the engineering.
So many very good options! 🙂
Just as a small point of interest I compared a 30L sealed box to two 4th-order bandpass designs totalling 54L and 40L internal volumes respectively. Without exceeding Xmax, I got the following output
Unsurprisingly the bigger BP is slightly louder around 25hz which is the kind of low-end that I'm targeting.
But interestingly (to me, anyway) the displacement chart shows how quickly excursion drops off with the bandpasses, particularly the smaller one
suggesting that there would be opportunuty to EQ the lowest frequency up a little bit. As I've got plenty of power available it might be possible to use it in allowing a smaller rear BP4 chamber.
I can't do quite as much to reduce the front chamber because of the necessary tuning; the port length starts to become undesirably long. In the above examples it is about 45-50cm long and only just large enough to keep port velocity below 27 meters-per-second at Xmax. However I've been told there are ways to handle the higher frequency peak resonance, which I wasn't aware of when previously considering bandpass. So that is quite intruiging, potentially reducing one of the undesirable limitations, and so something I want to inverstigate further.
Unsurprisingly the bigger BP is slightly louder around 25hz which is the kind of low-end that I'm targeting.
But interestingly (to me, anyway) the displacement chart shows how quickly excursion drops off with the bandpasses, particularly the smaller one
suggesting that there would be opportunuty to EQ the lowest frequency up a little bit. As I've got plenty of power available it might be possible to use it in allowing a smaller rear BP4 chamber.
I can't do quite as much to reduce the front chamber because of the necessary tuning; the port length starts to become undesirably long. In the above examples it is about 45-50cm long and only just large enough to keep port velocity below 27 meters-per-second at Xmax. However I've been told there are ways to handle the higher frequency peak resonance, which I wasn't aware of when previously considering bandpass. So that is quite intruiging, potentially reducing one of the undesirable limitations, and so something I want to inverstigate further.
Haa
Sonos sub looks like
That one you posted
https://www.bestbuy.com/site/sku/6597168.p?skuId=6597168&sb_share_source=PDP
Ha, yes; I suppose it might be a similar design. Though my ideas moved on slightly; the plywood to make a stand-alone slot took up internal volume, so once I got the box dimensions down it was more space-efficient to have a full-height slot that could use/bridge the existing outer cabinet walls for at least three of its sides.
It's always attractive to use a PR system and make a very small subwoofer, but you will have the same problems that you get when you make a very small closed box subwoofer: increased distortion. Linkwitz provided an Excel spreadsheet to calculate the distortion contribution from the box size alone (the driver non-linearities will compound on top of this):
https://www.linkwitzlab.com/closed-box1.xls
Open the spreadsheet, enter your driver parameters under "Driver Data" and the closed box data under "Box Data". For a PR system just enter the rear (sealed) chamber volume. Then enter the power that you will be delivering to the woofer under "Power amplifier specs". Finally check out cell U14. This is the distortion that is only due to the driver changing the pressure inside the sealed chamber. Because the pressure is changing as the cone moves in and out, this causes the amount of cone movement in either direction to differ, and that gives rise to distortion (even order).
As the box gets smaller and smaller compared to Vas this source of distortion can rise, and it is something that everyone should be aware of, whether it is a small PR system or a small closed box system that is equalized with a Linkwitz Transform.
For lower distortion you must make the sealed chamber larger to make the displaced volume of the cone's movement (Sd*Xmax) a lower percentage of the total air volume in the box.
Take a look at the spreadsheet and see if the distortion is acceptable, keeping in mind that there will be additional distortion coming from the non-linearities of the driver itself, e.g. suspension and motor non-linearities that produce their own distortion.
This might be why small subwoofer systems that do not use e.g. motional feedback or other corrections, can sound very muddy.
Thank you (once again!) Charlie. That was a piece of the jigsaw that I hadn't quantified.
The spreadsheet indicates that at large volumes (say half of VAS or more) there is not a massive difference for quite big volume changes. Yet at smaller volumes, distortion becomes much more affected. Which focussus the mind on the repercussions other than available power, when looking to shave off size on already small chambers.
There is a lot of disagreement over what level of distortion is noticeable, let alone disagreeable, at bass/subwoofer frequencies - especially in music rather than test tones - which is further complicated by there being different types of distortion. But given how rapidly it rises (in the Linkwitz spreadsheet) with small chambers, there is clearly going to be a sensible minimum. Irrespective of subtleties or further refinements, once the chamber is reasonably small I will stop trying to make it ever more so.
For me personally, I'm in the camp that doesn't want my speakers to have a noticeable character, overlaid onto all music, even if said character were pleasant and certainly not if it were e.g. muddy. For right or wrong, I'm fairly comfortable with a few percent distortion (WRT bass drivers, rather than electronics), maybe I'd be content with several percent below 80-100hz. If I'm using the spreadsheet correctly, that means (with my driver) I'd be reasonably happy with roughly 30L, around 4% distortion, but much less is starting to seem rapidly more poor.
Very useful, thanks again!
The spreadsheet indicates that at large volumes (say half of VAS or more) there is not a massive difference for quite big volume changes. Yet at smaller volumes, distortion becomes much more affected. Which focussus the mind on the repercussions other than available power, when looking to shave off size on already small chambers.
There is a lot of disagreement over what level of distortion is noticeable, let alone disagreeable, at bass/subwoofer frequencies - especially in music rather than test tones - which is further complicated by there being different types of distortion. But given how rapidly it rises (in the Linkwitz spreadsheet) with small chambers, there is clearly going to be a sensible minimum. Irrespective of subtleties or further refinements, once the chamber is reasonably small I will stop trying to make it ever more so.
For me personally, I'm in the camp that doesn't want my speakers to have a noticeable character, overlaid onto all music, even if said character were pleasant and certainly not if it were e.g. muddy. For right or wrong, I'm fairly comfortable with a few percent distortion (WRT bass drivers, rather than electronics), maybe I'd be content with several percent below 80-100hz. If I'm using the spreadsheet correctly, that means (with my driver) I'd be reasonably happy with roughly 30L, around 4% distortion, but much less is starting to seem rapidly more poor.
Very useful, thanks again!
Last edited:
Looking at the complex spreadsheet in a little more detail, it seems the formula for box spring distortion is actually quite simple:
Box air-spring distortion in % = 140 x (one-way driver swept volume) / (chamber volume).
So a nice direct relationship; half the box size (for a given driver) means double the air-spring distortion. Or, put another way, such distortion is related to a ratio between the driver's swept volume and the chamber size.
Rightly or wrongly, it has been suggested that at lower subwoofer frequencies, most people would not find 5% distortion audible, so maybe a target of <5% at 30hz (though much less towards 50hz) at full SPL would be reasonable. The driver's swept volume at Xmax is 0.85L, so the sealed chamber should be at least 24L (+3L for driver displacement, might as well round that to 30L total).
Of course, there will be other causes of distortion and some might be cumulative, but there is a lot of leeway really. Even when used as full-room subs, I don't plan to normally listen to music at the subwoofer's peak output, that is for occasional headroom in movies etc where I'm not too bothered if exploding planets aren't reproduced at highest fidelity (from the formula, airspring distortion drops proportionally with excursion). I'm also not too bothered about preserving fidelity down through the lower roll-off, since I struggle to distinguish tonality as frequencies drop towards 20hz, even though I can hear the sound.
So I think 27L-30L is what I'll consider to be about the minimum desirable chamber volume.
Box air-spring distortion in % = 140 x (one-way driver swept volume) / (chamber volume).
So a nice direct relationship; half the box size (for a given driver) means double the air-spring distortion. Or, put another way, such distortion is related to a ratio between the driver's swept volume and the chamber size.
Rightly or wrongly, it has been suggested that at lower subwoofer frequencies, most people would not find 5% distortion audible, so maybe a target of <5% at 30hz (though much less towards 50hz) at full SPL would be reasonable. The driver's swept volume at Xmax is 0.85L, so the sealed chamber should be at least 24L (+3L for driver displacement, might as well round that to 30L total).
Of course, there will be other causes of distortion and some might be cumulative, but there is a lot of leeway really. Even when used as full-room subs, I don't plan to normally listen to music at the subwoofer's peak output, that is for occasional headroom in movies etc where I'm not too bothered if exploding planets aren't reproduced at highest fidelity (from the formula, airspring distortion drops proportionally with excursion). I'm also not too bothered about preserving fidelity down through the lower roll-off, since I struggle to distinguish tonality as frequencies drop towards 20hz, even though I can hear the sound.
So I think 27L-30L is what I'll consider to be about the minimum desirable chamber volume.
Last edited:
Here is some more info that might be useful:
For an example (to provide context about the level of distortion you can expect) of a very good, pretty low-distortion 12" subwoofer driver, see:
https://www.linkwitzlab.com/woofer3.htm
There are plots for 30Hz and 20Hz at 12mm excursion. At 30Hz the system is producing about 5% 2nd and 1.5% 3rd, with higher order products much lower. This is VERY clean. At 20Hz the system produces about 9% 2nd and 8% 3rd order, with higher order products still below 1%. That's still very good.
NOTE: the distortion measurements are for the driver in a W-frame, so there is NO AIR SPRING DISTORTION at all. This is just the distortion contribution from the driver at Xmax. You would add the box air-spring distortion on top of this.
How much distortion is OK? One way to evaluate sound quality for a subwoofer is via CEA-2010-A testing, which specified a separate level maximum for each distortion product, and the SPL needed to exceed the threshold is determined. This tell you the level of distortion at which the sub will start to sound stressed. These are:
Reference:
https://audioxpress.com/article/cta-2010-a-better-way-to-measure-subwoofers
CEA-2010-A data can be found for many DIY and commercial subs in this spreadsheet:
https://docs.google.com/spreadsheet...6ZL6anVTW2_M/edit?gid=834598950#gid=834598950
For an example (to provide context about the level of distortion you can expect) of a very good, pretty low-distortion 12" subwoofer driver, see:
https://www.linkwitzlab.com/woofer3.htm
There are plots for 30Hz and 20Hz at 12mm excursion. At 30Hz the system is producing about 5% 2nd and 1.5% 3rd, with higher order products much lower. This is VERY clean. At 20Hz the system produces about 9% 2nd and 8% 3rd order, with higher order products still below 1%. That's still very good.
NOTE: the distortion measurements are for the driver in a W-frame, so there is NO AIR SPRING DISTORTION at all. This is just the distortion contribution from the driver at Xmax. You would add the box air-spring distortion on top of this.
How much distortion is OK? One way to evaluate sound quality for a subwoofer is via CEA-2010-A testing, which specified a separate level maximum for each distortion product, and the SPL needed to exceed the threshold is determined. This tell you the level of distortion at which the sub will start to sound stressed. These are:
2nd harmonic: -10dB
3rd harmonic: -15dB
4th and 5th harmonics: -20dB
6th, 7th, and 8th harmonics: -30dB
9th and higher harmonics: -40dB
Keep these in mind when thinking about how low your sub can play and still sound good, taking into account ALL the different contributions to distortion in the system.Reference:
https://audioxpress.com/article/cta-2010-a-better-way-to-measure-subwoofers
CEA-2010-A data can be found for many DIY and commercial subs in this spreadsheet:
https://docs.google.com/spreadsheet...6ZL6anVTW2_M/edit?gid=834598950#gid=834598950
Below 50Hz distortion% should be more that 20% to be heard with sine tone. With music it can be higher without problems.
https://www.audioholics.com/loudspeaker-design/audibility-of-distortion-at-bass
https://www.audioholics.com/room-acoustics/human-hearing-distortion-audibility-part-3
https://www.audioholics.com/loudspeaker-design/audibility-of-distortion-at-bass
https://www.audioholics.com/room-acoustics/human-hearing-distortion-audibility-part-3
@Juhazi An audibility threshold of 20% distortion at 50Hz seems a bit high to me. Where did you get that number, exactly (I didn't find it in the linked Audioholics web page). For example, this table from that web page has very different numbers:
The other thing to bear in mind is that we are all very used to hearing lots of distortion in the reproduction of very low frequencies, below 50Hz, but that doesn't mean we should not try to do better...
The other thing to bear in mind is that we are all very used to hearing lots of distortion in the reproduction of very low frequencies, below 50Hz, but that doesn't mean we should not try to do better...
Thank you, both. Yes it is surprising to me how wide ranging quoted values are for distortion audibility , especially when considering music where masking becomes part of the picture, and beyond that is the further question of what might be audible yet acceptible or sufficiently unobtrusive. I've seen everything from 5% to 100% quoted at 20hz.
So I think it is probably sensible to just look at well documented methods like CEA-2010-A tests and reputable measurements of existing subwoofers to put things into context. At this stage I can plan for some things like the air-spring distortion but will not know others until a prototype is built and tested, so it is all a bit of guess work.
Re the guessing, it seems that with a Lab12 driver Art Welter once measured about 9% THD at 30Hz and 3% by 40Hz at very high power sine wave at 1m outdoors (I believe from a bass reflex design). My drivers are not quite the same, but similar enough that I might expect them to achieve something vaguely comparable. But then my interest in smaller sealed designs would add noticeably more air-spring distortion; a few percent for the kind of box sizes I can tollerate. So, maybe in total I could expect something around 12% THD at 30hz. All very rough guessing, but possibly enough to manage expectations.
So I think it is probably sensible to just look at well documented methods like CEA-2010-A tests and reputable measurements of existing subwoofers to put things into context. At this stage I can plan for some things like the air-spring distortion but will not know others until a prototype is built and tested, so it is all a bit of guess work.
Re the guessing, it seems that with a Lab12 driver Art Welter once measured about 9% THD at 30Hz and 3% by 40Hz at very high power sine wave at 1m outdoors (I believe from a bass reflex design). My drivers are not quite the same, but similar enough that I might expect them to achieve something vaguely comparable. But then my interest in smaller sealed designs would add noticeably more air-spring distortion; a few percent for the kind of box sizes I can tollerate. So, maybe in total I could expect something around 12% THD at 30hz. All very rough guessing, but possibly enough to manage expectations.
The size of chamber that I can tollerate will be biggest in a simple sealed design, because no space is taken up by additional chambers or ports. Even so, I spent all of yesterday looking at 4th order bandpass designs; there are some lovely looking results possible which are extremely attractive to me. But unfortunately, however I arrange things they are simply going to be too big for practicality in my particular situation. Sad, but it is a fact that I've reluctantly now accepted.
After which I came back to look at simpler, smaller slots; specifically at what can be done about their main detractor, that huge 1/4-wave resonance from the slot depth. If the once-parallel slot becomes more of a triangle (i.e. with almost no back wall) and then fillers are added top and bottom to make even the point of the triangle more cone-like, it looks like this will very significantly reduce the resonant peak. I think perhaps sufficiently for it to become acceptable even without EQ, though of course I would use EQ because it'll be available anyway. So my feeling at this point is that it'll be at least worth building a simple prototype (when the weather permits), to test if my belief is true.
If it proves poor then I'll just go back to basic sealed boxes. A sealed box could still be positioned very nearfield with a bit of inventiveness, and/or could later be combined in a closely-coupled manner if/when future situations want more distance and SPL; it needn't be via a slot if that method is problematic.
After which I came back to look at simpler, smaller slots; specifically at what can be done about their main detractor, that huge 1/4-wave resonance from the slot depth. If the once-parallel slot becomes more of a triangle (i.e. with almost no back wall) and then fillers are added top and bottom to make even the point of the triangle more cone-like, it looks like this will very significantly reduce the resonant peak. I think perhaps sufficiently for it to become acceptable even without EQ, though of course I would use EQ because it'll be available anyway. So my feeling at this point is that it'll be at least worth building a simple prototype (when the weather permits), to test if my belief is true.
If it proves poor then I'll just go back to basic sealed boxes. A sealed box could still be positioned very nearfield with a bit of inventiveness, and/or could later be combined in a closely-coupled manner if/when future situations want more distance and SPL; it needn't be via a slot if that method is problematic.
Charlie, ~20% treshold for lows (below 50Hz) is my observation, for myself and some friends, playing sine in-room from single speaker and slowly adding spl. Room modes will affect audibility enormously. No science behind it. -15dB is 16.2% which is said to be treshold here https://www.audiosciencereview.com/...audibility-of-low-frequency-distortion.18187/
I could not find AES articles or other source for audibility listening tests, about how the limits were defined. I am sure that there must be individual variance too. And of cource tests should be done in anechoic chamber or at least outdoors with very low ambient noise (in desert without wind or jetplanes).
https://audioxpress.com/article/cta-2010-a-better-way-to-measure-subwoofers
With music we know from scientific tests that low harmonics add spl we hear, and fundamental is not so important (psychoacoustics). This is tonal coloration similar to music instrument's spectrum of fundamentals (tonality). https://en.wikipedia.org/wiki/Missing_fundamental
With REW anyone can test distortion and audibility by sines or CE-2010, chapter 19. https://www.roomeqwizard.com/REWhelp.pdf
I could not find AES articles or other source for audibility listening tests, about how the limits were defined. I am sure that there must be individual variance too. And of cource tests should be done in anechoic chamber or at least outdoors with very low ambient noise (in desert without wind or jetplanes).
https://audioxpress.com/article/cta-2010-a-better-way-to-measure-subwoofers
With music we know from scientific tests that low harmonics add spl we hear, and fundamental is not so important (psychoacoustics). This is tonal coloration similar to music instrument's spectrum of fundamentals (tonality). https://en.wikipedia.org/wiki/Missing_fundamental
With REW anyone can test distortion and audibility by sines or CE-2010, chapter 19. https://www.roomeqwizard.com/REWhelp.pdf
Yes it is an option, the floor effectively becomes another way of making a detachable slot! Unfortunately though I've not had the best experiences with largish woofers positioned in that orientation, due to the suspension sagging over (lots of) time. I'll still use it in opposed situations where one can turn the box upside-down periodically though.
Thanks for the distortion info, too. I'm going to have to read up on how to measure these thing in REW, when a prototype exists to test. Though yes, I'm also interested in what is actually noticeable (by me and other human adults). I have little doubt that at my normal listening levels things will be okay, and especially when in nearfield mode, after that I don't mind 'too much' if the measurements are a bit more lacklustre when pushed to extremes for fun or sound-effects. But as a minimum I still want to avoid any clearly audible artefacts even then - noticeably muddy or strained, port chuffing, rattling etc. would be an indication of having gone too far.
Thanks for the distortion info, too. I'm going to have to read up on how to measure these thing in REW, when a prototype exists to test. Though yes, I'm also interested in what is actually noticeable (by me and other human adults). I have little doubt that at my normal listening levels things will be okay, and especially when in nearfield mode, after that I don't mind 'too much' if the measurements are a bit more lacklustre when pushed to extremes for fun or sound-effects. But as a minimum I still want to avoid any clearly audible artefacts even then - noticeably muddy or strained, port chuffing, rattling etc. would be an indication of having gone too far.
Last edited: