Yeah got thinking, system is limited by excursion anyway, one is not going to get more 40Hz or 50Hz, than the system is capable of regardless of EQ but the context is what content is played as its not single frequency but a bandwidth with transients and all. maxSPL at any frequency is defined at the knee, where the response graph touches the max volume displacement/max power graph.
Ok, so it's subjective assesment of sound. Take a party track you remember, think kick drum hit and what it takes for it to sound nice and hit like it was ment to sound to make people party. The system just needs to be able to reproduce that, extension needs to extend sufficiently below whats on the recording. The track is most likely high passed already by the mixing engineer to get as much hit as possible, with the sound they wanted. If playback system high passes too high the sound is gonna be ruined. So, its just the volume displacement capability you need, something that can't be fixed with DSP. Only thing that DSP can fix is to ruin the sound by high passing higher than where the meat is so all it does is it takes away what's on the track. Ok, so we do not want to lose what's on the track, so there is two options which are leave it be and just push it to some distortion when needed, or multiply the box or any other means to increase volume displacement to get distortion free nice SPL with low enough extension, low enough for the content that's on the platter.
So, distortion is by necessity if cannot make the system bigger. Do not rob the lows. Assuming tracks are already high passed, another high pass filter for a closed box is unnecessary if box is small enough to limit excursion already so group delay is what it is. Only EQ needed is to level out above the preferred low knee.
So, small closed box, that does not need high pass filter, what can we do for best sound? equalize it so that extension is sufficiently low for the music and flat/house curve above. If there is not enough volume displacement available for the occasion there is going to be some distortion, that's it.
Turbo button from previous post could increase maxSPL but cost in sound quality. Perhaps make turbo button select low knee so that there is about 6db difference in max system SPL, meaningfull difference between the two states. Difference would be about half an octave for closed box? For max kick, perhaps one at 50Hz so sacrificing sq below for some genre gets meaningfully more output. The other knee, with 6db less headroom for the whole system would then be around 35Hz or so, seems reasonable.
You could advertize the product has crazy loud output* and very low extension*.
*)the turbo button switches between the two.
Distortion always happens, there is never enough bass, group delay is what it is.
Ok, so it's subjective assesment of sound. Take a party track you remember, think kick drum hit and what it takes for it to sound nice and hit like it was ment to sound to make people party. The system just needs to be able to reproduce that, extension needs to extend sufficiently below whats on the recording. The track is most likely high passed already by the mixing engineer to get as much hit as possible, with the sound they wanted. If playback system high passes too high the sound is gonna be ruined. So, its just the volume displacement capability you need, something that can't be fixed with DSP. Only thing that DSP can fix is to ruin the sound by high passing higher than where the meat is so all it does is it takes away what's on the track. Ok, so we do not want to lose what's on the track, so there is two options which are leave it be and just push it to some distortion when needed, or multiply the box or any other means to increase volume displacement to get distortion free nice SPL with low enough extension, low enough for the content that's on the platter.
So, distortion is by necessity if cannot make the system bigger. Do not rob the lows. Assuming tracks are already high passed, another high pass filter for a closed box is unnecessary if box is small enough to limit excursion already so group delay is what it is. Only EQ needed is to level out above the preferred low knee.
So, small closed box, that does not need high pass filter, what can we do for best sound? equalize it so that extension is sufficiently low for the music and flat/house curve above. If there is not enough volume displacement available for the occasion there is going to be some distortion, that's it.
Turbo button from previous post could increase maxSPL but cost in sound quality. Perhaps make turbo button select low knee so that there is about 6db difference in max system SPL, meaningfull difference between the two states. Difference would be about half an octave for closed box? For max kick, perhaps one at 50Hz so sacrificing sq below for some genre gets meaningfully more output. The other knee, with 6db less headroom for the whole system would then be around 35Hz or so, seems reasonable.
You could advertize the product has crazy loud output* and very low extension*.
*)the turbo button switches between the two.
Distortion always happens, there is never enough bass, group delay is what it is.
Last edited:
do you mean boomy bass? not sure if that happens with closed box ever, except how the system interacts with room? or does it? Perhaps one note bass is possible when the box is too small making high Q and lots of power pours in, heat increasing DCR making the Q even bigger.
Last edited:
It's a very theoretical question with vague parameters, but I probally choose SPL as compromise. I don't need to have it very loud at home. THX standards are nice, but many do with much less and in most homes, you need to take the neighbours in account (so don't go too loud).
@tmuikku
Your DSP idea is not that crazy, I actually have designed something very similar for a prototype speaker for a certain company.
It had a couple of presets that could be changed by the user. Or maybe would go on automatically after abuse for x-amount of time. (we have never implemented that yet).
It had a party mode, where low-end was limited as well as a compressor being used.
It also had a night mode, to squeeze down the dynamic range and low end to be a little less annoying for roommates/neighbors.
It worked extremely well 🙂
People call this thought experiment "vague", but those are actually real choices you have to make when designing for certain customers.
Especially with a limited budget and box volume.
For me they are everything except vague, it's an inherent part of (any) system design.
One of the first things I ask almost any client.
If they rather want lower, louder, tighter or cleaner. (these are obviously not quite correct subjective words one has to use with non-technical people) .
Your DSP idea is not that crazy, I actually have designed something very similar for a prototype speaker for a certain company.
It had a couple of presets that could be changed by the user. Or maybe would go on automatically after abuse for x-amount of time. (we have never implemented that yet).
It had a party mode, where low-end was limited as well as a compressor being used.
It also had a night mode, to squeeze down the dynamic range and low end to be a little less annoying for roommates/neighbors.
It worked extremely well 🙂
People call this thought experiment "vague", but those are actually real choices you have to make when designing for certain customers.
Especially with a limited budget and box volume.
For me they are everything except vague, it's an inherent part of (any) system design.
One of the first things I ask almost any client.
If they rather want lower, louder, tighter or cleaner. (these are obviously not quite correct subjective words one has to use with non-technical people) .
With an active EQ you can give a closed box any response you want?
If boomy bass is the goal of the project, it's just a matter of EQ'ing it accordingly or use a LW-transform.
(I consider that the same thing as EQ'ing).
Cone excursion will quickly become the limit here obviously.
On the same way you can also give a BR system the same response as a closed box system.
Just tune it very low and add a bit of EQ to follow a 0.5 Q factor target curve.
When the box volume isn't to small, this works actually very well 🙂
In general I always add at least a 1st order filter or very low param EQ dip to even closed systems to protect them from excessive cone excursion.
For bigger subwoofers, this is obviously a bit less of an issue.
But when one wants to squeeze out every bit of a bunch of 6 inch woofers or so, that's a very different story.
The penalty here is mostly group-delay and a tiny bit of low-end (although not much if you do it well, couple of Hz).
In the end most customers complain more about damaging a speaker than group-delay.
In fact, I have never heard a complained about group-delay in the last 15 years... ever
Not even from the most picky audiophile people (often not knowing that the system had a bit of extra group-delay).
Hi mrKlinky, yeah if there is DSP the system Q can be made almost anything sensible, including 0.5. Boomy bass is very annoying, but its easy to EQ if its the box. Also room boom can be EQ:d at particular position.
How would you listen such thing, room makes the response a rollercoaster regradless? Is there some audible quality with driver/box Q, after making sure with EQ its not boomy in room?
^ditto, closed box can and should be EQ'd, like any bass system, because of effects of room alone and its easy to adjust response of the box at the same time. In fact, I'm not sure if anechoic response matters at all, all we need is enough volume displacement ad EQ. EQ would depend on room and positioning, it is only possible to make some sensible average response like Q0.5.
How would you listen such thing, room makes the response a rollercoaster regradless? Is there some audible quality with driver/box Q, after making sure with EQ its not boomy in room?
^ditto, closed box can and should be EQ'd, like any bass system, because of effects of room alone and its easy to adjust response of the box at the same time. In fact, I'm not sure if anechoic response matters at all, all we need is enough volume displacement ad EQ. EQ would depend on room and positioning, it is only possible to make some sensible average response like Q0.5.
Last edited:
Ok vague was maybe a bit too much.
In my case let's say some information was missing: type of load for sub ( reflex, sealed, TL,...), value for max spl expected, box volume and field of application ( P.A., monitors, theater, casual stereo listening).
I tend to favour closed box and low Qtc ( ~0.577 through Linkwitz Transform via dsp) as i'm sensible to accuracy in time domain and i expect my loudspeakers to be able to reproduce spl i ask. I'm ok with a bit less frequency extension and a bit of distortion if i have to compromise on something for home listening in low end. If it wasn't sub/low way of multiway these requirements could change a bit or a lot.
But it's for a given field of application ( home listening which i expect to be like in studio more or less).
For PA (and/or partying- theater) it would be a whole different set of compromise hence the issue about 'vagueness'.
Last edited:
That's in the first post; closed box.
Although, in an active system this is not always very interesting to know.
Other absolute values (like value for max SPL) are not important for the question.
The question is rather what you would like to go first if you just need more out of the same system in a particular case?
I deliberately left out the field of application, mostly because I think most people would automatically think of home/studio use.
More importantly, I am just curious about what people say and why.
Regardless of the application or why in such and such application.
So maybe someone only works with or likes PA/sound-reinforcement type speakers and designs accordingly because of taste.
That is also totally fine! 🙂
I am actually quite surprised about the results.
Mostly just very interesting to read the thoughts behind the votes! 🙂
Mostly just very interesting to read the thoughts behind the votes! 🙂
Distortion isn't as important as some believe. Typical distortion levels from a good design are not audible. However there are many other problems that seem to sound like harmonic distortion, so HD gets blamed unfairly (and causes us to spend too much time looking for the real problems).
I think if you look at Erin's reviews and measurements @bikinpunk , this correlation holds up extremely well.
Some of the best speakers he reviewed are not the best performing in distortion.
Sure, at some level, there is a argument to be made to want everything as "good" as possible.
But a resonance or some other artifact ALWAYS sounds bad, doesn't matter if the distortion (or other parameters) are super low.