Now that I am designing another set of speakers
Corrugated Cardboard Bookshelf Speaker Project
I am left with the realization that the speakers will not sound good without equalization.
I have been using equalizers and tone controls all my life, and especially in my car. For most music, in my car, that is, I simply cannot do without it, and most music sounds unacceptable without it. I just tried listening at the FLAT setting and really was very disappointed, as I am most of the time, except if I suppose my music is loud enough so I cannot hear ambulance sirens. My cardboard speakers will sound OK with tone controls and loudness but is that the point?
So do I design the speakers to work with an inbuilt equalizer or L-pads or something like that? I listen at 70 dB where the loudness effect occurs, Fletcher Munson curves etc.
I hope this is not a duplicate thread but I have already looked at the four or so threads on the topic.
equalizer
Including the above.
Corrugated Cardboard Bookshelf Speaker Project
I am left with the realization that the speakers will not sound good without equalization.
I have been using equalizers and tone controls all my life, and especially in my car. For most music, in my car, that is, I simply cannot do without it, and most music sounds unacceptable without it. I just tried listening at the FLAT setting and really was very disappointed, as I am most of the time, except if I suppose my music is loud enough so I cannot hear ambulance sirens. My cardboard speakers will sound OK with tone controls and loudness but is that the point?
So do I design the speakers to work with an inbuilt equalizer or L-pads or something like that? I listen at 70 dB where the loudness effect occurs, Fletcher Munson curves etc.
I hope this is not a duplicate thread but I have already looked at the four or so threads on the topic.
equalizer
Including the above.
I did find some information in an article here:
What they admit is:
1. Equipment is not perfect
2. Recordings are not perfect
Therefore use an equalizer to make the sound more even.
No one addresses the loudness effect, though.
Good advice here
What is the best volume to listen to music? - Quora
What they admit is:
1. Equipment is not perfect
2. Recordings are not perfect
Therefore use an equalizer to make the sound more even.
No one addresses the loudness effect, though.
Good advice here
What is the best volume to listen to music? - Quora
Last edited:
Lets separate this into two sections: Speaker EQ and loudness compensation (Fletcher-Munson).
No passive equalizer or crossover can adequately equalize a speaker. There are two reasons: 1. The response excursions from flat are far too many and too small for low Q passive EQ, simple graphic equalizers, or even 1/3 octave graphics, and 2. The speaker interacts with its acoustic environment to further complicate the result.
The only way to really EQ a speaker is to do it in situ…in the room, at the listening position…by taking detailed multi-point measurements, then reacting to these with one of several DSP-based equalizer systems that can compensate well, some even taking the time domain issues into account.
Have a look at miniDSP.com for a few relatively low cost products, but also study the differences between those and their DIRAC product line.
Realize that simple 1-octave EQs are like performing brain surgery with an axe, and 1/3 octave graphic EQ guarantees that no control will be centered where you need it. Something with full parametric EQ would be a reasonable compromise, something like a Behringer FBQ2496 is actually quite useful. Troll eBay for a used deal.
And also realize that your cardboard box speakers are going to have lots of resonances that you won't be able to EQ, the penalty for not building a real speaker box of rigid materials. But, oh well.
Now, loudness....
Real Loudness Compensation is, first of all, not really Fletcher-Munson curves (they’re a bit wrong), and not a fixed amount of compensation based on a volume control setting. The the requirement for proper loudness comp is dynamic and depends on several things, most prominently the instantaneous volume of the reproduced sound, but also very important is the specific differential between the volume at which the material was mixed and your personal playback volume. And there’s a problem. Outside of the film industry where there are well defined reference levels at which all soundtracks are mixed, there is no reference level. The music industry has not standardized on mix levels at all, so at best, the differential between your level and the original is an educated guess, at worst, it’s completely wrong.
Therefore, practical loudness compensation of music becomes subjective, something you’ll have to adjust to taste, and realize that it won't ever be spot-on dynamically. However, practical loudness compensation of film mixes is actually quite easy if you use something like Audyssey Dynamic EQ, which “knows” the original level, and “knows” the playback level moment by moment, and adjusts dynamically in response to the differential of specific frequency bands. Audyssey DEQ only comes when you use Audyssey MultEQ to calibrate and equalize your system, however. Some feel Audyssey MultEQ is a bit lacking (and it will definitely be challenged with your cardboards..), and there is a very significant difference between the various flavors of it, with the best being MultEQ XT32 with the Pro Calibration kit. However, the cost for all of that makes the miniDSP solution fairly attractive for two-channel systems, lacking, of course, real loudness comp.
If you don't mind an eBay search you could come up with a Denon AVR of a few years back with the features you need, but at a huge discount because it doesn't do 4K video. The Audyssey Pro kit shows up there too, but they will ding you for a Pro Cal license. I feel it's worth the expense, but that's just me.
Taking measurements, crunching the results, and setting an equalizer with it is not exactly simple. The miniDSP along with a USBMic and RoomEQ Wizard software is a good way to learn by doing, but the learning curve does have an upward pitch. Audyssey's setup is fairly well automated, and therefore not a big deal to learn. miniDSP's DIRAC product is also automated, but there's a cost. And you still need the mic.
I gotta admit, all of this seems a bit of overkill for cardboard speakers!
Ugh. That's just a terrible explanation. Especially 2.
Well, sort of. His second paragraph is assuming way too much, but 85dB is the "educated guess" I referred to in the last post. From that you can derive basic loudness compensation, though because it's static, it won't be right all the time.
You'll end up adjusting to taste.
Modern "cheapo" "hi-fi" units have equalizers to compensate for their less than ideal speakers. Hook a set of these "boombox" speakers to a nice audio system and they'll be sorely lacking. Hook a nice set of speakers to the "boombox" and they'll almost certainly sound terrible. What gives?
The equalizer (not switchable) compensates for some of the speaker's deficiencies; typically bass deficiencies. This works pretty well in practice. I reverse engineered some dumpster "tabletop hi-fis" and they all worked this way. Furthermore some of the amplifiers could drive real "hi-fi" speakers pretty good when the equalization was wired around.
I applied this principle to some nice equipment, correcting the "Q" of overdamped air suspension speakers with an electronic circuit. With powerful and efficient woofers and a "real" amplifier it works fantastic. The penalty is that you lose 3-5 dB maximum volume but it plays plenty loud and clear.
You can accomplish this with a DSP as suggested or build an analog circuit like I do (the hard way). For your case I would recommend a DSP.
The equalizer (not switchable) compensates for some of the speaker's deficiencies; typically bass deficiencies. This works pretty well in practice. I reverse engineered some dumpster "tabletop hi-fis" and they all worked this way. Furthermore some of the amplifiers could drive real "hi-fi" speakers pretty good when the equalization was wired around.
I applied this principle to some nice equipment, correcting the "Q" of overdamped air suspension speakers with an electronic circuit. With powerful and efficient woofers and a "real" amplifier it works fantastic. The penalty is that you lose 3-5 dB maximum volume but it plays plenty loud and clear.
You can accomplish this with a DSP as suggested or build an analog circuit like I do (the hard way). For your case I would recommend a DSP.
No, you don't compensate for ear frequency response. You can EQ some room effects, not all.
It’s about trade offs. With a small speaker you do what you can enclosure-wise to get reasonable bass extension, and then eq gets you a flat pass and down to the point where the speaker isn’t really working. In my experience passive eq just doesn’t work, as you’re trying to provide gain around the speaker resonance, so have an impedance that’s all over the place. This means you’ve got to do it before the power amp, either using active opamp circuitry or else with dsp.
Of course every dB that you add is a dB that you lose in dynamic range. So if you want a speaker that’s got good bass extension and plays loud without distorting, well you’re going to have to use a bigger driver, or several small ones.
Of course every dB that you add is a dB that you lose in dynamic range. So if you want a speaker that’s got good bass extension and plays loud without distorting, well you’re going to have to use a bigger driver, or several small ones.
Ears differ on an individual level, but across total population they are remarkably similar.
Regardless, hearing response is the "reference", as it's what we have heard our entire lives (barring damage or degeneration). Basic hearing response is how we hear life. The goal is for reproduced sound to match a live performance. Both are heard through the "hearing response filter", so hearing response cannot be removed if that match is to be correct. When hearing is damaged or there is hearing loss due to age or disease, there can and should be a correction applied because the resulting response provides inferior hearing performance in life, and of course that individual doesn't match "normal", for which life and all recordings should match.
So no, we don't attempt to compensate for hearing response until that is done in hearing corrective devices like hearing aids. High frequency loss is common, and correctable to a point, then not correctable. 90dB of loss at 10kHz is not uncommon for people in their 80s, but is not correctable.
I was afraid of that. I guess we are all sound engineers creating the final master.
OK
The speaker response curve looks like a mountain range, and can just as easily be flattened.
I can remember reading about a DSP that could compensate for a speaker placed in a cardboard box. I suppose that is what I would need.
Got it. ( I mean the idea) That one goes for a few hundred dollars.
Maybe I will try baked earth next.
The volume it was mixed at will be between 85dB and maybe 90, which narrows it down somewhat, so what I would have to do is to achieve the same balance as a 85dB mix at 70 dB which I listen at, and also have the flexibility if I decide to play it at realistic levels if I have the opportunity.
OK
Yes it is, but if i ever get to a true Hi-Fi system whatever that is, would a DSP be essential or will it be good value for money in the system?
I have hooked up cheaper amplifiers to large Hi Fi speakers and the other way around they don't sound too bad except for lack of power.
Do you have more details on this? A circuit diagram perhaps, since I am using cheap PC speaker amps for my experimental builds right now, notably the TEA 2025 and the BA5406 hopefully which I can get repaired.
What does the circuit look like? bandpass filter?
Well...yes...sort of. But in that lies technique, and an understanding of the practical. The learning curve is also mountain shaped.
Or MDF, which is easier to cut and shape into a box.
Of course we don't KNOW that...and it could easily have been mixed another 5dB higher in some genres.
Generally correct, but the "same balance" as an 85dB mix cannot be achieved with a single fixed correction. Just sayin'.....
I have yet to find a system at any cost or level of sophistication that didn't benefit from EQ, and the more sophisticated (like FIR-based), the better. A client once had me work on his high-end system, and that was a system totaling over $350K in hardware. After EQ (calibration with Audyssey), he was blown away at how much better it sounded. Of course, the room was crap...but anyway....
Ever single system should be carefully analyzed and equalized. It's one of the most audible benefits you can buy, right up there with room acoustic treatment (which is first), only cheaper and easier.
It would be wiser to invest in a good second-hand "real" amplifier.
Depends on the measured response.
Good point. Hearing loss.
Need to compensate for that also.
1 or more high efficiency horn tweeter(s) and a dsp will take care of that.
Don't compensate for hearing loss in the sound system! Your reference is still "life" around you. You'll make the system sound way different from "life", and that will sound wrong.
You compensate for hearing loss with a modern DSP-based hearing aid, so everything you hear is equally compensated for, and the compensation is calibrated to your individual ears based on measurement of their sensitivity to various frequencies.
I doubt the hearing aid will sound as good as a good sound system.
And we all want the good sound. No?
Sure we want good sound. But our perception of "good" is what we hear every day in life. Compensating for HF hearing loss in system EQ won't make it sound "good" because it's different from your life-reference, you're not measuring your exact hearing loss. HF loss will be different in each ear, and there's no way to compensate for individual ear response with speaker/system EQ as both ears hear both speakers and all the room reflections of direct and off-axis sound.
No, system EQ to compensate for hearing loss is completely wrong.
On a practical note, I do suffer from hearing loss on the right side. The EQ compensation I use is for the lower frequencies where I do not have a hearing loss, also, I do not hear much difference between right and left ears, or what I hear from them, except I used to adjust the headphone levels on the right side.
One opinion at: Feature Article
One opinion at: Feature Article
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.