I'm kind of hoping Doug (tanstaafl) is floating around to read this (assuming he's had enough time to get that radio station ticking over like a well oiled clock).

I am unsure as to whether my speakers are in phase. I am running just a two channel installation for the moment until I get my second amp organised. I just replaced my original amp last week and I am now finding that whilst I now have better overall sound quality (notably due to using a compact subwoofer in free air as recommended by tanstaafl - good tip, young sir!), I am getting a very dead, boomy sound. Using a phase check track, I am unable to tell whether the speakers are correctly phased or not. I only get the feeling that something is not right - the stereo image does not seem to give very good seperation. What I lack is a track containing a constant-frequency tone (say, middle C) that pans from extreme L to extreme R with a pause in the centred image position.

I thought I had been very careful to ensure all the cabling is marked for phase and that it is all correctly matched on both speaker sets and crossovers, but I am now not so sure.

Is there any obvious characteristic I should be noticing in an anti-phase setup up that tells me that it is wrongly phased?

My next steps will be:

1) Re-trace all the wiring and check the phase markings, correcting where neccessary
2) If the sound is the same, reverse one pair of speaker cables and check the sound quality again to see if it is better or worse.

Anyone got any tips?

Additionally, I am going over to a 2-amplifier, 4 channel setup this weekend; how do you ensure correct phase relationships with 4 channels?

Over to you....

Well, last I heard, Doug was going to be offline until later this week. So you probably won't get an answer from him any time soon. But I think I can help, though.

I just happened to have experienced a phase problem day before yesterday, and I can tell you exactly what it sounds like.

Here's what happened: I just bought a new car, and I'm in the process of transferring all of the components that I ripped out of my old car. I'm doing it in little stages. The first step was to replace the cheap factory 6x9 rear-deck speakers with my bitchin' aftermarket 6x9's.

Well, the factory wiring on the speakers was not labeled +/-, and the color-coding didn't help, either. The factory speakers were likewise not labeled, and I don't have one of those handy-dandy Crutchfield wiring harnesses which takes care of the problem for me.

In the end, I'll be wiring my own amplifiers directly to my own speakers, so I'm not concerned with being able to wire them correctly when that time comes (all of my aftermarket components are properly labeled). But just for that evening, all I wanted to do was to get the factory radio working through those aftermarket rear deck speakers. So I figured, "as long as I get the phase between the left and right rear speakers the same, the worst that can happen is a phase difference between the fronts and the rears, and that's not as bad because the delay time difference screws up the phase between fronts and rears anyway".

So I just plugged in the wires, crossed my fingers, and turned on the radio. I set the fader all the way to the rear and played with the balance. Almost instantly I could tell that the phase was reversed on one of the speakers. I didn't need a special test track or a constant-frequency tone. Here's how:

The radio was tuned to a Jazz station, and there was a lot of plucked acoustic bass in the song. When the balance was in the middle, I could hardly hear the bass. The incorrect phase on one of the speakers caused the bass to cancel itself out when both speakers were playing the same bass notes. When I set the balance hard left or hard right, the bass came through loud and clear. I checked my wiring, and sure enough, I had wired one of the two rear speakers backwards. Switching the wiring on that speaker fixed the problem instantly.

In theory, if the fronts and the rears were similar in timbre, you could do the same thing in a 4-channel setup. But usually, they're not, so you might be out of luck in that case.

I assume that your problem is similar to mine, in that you're using the factory speakers and the factory speaker wires, and they aren't clearly labeled. One way to figure it out is as follows: Buy an aftermarket wiring harness for the head unit (mine is on order right now, which is why I couldn't do this yet). The aftermarket wiring harness should match up to your factory wiring plug in the dash perfectly, and it will have the speaker wires labeled +/- properly. Then, look at the color-coding of the wires coming out of the back of that factory plug. Those colors should be the same as the wires leading to the factory speaker connectors. Use that as your guide.

As far as your complaint that you're not getting "good separation", that can be caused by a lot of problems, not just phase. Incorrect wiring, an amplifier set to mono subwoofer mode, or simply the fact that in-car audio traditionally doesn't have good separation can all be the cause.

Now you also said that the sound is dead and boomy. That could be phase, or again, it could be any number of other problems. If this is your first time hooking up the Empeg, you may be surprised that its default EQ settings are totally flat. There is no bass or treble control, and the loudness is off by default. Many aftermarket stereos have their outputs tweaked a little bit, even at the default settings. So if you compare a flat-set Sony CD player to a flat-set Empeg, the CD player will probably have a bit more bass and treble there by default. After you fix your phase, play with the Empeg's EQ a bit and see what you think.

Tony Fabris
Empeg #144

Well, I'm back in the world again, after a pretty intense 6,000 mile cross country trip. (Intense? When I hit Canada, I bought a bucket of Kentucky Fried Chicken and a two litre bottle of root beer, and for the next two days meal time meant slowing the car down from 75 MPH to about 60 to facilitate throwing the chicken bones out the window. They're biodegradable, right? One day I went well over 1300 miles, about 2100 kilometers.)

Rob, I'm afraid I won't be much help to you in this area -- Dominic at Cambscar would be more useful. I have not done any of my own installation work -- I leave that to the pros. Instead, I rely on what I unabashedly consider to be a very good ear to determine what is right and what is wrong with my stereo. I know quite a lot about what to do to fix the problems if it involves tuning and equalization. But hardware and wiring problems I don't get into. My IASCA competition disk has a track on it specifically designed to demonstrate out-of-phase problems: it is a vocal track with a woman saying "My voice is in phase now, and should sound coherent and be placed in the center of the sound stage. My voice is out of phase now, and should be diffuse and difficult to locate. In addition, it may lose low frequency response." or words to that effect. When she goes out of phase, if your system is properly set up, you can really tell. On cars with speakers out of phase, she sounds the same both ways. Another useful track on the IASCA CD is a series of seven drumbeats that pans left to right across the sound stage.

Having always relied on professional installers (I pick the equipment and tell them where I want it installed, and they do the grunt work) I have never had any sort of wiring or phasing problems, so don't know exactly how I'd deal with them. But you are probably on the right track (or was that Tony) by swapping wires and seeing if it makes things better or worse. Another thing you might be looking for could be phase cancellation -- you get a pair of speakers pointed at each other and if you are unlucky... what am I saying, you already know far more about this sort of thing than I ever will, I think, judging from some of your posts and emails.

What size subwoofer did you install? I have found that a single 10" works very well indeed. No, lets see, in the Netherlands you don't use inches do you? It's pounds or liters or something... I've got it, a 10" speaker is exactly .0039665404 furlongs in circumference, or to put it another way, about .000012521043 acres in cone area, give or take a few kilograms. That should allow you to easily visualize the size speaker I'm talking about.

Say Hi to Henno for me...

tanstaafl.











"There Ain't No Such Thing As A Free Lunch"

As a follow up to the original post, I dug through my wiring and found that I had indeed reversed my left channel output pair, resulting in phase problems. With the wiring corrected, the stereo image and bass response improved. A noticable difference.

So thanks for your help you guys, I will be applying this wisdom in the future.

The woofer I put in was a compact 16 cm (6 inch). It is pretty tiny, but it is about the only size I could get into the car (a Mini). It is currently running open-air with no enclosure, resting on the car floor. This has made quite a subtle difference to the listenability of the music of the car - I am finding at higher speeds, I am not having to jack up the volume so much.

So, on to 4 channels!

It is currently running open-air with no enclosure, resting on the car floor.

Speakers need the proper enclosures to sound their best. As soon as you can, make sure that all of your speakers are properly enclosed. Of course, they have to be made to go into enclosures. Some are designed specifically for open-air use, but when they say "open air" what they mean is that the unit shoud have the cone facing into the vehicle and the back of the cone facing out of the vehicle (for example, mounted to a sealed baffle board with the back end facing into the trunk or into the door frame). Open-air woofers depend upon there being a seal between the inside of the car and the outside- with the pressure difference between the two supporting the woofer's ability to create sound.

Low-frequency speakers have to be able to move air, you see. If you have a speaker just "sitting there", then the air behind the speaker is getting moved just as much as the air in front of it, and since it's the same air, it doesn't get much of a chance to make noise (at least not the way it was intended). If you put it in an enclosure (properly sized, of course), then the enclosure resonates properly and provides backpressure on the cone, and the front of the cone moves the air to your ears the way it was intended.

It's amazing how much better a properly enclosed speaker sounds than a speaker that's just "sitting there". A properly enclosed speaker will play louder, deeper, and tighter than you might think.

Tony Fabris
Empeg #144

OK, but then why do most bass speaker enclosures have a reflex port?

I was always told the enclosure channels the sound energy, hence the need for a port to allow the speaker diaphragm to move air on the return movement. I was under the impression that this was why reflex ports had exponential horns built in for "impedance matching", and that sealed enclosures were a "bad thing" due to the effect of the case pressurising on a return movement, which added a damping factor to half the amplitude of the sound waveform.

Surely with bass tubes and reflex enclosures being so popular, this isn't so?

OK, but then why do most bass speaker enclosures have a reflex port?

This is a very good question, and I anticipated it.

A bass reflex (ported) enclosure is one of the two possible ways to build an enclosure. A sealed-box enclosure is also called an "Acoustic Suspension" enclosure.

Some speakers are designed to work in ported enclosures, others are designed to work in sealed boxes (but you have to get the cubic volume of the sealed box correct).

With a ported enclosure, the "port" is specifically tuned to allow only a certain volume of air to escape per second. There is still a little bit of backpressure on the speaker. In a ported enclosure, the bass radiates from both the speaker and the port. The port usually emits frequencies in a certain narrow range, depending on how it's tuned. Keep in mind, though, that a ported enclosure is still an enclosure- the speaker's not just sitting out in the air.

A ported enclosure will sound louder than a sealed enclosure, requiring less power to drive it. But it will sound less accurate, more "boomy", its frequency response will not be flat across the speaker's range (it'll have a spike around the port frequency), and it won't extend to very low sub-bass notes. Ported enclosures are popular because they can get louder with less amplification.

A sealed enclosure (when properly sized to the speaker), will require a bit more power to drive properly. But it will have a flat frequency response, it will extend to lower bass notes, and it will sound tighter and more accurate.

I have firsthand experience with this- I had a ported, pre-assembled, self-powered subwoofer in my hatchback for a while. After it got stolen, I assembled my own sealed-box unit from component parts. For about the same amount of money as I'd paid for the ported unit, I got a sealed-box unit that sounds orders of magnitude better, and that's what I'm using now. I was surprised at how much better the sealed-box unit sounds.

Okay, I lied above. There's actually a third way to create a bass enclosure: it's called a "Bandpass" enclosure, where the speaker sits between two chambers inside the box, and bass radiates only from the port(s). This design has the problem of narrowing the frequency response to a very narrow range- the range of the port(s). It's fine for people who just want a loud "boom" and aren't interested in the nuances of all the low frequencies.

Tony Fabris
Empeg #144

I am now sufficiently interested to look deeper at this; I originally went for the open speaker approach based on Doug's descriptions and experiences, and yes, when I tried it, it appeared to make a difference. It was also due to the fact that there are no real locations inside the cabin of a Mini to place an enclosed bass unit of sufficient volume.

What I would like to do is design and build an enclosure from damped marine plywood to a custom shape to fit in to one of the few locations I can place the woofer. So, have you got any URLs or reference materials on the design of bass unit enclosures?

What I would like to do is design and build an enclosure from damped marine plywood to a custom shape to fit in to one of the few locations I can place the woofer.

That's a great option. It's possible that there are companies out there who've already done this and have pre-made enclosures for those odd locations. In the Crutchfield catalog, there's a company that has a line of these enclosures for American cars (mostly pickups). Perhaps there's a European company specializing in Minis?

I don't have any enclosure-building URLs handy, but Crutchfield has a book on loudspeaker design that I think you can order.

Tony Fabris
Empeg #144

Neither ported or sealed speakers have flat responses. This is almost impossible to do with a passive speaker (transducers and crossovers with amplified input) without a huge amount of expense.

To my knowledge, sealed speakers have the advantage of a well known response - which is something like 6dB/octave below the resonant frequency. Couple this with an appropriate crossover and you can get a reasonably flat response with only a few components.

While it's true that cheap speaker designs use porting simply to get some extra oomph at low frequencies, it has valid uses in "real" speakers too. Some speaker manufacturers tune the bass port to be resonant at frequencies which the bass cone naturally isn't because of the material it's made of. I remember reading that the B&W DM-60x series have it aligned to the null frequencies of the kevlar cone. There's also the issue of phase that porting can either address or destroy - depending on how well it's tuned. A well tuned bass port will deliver bass in-phase to the midrange/tweeter cones.

The ideal speaker design would use all active components (powered at line or digital, not amp level). I've been toying with the idea of buying myself a decent measurement microphone, recording a frequency sweep from my speakers, and using that to adjust the response in software, digitally.

For the measurement mic, www.linearx.com's M31 microphone ($150) comes with a floppy disk with all the calibration data (frequency response measured in anechoic room with known speakers). If I aligned my speakers to that, I'd have a response as good as the microphone, which is +/-1dB. Has anyone had any experience with measurement microphones, speaker measurement or anything related? In theory, I can get some absolutely superb sounding speakers simply by applying the inverse of their response. No idea how it'll turn out in practice, but I'm sure I'll have fun trying :)

Anyway - just a few thoughts I wanted to throw in...


- John (from empeg)

(The above may not represent the views of empeg :)

This is pretty much exactly what I am trying to do, on the cheap.If you want to subsidise your effort John, then let me know as I am interested too.

The only real experience I had with this type of measurement was at GEC when we were designing the truly horrible IXT telephone, where we had to take "in head" accoustic measurements using a Bruel & Kjaer stereo measurement system, the mike of which was planted inside an accoustically neutral dummy head. We then ran full audio spectrum sweeps (even outside the classic 300 - 3400 Hz telephone audio bandwidth) to asses the audio properties of various bits of the telephone handsets and so forth. The Anechoic chamber became quite popular for private "after hours" testing of headphones and loudspeakers, with one bloke taking his entire stereo in to characterise the system so he could tweak it when he got it home.

I wish I had that dummy head and the measurement set now to do my Mini with. **sigh**. Unlimited tweaking possibilities....

Actually, it has just struck me - the Mk2 has got a speech recognition input, right? So what would be so difficult about the empeg playing back a pre-ripped full-spectrum, flat -3dB sweep while the empeg sampled the input from a microphone placed in the driver's head position? It could then look at the response and compensate for any dips/peaks introduced by the amp, speakers or listening environment.

Just think, you'd be able to dial in "Flat", "Bass-cut", "Treble-cut" playback characteristics, it would sweep and analyse the response, then kick the equaliser settings to match the environment and characteristic you just asked for!

A self-configuring head unit with automatic spectrum compensation! Brilliant! Talk about a USP!!!

I suspect the supplied microphone would be more tuned for voice, with appropriate filters in there to block out cabin noise and pinpoint a voice. Not the ideal sort of thing to measure a full spectrum sweep.

Calvin

Here are my thoughts, and is somewhat pie in the sky but I'll throw them out as well.

I suspect that with speakers, signal response varies by frequency AND amplitude. So you won't be able to get away with an adjustment function that uses frequency alone. You would need to run frequency sweeps at a large number of different volume levels in order to map out a plane adjustment function of volume and frequency. Considering this, if the equalizer software can be adjusted to take values out of this plane table, in theory you would write some software that runs for a few hours doing many many frequency sweeps recording the proper equalization plane. It might also be a good idea to record equalization planes from all the possible seating positions because in a car, the distance to various speakers varies so greatly that you'd need to have multiple planes. So you can chalk up "number of passengers in a vehicle" as a component of equalization.

In my own car, I have a convertible top so sound quality varies by whether the top is up and down and the vehicle speed-- because wind noise begins to mask bass above a certain velocity. So if the empeg is aware of these other parameters and adjust them in real time that would be a godsend. With my alpine I'm constantly pushing and turning and adjusting the sound at highway speeds to make it sound proper, and then when slowed to street speeds it sounds awful and the need to readjust comes up again :(.

So the second pie in the sky thought is, I would be willing to give up voice recognition for a much more expensive microphone that is dedicated to monitoring cabin noise and adjusting for it automatically. (I've seen volume adjustments based on a speed sensor, but that's not quite right... because response doesn't uniformly decrease with higher speed).

I think if I had to put a value on adjustment software like this, I'd say $500 or $1000. So I'd and probably audiophiles and gadget freaks would be willing to pay for add on software to the empeg for doing this very thing. It wouldn't make sense to develop it for people who wouldn't use it or understand why sound might need to be tuned for number of passengers, vehicle speed, presence of a top, volume level, frequency, etc... but there are some that might be willing to pay :>

Calvin

Slight problem that the mk2 can only sample at up to 22.05kHz...

Hugo

Hmmmm.... but you could use it as an input if the fidelity of the input circuit was reasonable and you were prepared to invest in a reasonable pickup microphone. Overkill for just voice recognition, yes, but if you bought yourself a quality input mike for sweeps and compensation work, you could swap this in and out as needed. A feature I would be willing to pay for, even though the use would be primarily occasional.

Sampling at a rate of 22 kHz, or sampling up to a frequency of 22 kHz?

Up to 22kHz audio range would surely be enough for a full audio spectrum sweep at different levels, or am I missing something?

Even if it was only a rate of 22k, that would still give you a sweep range of 11kHz. I can't hear much above 16k anyway, and in a car I'm not so interested in high frequency response, more low- and mid-range.

Rate of 22.05kHz, ie 11kHz bandwidth.

Not ideal I'm afraid. We might be able to push it higher (maybe 30k) with some trickery, I've not tried. No promises ;)

Hugo

Hugo, I never count on it - it would be unfair, no matter how nice - but in this case, I await your and Patrick's normal levels of creative magic

0 - 15kHz sampling capability would still go a significant way to providing a way sample a useful chunk of the audio spectrum. Being able to just tell the empeg to sit and perform scans for half an hour while I go for a cuppa to come back to a car with equaliser compensation automagically put in to play after it has been sniffing the air would be a dream come true. Pwetty, pwetty pweeeeaaase!

Oh damn, I've just remembered I have a Mk1 . Still that would be a good reason to buy a Mk2! "Darling, I need to buy another empeg 'cos it can intelligently setup it's own spectrum compensation for me" (smack) I suppose sex is out of the question, then? SLAM!!!

There is another alternative.

Sony and Pioneer both sell sampling equalizers, complete with microphone. Some IASCA competitors use them on the pink noise track of the RTA analysis portion of the competition. The competitor places the microphone on the judge's microphone stand right next to the judge's microphone, presses some button or other, and the equalizer self-dials itself into a perfectly flat response. Takes about 30 seconds or so, I believe.

Understand, however, that if you tune your stereo to give you a flat response to a pink noise track, it will sound *terrible* when you play regular music. Don't get too carried away with this auto-equalization idea. Perfect, flat equalization doesn't sound like what the human ear is expecting to hear. Other than for esoteric competition purposes, you might be better off equalizing by ear, making it sound the way that seems best to you, rather than trying to conform to some arbitrary "ideal" of flat equalization.

Some competitors actually have their cars equipped with multiple equalizers and switch between them during competition: one for sound quality, one for RTA, and one for SPL. Others (such as myself) just make notes of the system settings that are optimal for each phase of the competition, and re-adust them as appropriate. This is a bit risky on a complex system, as it is quite easy under the pressure of competition (you have very limited time to make changes) to make mistakes, tweak the wrong band or forget to turn a Loudness switch on or off. My system is simple enough (not even an equalizer) that I can do it pretty easily. For example, to tune for SPL, I just swap the output plugs on my amps so the "big" amp (a mighty 100 watts) feeds the subwoofer, and set my head unit volume to 34. At that point I produce an incredible 112 decibels, and other competitors laugh at me because a lot of them actually *listen* to their stereos at volumes louder than that. (By the way, they stop laughing when the sound quality points are tabulated...) The judges have told me that I have the quietest car that has ever competed seriously in IASCA.

Anyway, I still think I would trust a good ear more than I would turst any automated equalization process.

tanstaafl.

"There Ain't No Such Thing As A Free Lunch"

In this discussion, you're forgetting one thing...

Microphones have their own unique response characteristics. In order to do something like what you're talking about, you'd have to use a pretty expensive microphone. Also, the software would have to be pre-calibrated to that specific mic so that it could compensate for its idiosyncracies.

Tony Fabris
Empeg #144

We might be able to push it higher (maybe 30k) with some trickery, I've not tried. No promises ;)

Wow, Hugo, now you've got my attention. You'd get enormous publicity if you were able to get self adjusting equalisation into Emma (and/or her home equivalents). Recently, a couple of home amplifiers were introduced that included that capability. I can't find the issue of the (German HiFi/High-End) magazine Stereo right now that ran an article on them. I'll hunt some more and will send you a copy if you're interested.

Henno
# 00120 (6GB+18)

I tend to agree with tanstaafl; equalisation seems highly dependent on the ear of the listener. Tony sort of chimes in when he mentiones calibrating for the idiosyncracies of a mic -- which is effectively the "ear" used for auto-equalisation.

Remember -- unless you're competing it's what YOU hear that counts, not what a tuning mic hears. This is what equalisation presets are all about: letting you take the time to tune the settings to account for specific cirmstances and then allowing you to quickly/easily recall those settings "on the fly".

-- Bleys

"If you would judge, understand." -- Seneca

What about the possibility of making some basic equalization adjustments? Like increasing bass to compensate for road noise or simply altering volume according to ambient noise? These may not sound as sexy, but they would probably be more usefull to your average user.

-Mike

it's what YOU hear that counts, not what a tuning mic hears

Of course, but starting from a flat response would be much easier than having to do all the adjustments by ear from the response of four usually mediocre speakers playing within the odd-shaped car enclosure. I bet it would give many of us a much better chance to arrive at nice equalizer settings. I'll look up the article and let you know how they did it.

Henno
# 00120 (6GB+18)

Exactly what I am trying to do - if I could start out with a level playing field (excuse the pun) then it will be far easier for me to then start changing settigns for musicality. I am having a bad time of setting up at the moment using the empeg equaliser because although the equaliser is set up to be flat, I don't know what this translates into after it's been fed through the loudspeakers and amp. If I could look at what the empeg samples and then sets up as "flat after speaker output", then I would be able to say "too much treble" and wind it down. But I need to see what the spectrum looks like first, which says sampling. Since I haven't got access to a B & K audio spectrum analyser, then the empeg itself is the next best thing.

Microphones have their own unique response characteristics. In order to do something like what you're talking about, you'd have to use a pretty expensive microphone. Also, the software would have to be pre-calibrated to that specific mic so that it could compensate for its idiosyncracies.

I did think of this, which is why the microphone I'm looking at getting (the M31 from www.linearx.com) comes with a floppy disk with all the information about its response you need for compensation. They're the only company I've seen which provides this kind of information. Usually you'll find it's all integrated into a DSP in the back end of the mic, which is a less accurate and more expensive method than just doing it in your measurement software.


- John (from empeg)

(The above may not represent the views of empeg :)

I suspect that with speakers, signal response varies by frequency AND amplitude. So you won't be able to get away with an adjustment function that uses frequency alone.

You're right - what you're describing is nonlinearity (where wattage doesn't map in a straight line to displacement), which is what THD+N and IMF ratings indicate. It is possible to adjust for this too (class T amplifiers do this I think) but it's much more difficult. The difficulty here, I suppose, is actually building measurement devices which have very low distortion (expense expense!)

Something that springs to mind is that speaker which uses ultrasonics to direct sound into a very narrow cone. The nonlinearity of air is used predictively to generate (a few feet in front of it) lower audible tones with sound quality similar to that of an ordinary speaker. You could apply the same techniques to ordinary speakers to get lower distortion - if you can find a suitable measurement device. There are patents (pending) for the ultrasonic speaker, but I don't know if they are generalised or not.


- John (from empeg)

(The above may not represent the views of empeg :)

Understand, however, that if you tune your stereo to give you a flat response to a pink noise track, it will sound *terrible* when you play regular music. Don't get too carried away with this auto-equalization idea. Perfect, flat equalization doesn't sound like what the
human ear is expecting to hear. Other than for esoteric competition purposes, you might be better off equalizing by ear, making it sound the way that seems best to you, rather than trying to conform to some arbitrary "ideal" of flat equalization.

The fact that a perfectly flat response sounds terrible means that recording studios aren't using speakers with flat responses. As far as I know, most speakers have a slightly dulled response at 1-4kHz (some of the vocal band). The thing is, with enough data you can make any speaker sound identical to another. So if you like the sound of one competitor's system, you can just clone it


- John (from empeg)

(The above may not represent the views of empeg :)

Doug: Don't get too carried away with this auto-equalization idea. Perfect, flat equalization doesn't sound like what the human ear is expecting to hear.

John: The fact that a perfectly flat response sounds terrible means that recording studios aren't using speakers with flat responses.


I don't understand, Doug/John. Are you saying that without an equaliser a system sounds 'terrible'?

I have a pretty good sounding home system I think, that plays beautifully without any equaliser in the chain. All components are optimally tuned to deliver as closely as possible what was recorded: CD head delivers 10-20kHz between 0 and -0.2dB!; the amplifier section is also totally flat (within 0.1 dB!) between 20Hz and 20kHz; and the response of the speakers is as 'flat' as you can get them (with a bit of roll-off on the bass side - my wife doesn't like big speakers). No adjusting anywhere. Just lots of power (300W per channel) to accurately drive the speakers (limited to protect the speakers )

This system never sounds loud, even at pretty high volume levels. Everything remains easy, transparent; light; lots of 'air', if you know what I mean. Playing exactly as the recording engineers intended.

IMHO, there are two types of equalisers one could use: simple ones, like the one built into Emma to make gross corrections to the tonal balance to compensate for poor components / bad listening environment. With these, output can be polished up to make it sound impressive but the resulting sound will be very different from the sound that was intended. I agree, in the car Emma needs adjustments and I too have found a few Eq settings that make Emma shine. But the result is very different from the way the same album sounds at home. I have never measured the output of 'my' equaliser settings but I doubt that it will be flat: I guess that bass and treble are boosted more than speaker fall-off. Also, it is remarkable that with these adjusted frequencies, Emma can sound really loud.

The other type are the complex ones that adjust at many frequency points. Especially the new computer controlled ones are said to be very effective in optimizing listening environments by making (small) volume adustments at hundreds of different frequencies, compensating for speaker deficiencies and imperfect accoustics; basically making the output of the combined components as flat as possible at the intended listening position. I've never listened to one, so I can't judge. In a review of the German magazine Stereo they used one to analyse an already optimised room. It made many (small) adjustments and the resulting frequency curve was indeed more flat than before, but music played as beautifully as before. Proving that a good ear (and patience and an understanding partner) can be equally effective.

For me, equalisers are a last resort; I try to avoid them whenever possible. In a car, however, I find them unavoidable and I'm glad with Emma's embedded Equaliser controls. Above anything else though, I prefer unadjusted playback in a good playback / listening environment that allows the skills of the recording engineers to shine.

Henno
# 00120 (6GB+18)

AgreeMsg

Calvin

Henno: In a car, however, I find [equalizers] unavoidable

Well, yes and no.

I have what might be considered a distributed equalizer -- no equalizer per se, but enough individual controls to do a pretty good simulation of an equalizer. That is, my bass is on its own amplifier -- so I can adjust bass response separately from the rest of the system. My mid-range speakers each have a separate L-pad (resistor pot) on them so I can adjust both the amount and location of mid-bass. My treble speakers can be adjusted by the gain on the head unit. And then there are the coarse bass/treble and loudness settings on the head unit itself. So, while I have no equalizer in the car, I can in actuality do quite a lot of equalizing.

John: The fact that a perfectly flat response sounds terrible means that recording studios aren't using speakers with flat responses.

Well, maybe... I was referring specifically to a pink noise track in competition, in which the object of the game is to divide the 20Hz -- 20KHz spectrum up into 1/3 octave intervals, giving 32 separate bands, and then make each one play at exactly the same decibel level as all the others. That gives a straight (flat) line across the spectrum analyzer, and it is pretty impressive when a competitor has the kind of control of his system that allows him to tune like that. But listen to music with that equalization, and it sounds terrible--not enough bass response, not enough treble response, it seems like it's all mid-range.

I kind of think we're not talking about the same things here when I refer to "flat" on a pink-noise competition, and you refer to flat equalization of music.

tanstaafl.



"There Ain't No Such Thing As A Free Lunch"

I kind of think we're not talking about the same things here when I refer to "flat" on a pink-noise competition, and you refer to flat equalization of music.

I'm pretty sure that a flat spectrum test (the pink noise calibration you refer to) is actually not very flat in terms of the EQ settings required to pass that test.

Most stereo equipment is designed to sound good to the ear rather than to a calibration test. So most of the stuff you buy off-the-shelf is going to have far from flat response. So when you (Doug) say "flat", you're referring to the pink noise test. But when your average consumer says "flat" he means "whatever my stereo sounds like natively with all the tone and EQ controls set to zero".

Tony Fabris
Empeg #144

I kind of think we're not talking about the same things here when I refer to "flat" on a pink-noise competition, and you refer to flat equalization of music.

A "flat" response is the audio equivalent of wysiwyg for graphics. What goes in comes out exactly as the recorded medium asked. No matter how you do your equalisation curve you should get the same result. It doesn't matter whether you use pink noise, white noise, or plain music - you're trying to fit the output curve to the input curve.

The ugly truth is that if you have a perfect wysiwyg (and no distortion, intermodulation etc) system and it sounds awful - then it's telling you the recording you have is awful. Because what you're hearing IS the recording, unaltered. The problem may be that the recording studio decided to equalise for the average speaker system.


- John (from empeg)

(The above may not represent the views of empeg :)

if you have a perfect wysiwyg (and no distortion, intermodulation etc) system and it sounds awful - then it's telling you the recording ( . . ) is awful.

Makes sense. Thanks for clarifying

Henno
# 00120 (6GB+18)

I am now sufficiently interested to look deeper at this; I originally went for the open speaker approach based on Doug's descriptions and experiences, and yes, when I tried it, it appeared to make a difference. It was also due to the fact that there are no real locations inside the cabin of a Mini to place an enclosed bass unit of sufficient volume.

I've seen a product which takes advantage of one of those 'unused' spaces in a car - beneath the front seats. I can't remember what its called, but it functions basically as an enclosed speaker transmitting (pretty much) directly into your - spinal column.

What I don't know is anything more than the manufacturer's hype - like whether they actually work, or whether they're too big to fit under the seats of a Mini, or whether they just give thump and no tones.

All questions for someone else.

Save the whales. Feed the hungry. Free the mallocs.

They work ... well in fact. They're expensive, I'm told. I know someone who has a pair under the seats in his Corvette -- we checked his to see if they'd fit under the seats in my MR2. I've got about 5cm of clearance beneath my seats and it does not appear to be enough space.

I can get make/model information if you're still interested...

-- Bleys

"If you would judge, understand." -- Seneca

In the UK, Maplin Electronics sell these things, or at least they used to. A few years ago Patrick (empeg hardware designer) and myself worked with a company developing immersive simulators and we tried these little subs under the seats (which were fibreglass Sega seats at the time). They didn't work very well - "rattled" a lot I seem to recall.

The last time I saw the project they had uprated to 10" drivers in a ported enclosure right behind the player's back (and replaced the horrible Sega seat with a proper rally car seat). I don't think they need hydraulics to make the thing shake now :-)

Rob

It was a 450W 12" driver, actually, in a 45L enclosure. Driving right into the kidneys of the person sitting in the seat. Go in for a game, come out with a nosebleed and backache :) One of my better efforts, shame they scrapped it.

Patrick.

Opinions expressed in this email may contain up to 42% water by weight, and are mine. All mine.

Warning: pre-historic thread.

For the measurement mic, www.linearx.com's M31 microphone ($150) comes with a floppy disk with all the calibration data (frequency response measured in anechoic room with known speakers). If I aligned my speakers to that, I'd have response as good as the microphone, which is +/-1dB. Has anyone had any experience with measurement microphones, speaker measurement or anything related? In theory, I can get some absolutely superb sounding speakers simply by applying the inverse of their response. No idea how it'll turn out in practice, but I'm sure I'll have fun trying :)

Have you tried this yet? I'm very interested in this too.

For the M31, what would you connect it to to do the actual measurement? It seems designed for use with Linearx's boards and software and those are expensive.

How about the Radio Shack sound-level meter (cat# 33-2050)? People have good things to say about it on the net and there are published corrections that compensate for its response curve. I'll probably give it a try.

Borislav