To properly design the best sound system possible, it helps to understand things like how we hear, how speakers reproduce sound, how the speakers interact with one another, and how the sound they reproduce react in the environment. It is also helpful to know what modifications within each of these we can make to get to our goal. At times we may have to make compromises, and having a good understanding of all of these things will help us to know which compromises to make.
Our hearing is a very complex thing. There are physical aspects of it like the shape of our head, shape of our ears, and how damage over the years will effect the way we perceive sound. However I have found that there are some psychological aspects as well. Psychoacoustics plays a part in what we experience when we hear something, as well as personal experiences shaping our expectations.
We laugh at those who spend a fortune on expensive wire. However I do believe them when they say they hear a difference. Although the difference may not be a physical thing that could be measured, it is perceived by them psychologically and therefore the purchase was worth the cost. Most likely those people have already spent their money in all the places that measurably make a difference.
On the other side of the coin, you may also laugh at the guy who is bragging about how good his Jensen system sounds. The reason he thinks it is the best is because his personal experiences with listening to sound systems is very limited. Maybe he's only heard his buddies factory speakers distorting like crazy trying to keep up with the dual 15's in the pre-fab bandpass enclosure. So this Jensen system sounds so much better that he is happy. There is another reason he believes his Jensen system is so good. The reason is that when we spend a lot of money (from our own perception) on something, we will convince ourselves that it is the best thing out there to justify the choice. We see this every day on many of the online forums as people brag about how good brand X is. And others who have moved on from brand X to brand Y are arguing with him because obviously brand Y is the best. Of course even their tune will change when they move on to brand Z in the future. That's why you should be careful taking advice based on someones opinion. Look for facts to back up their recommendations if possible. Now back to the Jensen owner; It isn't until he finally gets a chance to listen to a really high end home stereo system that he finally realizes just how good audio can sound, and how lacking his Jensen system is. He didn't know what he was missing.
So we have looked at some of the psychological aspects of how we hear, now let's discuss the physical. This is a very complex topic with which our knowledge is expanding daily. But there are a few key points that I will discuss here. The first one is the Haas effect. Here is what Rane specified as the definition of this:
"Also called the precedence effect, describes the human psychoacoustic phenomena of correctly identifying the direction of a sound source heard in both ears but arriving at different times. Due to the head's geometry (two ears spaced apart, separated by a barrier) the direct sound from any source first enters the ear closest to the source, then the ear farthest away. The Haas Effect tells us that humans localize a sound source based upon the first arriving sound, if the subsequent arrivals are within 25-35 milliseconds. If the later arrivals are longer than this, then two distinct sounds are heard. The Haas Effect is true even when the second arrival is louder than the first (even by as much as 10 dB.). In essence we do not "hear" the delayed sound. This is the hearing example of human sensory inhibition that applies to all our senses. Sensory inhibition describes the phenomena where the response to a first stimulus causes the response to a second stimulus to be inhibited, i.e., sound first entering one ear cause us to "not hear" the delayed sound entering into the other ear (within the 35 milliseconds time window). Sound arriving at both ears simultaneously is heard as coming from straight ahead, or behind, or within the head. The Haas Effect describes how full stereophonic reproduction from only two loudspeakers is possible." http://www.rane.com/par-h.html
The second thing we should know about is where we get timing cues from to produce a good stereo image. This also has to do with how our ears are spaced apart. For frequencies 1000 Hz and above, we get our timing cues strictly from intensity differences. In other words, if the left speaker produces a frequency within this bandwidth louder on the left side than the right, the image will appear to come from the left side. For frequencies 500 Hz and down, imaging cues come strictly from timing differences. In other words, if the left speaker produced a frequency within this bandwidth on the left side sooner than the right, the image would be to the left. For frequencies between 500 to 1000 Hz, we get our imaging cues from both timing and intensity differences.
It is also important to understand distortion. As long as none of the equipment is being pushed beyond it's designed limits, the most likely type of distortion we will hear will be harmonic. What harmonic distortion is is unwanted artifacts of the fundamental that is harmonic in nature. So if the fundamental frequency was a 50 Hz tone, harmonic distortion will be a peak of sound at 100 Hz (second harmonic), 150 Hz (third harmonic), and so on. All speakers produce harmonic distortion, it's just that some are worse than others. Some companies are posting THD+N specs, but those aren't very helpful. The reason being that it isn't the total level of the distortion that is important, but the spectrum of the distortion. In other words, a high level of second order distortion isn't that annoying. But high levels of third harmonic distortion is. Here's a study performed by Dr. Earl Geddes and his wife Dr. Linda Lee showing this to be true: http://www.gedlee.com/results.htm.
Dan Wiggins once told me that 90% of a good soundstage is a flat frequency response. Having personally experienced his speaker designs and the amazing imaging they produce, I am very much a believer. However I have heard many people on the car audio forums in the past complain that systems with high RTA scores didn't sound very good. Well IMO there are two things going on here. First of all, I'm wondering if they formed this opinion on initial listening, or if they gave it some listening time. I've noticed that there is a break-in period for our ears. The likely reason they don't like the flat RTA sound is because they are used to their music being what I like to call "surreal". What happens is over time we have lost some of our hearing, especially at certain frequencies. To boost these lost frequencies can makes it sound better to us. Also, having a little boost in the bass is kind of fun, and boosting midrange makes the music sound warm and the lyrics more legible. We can see an example of this in the popularity of tube amps which add some second order distortion for that "warm" sound. Although more pleasant, the modifying of the signal isn't producing a faithful reproduction of the original recording. That's the difference of what I call objective SQ (high fidelity) and subjective SQ. So if a flat frequency response is so important to a good soundstage, do we have to give up our subjective sound? Probably not. What I recommend is starting from a flat response, then make those slight changes you enjoy so much. That way you can always go back to the objective SQ if tastes change or mistakes are made.
Speakers are the weakest link in any audio system. They have the worst frequency response and produce distortion magnitudes greater than any other piece of equipment in your system. Speaker design is also a very large topic because of this. There are generally three common speaker sizes we see in car audio, the tweeter, the woofer, and the subwoofer.
So let's start with everyone's favorite, the subwoofer. This write-up is about SQ, so I will ignore the SPL designed system. For the most part you want a driver that will produce a flat frequency response in your vehicle. Although a lot of source material doesn't have it, it is still good to be able to produce even the lowest octaves if possible. So what do we want to look for in a good subwoofer? Well first of all a good amount of excursion is handy. This is because it benefits us two fold. First of all for every halving of frequency, the excursion capability of the subwoofer must quadruple to keep the same output. The second reason for a good amount of excursion has to do with motor linearity and distortion. Earlier I talked about harmonic distortion and how third harmonic distortion is really bad. Well as it turns out this harmonic distortion is mainly caused by motor nonlinearities.
So what are these nonlinearities that were mentioned above? Well a speaker has a fixed magnetic gap with a coil of wire suspended in it. When current is applied to the voice coil it repels from this fixed magnetic flux due to the opposing magnetic polarities. The physical parameter BL (B being the amount of magnetic flux effecting the number of turns L of the coil) is derived from this and is measured in Tesla Meters or Newtons per Ampere. Well about 99% of subwoofers use a motor topography called overhung which is a small magnetic gap with a long voice coil. As it turns out there isn't only magnetic flux between the pole piece and top plate. There is also what is called stray flux or the fringe field that resides above and below the gap. You can see this here: http://www.adireaudio.com/Files/SGLC-Driver.gif What happens is at rest the BL will generally be at its highest. However as the coil starts moving, it will start to leave some of this stray flux and BL will start to drop. This creates distortion. Here's a DUMAX test of a Kicker Solobaric BL curve: http://www.adireaudio.com/Files/DUMAX/S12D.pdf You will also notice the Kms curve which is the suspension nonlinearities mentioned earlier (Kms=1/Cms). As you can see, the BL starts to drop off fast with excursion. The more of a change in BL, the higher the distortion will be. So how does having more excursion benefit us? Well a subwoofer with more excursion playing the same frequency as a subwoofer with less excursion and at the same volume level will be in a more linear part of the BL curve. So therefore distortion will be lower.
So what else do you want in a subwoofer? Well despite what many manufacturers want you to believe, a large motor size is not necessarily one of them. In other words, a larger magnet does not necessarily mean a better subwoofer. Huge BL numbers can actually be a hindrance rather than a help. This is because BL directly effects Qes which is a large part of Qts. So when we raise BL, Qes (and along with it Qts) goes down and raises Fb. So too much BL can actually hinder your low frequency extension from the subwoofer. You want just enough BL for the given design goals. Also note that thick magnets aren't as good as wide one's. Adding another slug (doubling the thickness) of magnet generally only increases the available flux by about 10%. It's surface area along the top and bottom plate's that make a difference. Also try to stay away from subs with bumped backplates. The biggest choke point for magnetic flux is at the back plate/pole piece junction. Bumping the backplate throws away a lot of your available flux. This is OK if the design allows for it, but many times it doesn't. For the cone, you want it to be rigid enough to stay pistonic without being too heavy. Now I know that people worry a lot about efficiency. But IMO this is really the last thing you should look for in a subwoofer. Because if you choose to go with an efficient driver, you have to give up either low frequency extension or enclosure size. And neither of those are things I'm willing to give up. Especially since power is cheap and trunk space usually isn't. After all, would you try to hold your breath to conserve oxygen in the middle of the forest? Also if possible look for a higher Qms and lower Qes. This means that Qts is dominated more by the magnetics rather than the suspension where driver tolerances are at their worst. I am also not a big fan of tinsel leads woven into the spider as they will have a negative effect on the life and performance of the suspension on the driver. I also don't recommend dual spiders as they will magnify nonlinearities in the suspension.
Midrange drivers are much more complex as they play the band of frequencies our ears are most sensitive to. You want them to be able to play low enough to match up well with the subwoofer without a problem. One weakness in many systems I've heard is the lack of midbass performance. The result is a sub system that sounds boomy. So how does it make the sub system sound boomy? Well every instrument has a harmonic signature that allows our ears to identify what it is. This includes instruments that play in the lower frequencies as well. However if you lose this harmonic content due to a lacking midbass section, the instrument will sound boomy with no definition. So it is important to have a rigid cone on the midrange to produce these lower frequencies. However you also don't want the cone to be too heavy since Mms will act as a low pass filter. Too heavy a cone might limit your upper frequency extension. Other important things are the dust cap design as well as it's attachment to the cone, and the surround/cone attachment as they will all effect the response of the driver. The dust cap often resonates creating stored energy within the usable frequency band. Or the glue joint might create a mode that produces a dip in the response. Phase plugs are rather popular as you can smooth out the upper frequencies nicely. However you do lose some rigidity you get with the dust cap on the cone. I personally recommend against metal coned mids as they tend to ring in the upper frequencies. And this ringing cannot be removed electronically as it is excited from nonlinearities in the drivers motor.
As I said earlier, you want to select a midrange to be able to play midbass frequencies without a problem. Many times this means selecting a driver with a good size cone. However the larger cone will also effect the upper frequency limit you want to take the driver to. This is because the speaker will start to beam. Beaming occurs as a drop in the off axis response of the driver. The further off axis you are, the more of a dip there is. The frequency the driver starts to beam at is determined by D^2/lambda. As an example, a 6.5" driver will start to beam at around 3.2-3.5 kHz. This is important to keep in mind when selecting a driver if you plan on using it up that high.
Tweeters come in many varieties and utilize many different cone materials. Tweeters can sometimes cause a balancing act as you try to choose the best bandwidth while not limiting it's output. The problem is that with any driver the lower you try to play it, the less output it will have before it starts to break up. Many manufacturers rate their tweeters to very low frequencies. However you must know at what output level this rating was obtained. Just because it is rated down to 1.5kHz, doesn't mean you can take it down that far without any problems. When selecting a tweeter, it is a good idea to know how low you want to take it. This is generally determined by the midrange's upper frequency performance. It might be a good idea to take a tweeter down lower to eliminate beaming concerns or poor response from the midrange. IMO it is best to start by selecting a good midrange, and then choose the best tweeter to fulfil your needs from there. Cone material IMO isn't as important as many people think it is. People attribute metal tweeters as more harsh, bright, etc. and silk tweeters as more smooth, natural, but early to roll off. However I believe this comes more from the voicing of the crossover than anything. I've heard soft dome tweeters with plenty of upper extension, and metal tweeters sound very smooth.
I actually had a big wake up call about how well we hear in the upper frequencies. I have extensive experience with the Creative Sound Solutions WR 125S full range driver. It starts to roll off on axis at 15 kHz and is 3 dB down at 18 kHs. And because of beaming as it's a 4" driver, it is much worse off axis. However I am utilizing one right now as my center channel speaker in my home theater, and barely notice the losses. There has to be a conscious effort to know those upper frequencies are missing.
Many times on car audio forums you will see claims that install is 90% of what makes a good sounding system. Well I disagree. I believe the crossover is THE most important part of any speaker system. Have you ever heard of the term "voicing" the speaker? Well this is done in the crossover design. It determines the final results of the speaker system as it determines how well the speakers interact. But it is important to understand that a crossover is much more than just an electrical slope at a certain frequency. A properly designed crossover will contour the acoustic response of the drivers in their intended location. It can include level matching, equalization, impedance correction, etc. Now I'm not saying that install isn't important. It does make a large difference as many people have proven. The problem comes from the fact that in a vehicle speaker locations and the environment are so problematic to sound that the small differences you can make with the install won't be enough to overcome those problems. And this is more true if the signal doesn't get reproduced very well to begin with. For the most part the best speaker design in factory locations will sound better than a poor speaker design in the best install within a standard vehicle. So getting the sound correct when it leaves the speakers is very important. I believe the claim became so popular because those who said it are unable to do anything about the crossover, so they put their focus on what they can do some thing about, the install. And because of that, it becomes the most important, to them. But it's still not an accurate claim IMO.
Band-aids are helpful too
OK, so now you have the perfect speakers working together in harmony, allowing superior quality sound to enter the vehicle. Unfortunately the environment that this great sound is entering is very hazardous and will greatly effect the quality of the sound that will finally get to our ears. Luckily we are now armed with the above knowledge. We now can find the best locations to install these speakers in so the environment has the least effect on them. Just keep in mind that install will only get you so far. Due to the horrible seating locations, you will never get the sound that the best home stereo can produce. But you also have some other tricks to help you out. A good processor can fix quite a few of the problems your system faces. The more complex the processor the more you can correct, but the harder it is to set up right. A good digital processor will give you memories to store different settings so you can have the sweet spot be at different seat locations, as well as a best for both seats setting. And if you are especially brave (and have deep pockets) you can go fully active and get even more control over each individual speaker. But again setting this up properly gets extremely complex, requiring very expensive test equipment to achieve the best sound possible. Good luck!