This article cannot possibly cover all the products out there that have the ability to afford great sound, so we shall jump into the "proper installation" category. First up is a discussion of a typical sound quality system and the theories behind its design. I can field any Q's you have as far as system component quality in a one-on-one basis in the Q&A section. And I am leaving the tuning section for later.
Typically, a system will contain a sub-woofer set-up with a mid/tweet combo and a possible extra set of tweeters somewhere. More complex systems will go one step farther by adding either dedicated midbass drivers or front-mounted subwoofers. In either case, the number of subs varies, but the midbass and midrange drivers are restricted to one pair. This is done due to the phenomenon known as "multiple point-sourcing", where having more than one pair of drivers playing the critical imaging-cue frequencies can cause double-imaging, phase errors, and loss of image focus because the sound emitted seems to emanate from more than a single location. There are instances where multiple midbass and midrange driverÃƒÂs can work very well, but tend to be impractical in the car due to restricted space. Some guys do two pairs of midbass drivers in the doors with good results, but careful crossover selection and/or time alignment between the drivers is of paramount importance here. Let's keep it simple, shall we.
As stated, multiple point sourcing is BAD. The IDEAL speaker is a "true point source", meaning it would be able to reproduce accurately ALL audible frequencies from the same cone (hence the name "point source"). At present time, there is no such speaker. Electrostatic home speakers don't count, as they are a "line-source" which emits sound from a large area having a vertical configuration. BUT, that is a whole other can of worms and we needn't be concerned with it at this time. So, this is one of the reasons many high-end SQ systems use a minimal # of drivers, because we net the best possible imaging characteristics and the least number of phase problems to correct (more on this below). Coaxial designs (with the tweeter mounted above the voice coil of the midrange) and true "point source" coaxials (with the tweeter actually resting INSIDE the midrange voice coil) were built on the "point source" premise, and can do a good job in terms of imaging because of this. Standard coaxial designs should not be considered high-end though, as they contain merely a 6dB/oct cap to crossover the tweeter, and likely have NO crossover for the midrange, allowing it to roll-off it's freq response on the top naturally. Furthermore, the bottom of the tweeter can present itself as an obstacle directly in the path or the sound wave radiating from the midrange. Some coaxial designs are true bi-ampable systems with outboard passive crossover networks and off-center tweeter mounting capabilities.
They also contour the backside of the tweeter mount to minimize interference with the mid's output. Likewise, true "point-source" coaxials have outboard high-quality passive crossovers and bi-amp capability, making them a great choice in any system. However, those are limited in terms of mounting flexibility and power handling, but in terms of correct phase relationships, they are hard to beat.
Let's no move on and talk about phase for a second. There are two types of phase relationships we encounter in a car, mechanical phase and absolute phase. Mechanical refers to the physical polarity of the drivers and takes into account the phase shift that occurs in any type of crossover network. Absolute phase pertains to the arrival times of the sounds these drivers produce to the listener's ears. We can take, for example, a high-end coaxial speaker and wire the crossover, midrange, and tweeter in correct polarity, and we would seemingly have correct mechanical phase. However, two important factors come into play---the phase shift at the crossover frequency, which depends on the type of filter and slope in dB/octave, and the physical position of the tweeter compared to the position of the midrange cone. In our coaxial, the tweeter will be an inch or so ABOVE the midrange cone. This will make the tweeter's sounds arrive at our ears earlier than the midrange frequencies. So we have a system that is supposed to be "mechanically" in phase, but are not in "absolute" phase.
When a system isn't in absolute phase, negative sonic effects will happen? For starters, this can have a bad effect on stage height and width, but the biggest problem area is in the focus and detail of the imaging. Each car will respond differently, so these are generalizations. What most guys do to try to get proper phase relationships with a bi-ampable coaxial design is they will wire the tweeter in reverse polarity. This is a way to mechanically delay the signal going to the tweeter, and the delay at high frequencies can be as much as 3 milliseconds with this method. By doing this, it allows the mid's sound wave a little extra time to reach our ears, and brings both the mid and tweet arrival times closer together, improving absolute phase response and image focus.
Aside from offering benefits in terms of speaker placement options and power handling, component sets with separate tweeters and mids can allow us to use different angles for the tweeters and midranges. If it is found that these angles are necessary for smoother response or proper phase relationships. It should be no secret that most tweeters are domes, and when compared to cone midranges, they tend to have greatly improved off-axis sound dispersion, lending to better off-axis response. A component set allows us to experiment with different mounting locations to find the best "blend" of tweeter and midrange response characteristics in our cars. For example, most hard-dome tweeters are very bright, and if not tuned properly, they are TOO bright, leaning toward sonic harshness.
By simply turning these tweeters more off-axis (more in a cross-firing configuration), you direct their point of sonic intensity away from you, thus allowing a better and more natural blend with the midrange without destroying tonality or imaging characteristics (they still will have decent response and dispersion patterns off-axis). This is kind of getting into the "path-length/sound intensity trading" as mentioned earlier, which we will delve into later on.
Soft-dome tweeters are not as bright as metal domes b/c their frequency response tends to exhibit a gentle "roll-off" above 16KHz. This makes them more natural sounding by virtue of design, but also makes them favor a more on-axis mounting scheme. There are a few soft dome tweeters out there that have extended high-end response (Dynaudio, for example) because they were designed SPECIFICALLY for off-axis mounting in cars. (The Dynes actually were designed for kick panel applications where extended freq response and off-axis mounting yield the best results and best path lengths). Conversely, hard domes, by virtue, do not exhibit much high-freq roll-off at all, and likely extend ABOVE 20KHz (which we cant hear anyways). Many mfgÃƒÂs of hard dome tweeters must in turn design their crossovers with switchable resistances to tame the tweeter's output depending on mounting location. These are the "-2dB, -3dB, -4dB, etc" settings for the tweeters. We can't possibly cover which is "best" in this article, as it depends on your car and your mounting location. But I am trying to give you the tools needed to use your own ears to find what works best for you.
As mentioned earlier, some SQ systems can benefit from an extra set of tweeters to help solidify the sense of stage height and assist with stage width. Recall in the previous article where we discussed having a "stable" soundstage height, where it presents itself at the same level from far left to far right. Often, when a system with kick panels using a single mid/tweet combo is designed, the combo of speaker angle/positioning and phase correction to get a really focused, "eye-level" center image will adversely affect the stage width and height at the sides. This is sometimes referred to as the "rainbow effect" where you get a sense of the stage height being nice and high in the center, but it gets lower and lower as it spans to the sides. This is not realistic, and will get scored accordingly. The use of a second set of tweeters can help skew the listener's perception of height and width when carefully installed. Often times, a second tweet will be placed in the A pillar or top of the door (behind the side mirrors in what's called the "sail panel"), but only proper tuning of this tweeter will get the desired results. Let's say you do some kick panels and you end up with mediocre stage height at the sides. You can use some A pillar-mounted tweeters in a cross- firing scheme to help, but you must keep the kick tweets playing as the MAIN tweets, and since the A pillars are closer to you, you may have to manipulate the polarity to get proper phase between the two sets of tweets.
TO ensure the closer A pillar tweet is not too loud (since it is closer to you) and drawing the sound focus towards itself, you can add passive resistance to it to reduce it's relative volume level (attenuation). 25Watt ceramic resistors are available in various ohm ratings, which can be wires in SERIES on the positive speaker wire running these tweets. It is a good idea to have several resistors on hand to try different levels of attenuation to get the right volume. For example, in the S10, the A pillar tweets are the MAIN ones, and the sail panel tweets are there for the needed ambience to help with stage width. We have an 8ohm resistor in-line to each door tweet, making the door tweets a 12-ohm load (4 ohm nominal speaker impedance + 8ohm resistance = 12 ohms). This was the best configuration that added just enough sonic info to get a better sense of width WITHOUT smearing the imaging toward the sides, like it did with lesser levels of attenuation. Furthermore, we needed to slightly delay the output of the nearer door tweeter in relation to the main A pillar tweeter. We did this by reversing the polarity on the door tweets, netting a small time delay. But again, this type of tuning MUST be done in "real time" while listening, and done AFTER the main speakers are installed in the best possible location.
I have found that with kick panels, it is best to use a second tweet with an IDENTICAL crossover setting to the one used on the main tweets. Many guys will argue that you must crossover the added tweets much higher (like 8-10KHz of so), but this simply is not the case. I like to simply connect a second passive network in parallel with the main passive network to run the additional tweets, and place the resistor(s) inline BEFORE the second crossover (if you put a resistor AFTER a passive network, you will alter the crossover frequency DRAMATICALLY!!!). In HLCD (horn) systems, a higher crossover point on the additional tweets is needed (typically 10-16KHz and up) as any freq.ÃƒÂs lower than that will cause massive phase problems, poor spatial resolution, and multiple point-sourcing. By design, HLCDs are meant to be the ONLY driver in the system producing upper midrange and high frequencies. Added tweets are done to assist mainly in the width department, and again, real-time adjustments are a must when adding them. But we'll get into horns later.
In an all-active system with dedicated amp channels, additional tweeters can be tuned properly by simply adjusting the output levels for the added tweet amp. And depending on the processor used, the arrival time of the tweets can be manipulated digitally (using time-alignment). Furthermore you can adjust the crossover point and slope digitally, and if it is so equipped, you can EQ the tweeters separately from the rest of the system. All-active systems take time to master, but can make proper system setup easier that making manual adjustments in a passive set-up. Passive tuning offers durable settings that WILL NOT change over time, either by accident or inadvertent adjustments, making a passive system a "set it and forget it" type.
Many guys live and die by passive tuning and crossover's, others swear by an all-active or digital layout. Great results can be obtained either way. It is totally up to YOU!