Car Audio Explained, Part 2: Aiming and Where Each Speaker Goes (Andy Wehmeyer's Lessons)

4 days ago

34 min read

This is Part 2 of 19 in the series Andy Wehmeyer on Car Audio, Explained for Beginners. Start here · Download the full guide as a PDF

Unofficial guide. Andy Wehmeyer's points are paraphrased from his public forum posts, blog, tuning guides, videos and podcasts, with a link to each source; the plain-English explanations were written for this guide. Not written, reviewed or endorsed by Andy Wehmeyer or Audiofrog.

1.5 Aiming: only where a driver beams

What this part is about: Pointing ("aiming") a speaker at the listener only matters at frequencies where that speaker sends its sound out in a narrow beam. His view is that in a car this is mostly true of the tweeter, so aiming woofers and midranges achieves little.

1.5 #1 · In a 3-way, aim "Just the tweeter."

A 6" below ~300 Hz and a 3" below ~3 kHz are identical from 0 to 90°; for a 2" pillar mid crossed at 2 kHz, aiming "HAS NO EFFECT WHATSOEVER" on reflections. Turn tweeters in a little; a gauge cluster in a mid's path changes little and EQ fixes it. [RARE] Sources: Audiofrog blog Jun 2021 · DIYMA 2016-04-22 · DIYMA 2021-10-29

In plain English: A 3-way system splits the music among three kinds of speaker: a woofer or midbass for low notes, a midrange for the middle, and a tweeter for the highs. Any speaker spreads low notes evenly in every direction and only starts to "beam" (send its sound mostly straight ahead) as the pitch rises. He says a 6" (about 150 mm) speaker below about 300 Hz, and a 3" (about 75 mm) speaker below about 3 kHz (3,000 vibrations per second), sound the same whether you are straight in front (0°) or fully to the side (90°). So turning them changes nothing. For a 2" pillar midrange crossed at 2 kHz (it hands the notes above 2 kHz to the tweeter), he says aiming has no effect at all on the reflections it makes off the glass and dash. Only the tweeter plays high enough for aiming to count. So in practice: turn the tweeters in a little. If a gauge cluster (the instrument panel) sits in a midrange's path, he says it changes little and EQ (adjusting the level of each frequency band) fixes it.

1.5 #2 · Door midbass aiming doesn't matter at all.

From 80 to 200 Hz a 6" radiates everywhere ("not one bit — zip — zilch — nada"); what matters is arrival time and the sub-to-midbass blend. [RARE] Sources: DIYMA 2010-02-14

In plain English: Midbass is the upper-bass range, roughly the punch of a kick drum, usually played by a 6" (about 150 mm) speaker low in each door. He says that from 80 to 200 Hz a 6" speaker sends sound out equally in every direction, like ripples spreading from a pebble dropped in a bathtub, so which way it faces makes no difference at all. What matters instead is arrival time (when its sound reaches your ears compared with the other speakers) and how smoothly it blends with the subwoofer. So in practice: don't spend effort angling door woofers; spend it on timing and on the sub-to-midbass blend.

1.5 #3 · Aiming off-axis is just slow, risky "acoustic EQ", and it adds crosstalk.

Pointing a speaker away moves its highs into the reflections, which is why A-pillar and into-the-windshield mounting give a narrower stage than door or sail-panel tweeters. A friend who liked "head between the tweeters" was just closer to them (louder). [RARE] Sources: DIYMA 2011-03-26 · DIYMA 2014-03-30 · DIYMA 2021-10-29

In plain English: "Off-axis" means the speaker points away from you, so you hear it from the side. Because a speaker beams its highs, turning it away lowers the highs you hear directly and sends them toward the glass and panels, so you hear them mostly as reflections. He calls this a slow, risky "acoustic EQ": a tone control you set by physically moving the speaker. It also adds crosstalk, meaning sound meant for one side spreading to the other, for example by bouncing across the windshield. That, he says, is why speakers in the A-pillars (the posts either side of the windshield) or aimed into the windshield give a narrower stage (the band of sound you seem to hear in front of you) than tweeters in the doors or sail panels (the small triangles where the side mirrors mount). A friend who liked having his "head between the tweeters" was, in his view, simply closer to them, so they were louder.

1.5 #4 · Tweeter rule: keep both listeners inside a ~±30–35° window and stop fussing.

0° versus 15° makes essentially no difference; turn sail-panel tweeters slightly toward the rear of the car. Laser aiming is "nonsense" unless you "strap your head to the headrest". Sources: DIYMA 2011-04-15 · DIYMA 2012-04-29 · DIYMA 2015-12-20

In plain English: Picture the tweeter's beam as a cone spreading out from it. He says that if both the driver and the passenger sit within about 30 to 35 degrees either side of where the tweeter points, that is good enough, and you can stop fussing. Pointing it straight at you (0°) versus a little off (15°) makes essentially no difference. His specific tip: turn sail-panel tweeters slightly toward the rear of the car. Aiming tweeters with a laser at an exact head position is "nonsense", he says, unless you "strap your head to the headrest", since a real listener's head moves.

1.5 #5 · A-pillar tweeters aimed straight back or at the listener just trade one error for another.

Straight back matches level between seats but the center wanders; aimed at the listener matches response but adds near-side bias. With time alignment available, move them to the sail panels. Sources: DIYMA 2011-11-05

In plain English: A-pillar tweeters sit on the posts either side of the windshield. He describes two common ways to point them, each with its own flaw. Aimed straight back toward the rear of the car, both seats hear similar loudness, but the center image (the sound that should seem to come from the middle, like a lead singer) wanders. Aimed at the listener, the tone matches better, but you get near-side bias: the speaker closer to you dominates and pulls the sound toward your side. If you have time alignment (a processor that delays the nearer speakers so all the sound arrives together), his advice is to move the tweeters to the sail panels instead.

1.5 #6 · Crossfiring A-pillar tweeters across the dash is "a band-aid for a 2-seat car".

The near tweeter loses its top octave off-axis, roughly matching the far one that is ~6 dB down by distance: better for both seats than aiming at each, worse than a one-seat tune. [RARE] [Conflict: a 2015 Harman/JBL/Infinity tip sheet (authorship unconfirmed) recommends it, and the Bose Fiat 500 he praised for two-seat imaging crossfires.] Sources: DIYMA 2012-04-07 · Harman tip sheet 2015 · DIYMA 2015-12-07

In plain English: Crossfiring means pointing each A-pillar tweeter across the dash toward the opposite seat. Take the driver: the near tweeter now points away from him, so he hears it off-axis and it loses its top octave (the highest band of treble; an octave is a doubling of pitch). The far tweeter points at him but is about 6 dB quieter simply because it is farther away (6 dB is a clearly audible drop). The two losses roughly balance, so both sides sound more alike. He calls this "a band-aid for a 2-seat car": a patch that works better for both seats than aiming each tweeter at its own listener, but worse than tuning for one seat. The [Conflict] tag notes the other side: a 2015 tip sheet from Harman/JBL/Infinity (whose author is unconfirmed) recommends crossfiring, and the Bose system in the Fiat 500, which he praised for imaging in both seats, crossfires.

1.5 #7 · He co-invented an aimable flush swivel tweeter at Infinity.

The premise: highs are directional and car locations often poor, so a flush tweeter can be aimed toward the listener (up to ~27° in the figures), consistent with his later view that aiming is a tweeter-only tool. [Era: Infinity, filed 1997] Sources: US 5,859,917

In plain English: While at Infinity (a speaker brand owned by Harman), he was a co-inventor on US patent 5,859,917, filed in 1997. It covers a tweeter mounted flush (level with the panel surface, not in a pod) that can swivel in its mount so it can be pointed toward the listener, by up to about 27° in the patent's drawings. The idea was that high notes are directional and the spots available in cars are often poor ones. The point notes that this fits his later view that aiming is a tool for tweeters only. This point relates to a specific product and period of his career (Infinity, 1997).

1.6 Where each driver goes: doors, sail panels, A-pillars, kicks, dash, corners

What this part is about: Where to physically mount each speaker in a car: doors, sail panels, A-pillars, kick panels, the dash and the corners. His recurring theme is that nearby glass and panels reflect sound and shape what you hear, so every spot is a trade-off between width, height, tone and effort.

1.6 #1 · His default layout: a 3-way in the doors, small mid high, tweeter in the sail panel.

6" midbass low in the door, a 2.5–4" mid high ("in front of the door handle"), tweeter in the sail panel. Every Audiofrog demo car (through the 2023 Volvo C30) and his MS-8 development cars used it: the glass becomes the baffle, the windshield is farther away, the stage is widest, and "No surprises." [RARE] Sources: DIYMA 2013-03-11 · CMA Networks 2023 @17:01 · Absolute Electronix 2023 @51:00

In plain English: This is his go-to setup. A 3-way means three speakers per side: a 6" (about 150 mm) midbass low in the door, a small 2.5–4" (about 65–100 mm) midrange high in the door, "in front of the door handle", and the tweeter in the sail panel (the small triangle by the side mirror). He used it in every Audiofrog demo car (Audiofrog is his speaker company), through the 2023 Volvo C30, and in the cars used to develop the JBL MS-8 processor. His reasons: the door glass becomes part of the baffle (the surface a speaker is mounted in, which guides its sound forward); the windshield is farther away; the stage comes out widest; and "No surprises." Many people believe the A-pillars are the best place for mids and tweeters; this point goes against that.

1.6 #2 · "The key is in getting a small midrange high in the door."

A 2.5" high in the door can outperform a 4" in the wrong spot; a GB25 in a door panel needs no enclosure, ideally ahead of the door pull. Choose door versus elsewhere by whether you value width or height more, since width follows placement. [RARE] Sources: DIYMA 2014-03-04 · DIYMA 2017-10-31 · OFCAP Ep. 38 @35:03

In plain English: He says "The key is in getting a small midrange high in the door." A 2.5" (about 65 mm) midrange placed high in the door can beat a 4" (about 100 mm) one in the wrong spot. The GB25, Audiofrog's 2.5" midrange, needs no enclosure (box) when mounted in a door panel, ideally ahead of the door pull (the handle you grab to close the door). Choosing between the door and other spots, he says, comes down to whether you care more about stage width or stage height, because width follows where the speakers physically are.

1.6 #3 · Midbass always sits in a corner in a car, so design and EQ for it.

Low door and kick are corners: the baffle adds ~6 dB and the corner ~6 dB more below ~500 Hz, which is why the GB60 is voiced flat and low-Q rather than "full". A woofer in the door's top front corner makes left and right responses much more alike; with a sub you don't need an 8". [RARE] Sources: Audiofrog blog Jun 2021 · DIYMA 2017-08-17 · DIYMA 2021-12-03

In plain English: A corner is where two or three surfaces meet. The low part of a door and the kick panel (the side panel in the footwell) are both corners. Nearby surfaces reinforce a speaker's lower notes: he says the baffle adds about 6 dB, and the corner about 6 dB more, below roughly 500 Hz. (6 dB is roughly double the sound pressure.) Because the car already adds this bass, Audiofrog's GB60 midbass is "voiced" (tuned by design) flat and low-Q, meaning smooth rather than bumped up to sound "full". He adds that a woofer in the door's top front corner makes the left and right sides measure much more alike, and that if you have a subwoofer you don't need an 8" (about 200 mm) midbass.

1.6 #4 · Cross door midbass high, 400 Hz–1 kHz.

He always does and has "never had" a 400–600 Hz console cancellation; he endorsed 80–800 Hz lower-door midbass with an upper-door mid, and a 9" door (70–600 Hz) plus 3.5" kick 3-way. Mini Cooper fix: add a 3" and cross the factory 6" at 1 kHz. Sources: DIYMA 2014-03-30 · DIYMA 2014-04-02 · DIYMA 2010-01-03

In plain English: A crossover is the filter that decides which speaker plays which notes; the crossover point is where one speaker hands over to the next. He crosses the door midbass high, somewhere between 400 Hz and 1 kHz, so the door woofer plays up to that point before the midrange takes over. When a forum member worried about a cancellation dip from the console at 400–600 Hz, he said he always crosses that high and has "never had" that problem. He has endorsed a lower-door midbass playing 80–800 Hz with a midrange high in the door, and a 3-way with a 9" (about 230 mm) door woofer playing 70–600 Hz plus a 3.5" kick-panel speaker. For a Mini Cooper, his fix was to add a 3" (about 75 mm) speaker and cross the factory 6" over at 1 kHz.

1.6 #5 · The sail panel (mirror triangle) is his tweeter location.

It's farther from the windshield, the door and glass act as its baffle, it gives a wider stage than dash or pillars, and it's easy ("Velcro them on and recalibrate"). About 1.5" from the glass puts the reflection dip near 2.2 kHz, below most of the tweeter's range. Tweeters in kicks? Move them to the sails. [RARE] Sources: DIYMA 2011-11-05 · DIYMA 2012-04-08 · DIYMA 2017-10-31

In plain English: The sail panel is the small triangle at the front of the door window where the mirror mounts. He prefers it for tweeters because it is farther from the windshield, the door and glass act as its baffle, it gives a wider stage than the dash or pillars, and it's easy: "Velcro them on and recalibrate" (re-run the processor's setup). Background: when a speaker sits near glass, the bounced sound arrives slightly late and cancels the direct sound at one pitch, leaving a dip. He says about 1.5" (about 38 mm) from the glass puts that dip near 2.2 kHz, below most of what a tweeter plays. If your tweeters are in the kick panels, his advice is to move them to the sails.

1.6 #6 · A-pillars are "a HORRIBLE place for a midrange"; their one benefit is height.

A pod hanging a foot or so from windshield, dash and side glass spreads the image over all three and combs the whole midrange: a 2-ft reflection path digs a hole near 250 Hz (the "mid won't play below 250" complaint). Baffle-less tweeter pods there give ~4 apparent sources per side. [RARE] [Conflict: Audiofrog's own copy calls GB10+GB25+GB60 "the ultimate active system for A-pillars", and the GB25 is designed for pillar pods.] Sources: DIYMA 2013-03-10 · DIYMA 2021-08-26 · GB60 page

In plain English: He calls the A-pillars "a HORRIBLE place for a midrange", with height as their one benefit. A pod hanging about a foot (30 cm) from the windshield, dash and side glass reflects off all three, so the image spreads across all three. Those reflections also cause comb filtering across the whole midrange: when a sound mixes with a slightly delayed copy of itself, a row of dips appears, like the teeth of a comb. He says a reflection path of about 2 feet (60 cm) digs a hole near 250 Hz, which explains the common complaint that a pillar "mid won't play below 250". Tweeter pods with no baffle there create about 4 apparent sound sources per side. The [Conflict] tag: his own company's product text calls the GB10+GB25+GB60 set "the ultimate active system for A-pillars", and the GB25 is designed for pillar pods.

1.6 #7 · A-pillar pods are enclosures: size the box to the driver, not the biggest driver to the pillar.

A big infinite-baffle driver in a tiny pod loses bass and peaks before rolloff, like a sub in a too-small box. The GB25 was designed so "you cannot build an enclosure too small": with a 200 Hz LR4 high-pass, 0.16 L, 0.25 L and huge boxes measure almost alike. [RARE] Sources: DIYMA 2017-10-01 · DIYMA 2017-07-02 · CMA Networks 2021 @17:00

In plain English: A pod on a pillar is really a small box, and a speaker's bass depends on the box it's in. Some speakers are "infinite-baffle" types, made to work with a large open space behind them (such as a door) rather than a small sealed box. Put a big one of these in a tiny pod and, he says, it loses bass and gets a peak just before its low end falls away, like a subwoofer in a box that is too small. So pick a driver that suits the pod, rather than the biggest driver that fits the pillar. Audiofrog designed the GB25 so that "you cannot build an enclosure too small": with a 200 Hz LR4 high-pass (a steep filter that removes the notes below 200 Hz), boxes of 0.16 L, 0.25 L (about one cup) and very large ones measured almost the same.

1.6 #8 · Windshield corner: push the mid tight into it; never hang a pod about a foot away.

Tucked into the dash–windshield–pillar corner, the reflectors are so close that the first null moves above the mid's range, which is why factories use it; its position relative to the glass matters more than mid-to-tweeter spacing. [RARE] [Changed view: 2012 and March 2013 said keep speakers (mids especially) away from that junction; from August 2013, and firmly in 2021, tight in the corner is best and hanging near it is the problem.] Sources: DIYMA 2012-09-06 · DIYMA 2013-03-20 (old view) · DIYMA 2013-08-10 · DIYMA 2021-12-24

In plain English: Background: the closer a reflecting surface is to a speaker, the higher the pitch of the first cancellation dip (the "first null") it causes. He says that if you push a midrange tight into the corner where the dash, windshield and pillar meet, the surfaces are so close that the first null moves above the range the mid plays. That, he says, is why carmakers use the spot. Hanging a pod about a foot (30 cm) away is the bad version. Where the mid sits relative to the glass matters more than how far it is from the tweeter. The [Changed view] tag: in 2012 and March 2013 he advised keeping speakers, mids especially, away from that junction; from August 2013, and firmly in 2021, he said tight in the corner is best and hanging near it is the problem.

1.6 #9 · Dash mounting gives big free boundary gain and turns the windshield into a "tweeter".

Each halving of radiating space adds ~6 dB (dash, then windshield), so a 2.5" dash speaker plays down to 200–250 Hz on little power; cut the hump rather than boost the top. Aimed at a raked windshield, its beamed highs reflect to everyone: the factory "middler" is a "poor man's tweeter" (fails with upright glass, e.g. Jeep). [RARE] [Conflict: see 1.6 #23 on firing mids into the glass.] Sources: Audiofrog blog Jun 2021 · CMA Connected Oct 2021 @32:43 · DIYMA 2022-01-03

In plain English: Boundary gain: a speaker near a surface can only send sound into part of the space around it, so its lower notes get louder. He says each halving of that space adds about 6 dB: once for the dash, again for the windshield. That lets a 2.5" (about 65 mm) dash speaker play down to 200–250 Hz on little power. His advice is to cut the resulting hump with EQ rather than boost everything above it. And because the small speaker beams its highs, aiming it at a sloped ("raked") windshield bounces those highs to everyone. That is why he calls the factory dash "middler" (a small dash midrange) a "poor man's tweeter". It fails with upright glass, such as a Jeep's. The [Conflict] tag points to 1.6 #23, where he criticizes firing midranges into the glass.

1.6 #10 · Dash-top left/right speakers multiply sources.

Add a center and you get "three left speakers, three right speakers and two center channels" (direct plus reflections), which confuses auto time alignment (try reversing the center's polarity). Splitting a 2-way so mid and tweeter sit together on the dash is pointless: you mostly hear the windshield. Sources: DIYMA 2011-05-12 · DIYMA 2012-04-06

In plain English: Speakers on top of the dash bounce off the windshield, so each one is heard more than once: directly and again as reflections. He says that adding a center speaker there gives you "three left speakers, three right speakers and two center channels". All those copies confuse automatic time alignment (a processor routine that measures when each speaker's sound arrives and sets delays); his suggestion is to try reversing the center's polarity (swapping its + and − connections). He also says splitting a 2-way set (a midrange plus a tweeter) so that both sit together on the dash is pointless, because what you mostly hear is the windshield.

1.6 #11 · Dash tweeters: far forward and outboard, or hidden under the factory grille.

Tweeter type matters "FAR less" than position: put it near the dash–windshield–pillar junction, where aiming up or forward barely matters. Behind a factory grille, clear the plastic and hide a small tweeter ("No need for ugly tweeter pods"), avoiding a ~5/8" gap between its base and the glass (that spacing puts a dip near 10 kHz). [RARE] Sources: DIYMA 2013-04-25 · DIYMA 2021-09-05 · DIYMA 2021-09-04

In plain English: If your tweeters go on the dash, he says to put them far forward and toward the outside, near where the dash, windshield and pillar meet. There, whether you aim them up or forward barely matters, and which tweeter you buy matters "FAR less" than where you put it. If the dash has a factory speaker grille, he suggests clearing the plastic behind it and hiding a small tweeter underneath ("No need for ugly tweeter pods"). He warns against the tweeter's base sitting about 5/8" (16 mm) from the glass, because that spacing causes a dip (a missing band of sound) near 10 kHz.

1.6 #12 · Kick panels fix path lengths but are "terrible for frequency response".

They cut the left/right path difference from ~28" to ~6" and give real two-seat midrange imaging, but the footwell makes a big peak and dip around 300–800 Hz, legs block them and the stage narrows. They work only when the vertical angle to your ears is under ~20° (sedans, not trucks or minivans). [RARE] Sources: Time Alignment Part 2 · DIYMA MS-8 FAQ #13 · DIYMA 2016-04-27

In plain English: Kick panels are the side panels down in the footwells, in front of your feet. From either front seat, the far-side speaker is much farther away than the near-side one; he says kick panels cut that difference from about 28" (71 cm) to about 6" (15 cm). That gives real midrange imaging (sounds that seem to sit in the right places) for both front seats. But he calls them "terrible for frequency response": the footwell causes a big peak and dip somewhere around 300–800 Hz, your legs block the sound, and the stage gets narrower. He says they only work when the angle from the speaker up to your ears is under about 20°, which fits sedans but not trucks or minivans.

1.6 #13 · Doors plus a center beat kick panels for most people.

Kicks create about as many problems as they solve; with a center, doors give a wider stage; for a one-seat car with delay, "don't bother with the kicks." With no delay, little EQ and no center, pick kicks or any layout keeping the left/right path difference under ~1 ft. [RARE] Sources: DIYMA MS-8 FAQ #13 · DIYMA MS-8 FAQ p4 #71 · DIYMA 2021-06-29

In plain English: For most people, he says, doors plus a center speaker beat kick panels. Kicks create about as many problems as they solve. With a center channel (a speaker in the middle of the dash), doors give a wider stage. In a car tuned for one seat with delay (time alignment) available, "don't bother with the kicks." The exception: with no delay, little EQ and no center, pick kicks, or any layout that keeps the difference between your distances to the left and right speakers under about 1 ft (30 cm). Two of the sources come from his FAQ for the JBL MS-8 processor, so they date from that product's era.

1.6 #14 · Every location has its own reflection dip, so choose the trade-off instead of chasing a perfect spot.

His own low-door mids have a side-glass dip (~300–500 Hz), the price of width; all-in-the-dash fixes that but loses width and adds a windshield dip; kicks add a midbass dip. "It's always something." [RARE] Sources: DIYMA 2021-11-02

In plain English: Every mounting spot has some nearby surface whose reflection cancels part of the sound and leaves a dip at some frequency. He gives examples. In his own cars, the low-door speakers get a dip from the side window around 300–500 Hz; that is the price of a wide stage. Putting everything in the dash fixes that dip but loses width and adds a dip from the windshield. Kick panels add a dip in the midbass. His summary: "It's always something." So in practice: pick the trade-off you can live with rather than hunting for a perfect location.

1.6 #15 · The late-model BMW layout is close to "optimum".

4" door mid, sail tweeter, 4" dash center and under-seat woofers: a good directivity match, central woofers that don't overload rear passengers, and seats that soak up woofer noise. His MS-8 BMW crossed the under-seat woofers (~80–200 Hz) at 200 Hz, as low as the 4" doors could play loudly; keep an octave of usable bandwidth around any crossover. [RARE] Sources: DIYMA 2013-03-10 · DIYMA MS-8 FAQ #12 · DIYMA 2021-08-04

In plain English: He calls the layout in late-model BMWs close to "optimum": a 4" (about 100 mm) midrange in each door, a tweeter in each sail panel, a 4" center on the dash, and woofers under the front seats. His reasons: the speakers' directivity (how widely each spreads its sound) matches well from one to the next; woofers near the middle of the car don't overload rear passengers; and the seats soak up woofer noise. In the BMW he set up with the MS-8 processor, the under-seat woofers played roughly 80–200 Hz and were crossed at 200 Hz, as low as the 4" door speakers could play loudly. His general rule: keep about an octave (a halving or doubling of frequency) of usable range around any crossover, so the speakers on both sides of it still play well past the hand-over point.

1.6 #16 · Multiple midbass locations even out the seats but wreck imaging above ~100 Hz.

He tried 4, 6 and 8 midbass drivers with delays, levels and EQ: the more even the seats, the worse the image, and it "doesn't really work very well" for the cost. Keep everything below ~100 Hz out of the rears unless it cancels the sub for the back seat (why BMW under-seat woofers and Mercedes "Front Bass" work). [RARE] Sources: DIYMA 2009-11-11 · DIYMA 2021-11-29

In plain English: Some people spread several midbass speakers around the car so every seat gets even bass. He tried 4, 6 and 8 midbass drivers, adjusting delays, levels and EQ for each. The more even the seats became, the worse the imaging got above about 100 Hz, and it "doesn't really work very well", especially for the cost. His rule: keep everything below about 100 Hz out of the rear speakers, unless that sound is used to cancel some of the subwoofer for the back seat, which is often much closer to the sub. That, he says, is why BMW's under-seat woofers and Mercedes' "Front Bass" system work.

1.6 #17 · Behind the rear seat is the worst subwoofer location, though the easiest to design for.

A deep null at the seat can come from where sub and listener sit, not crossover phase; a dip with only one of two subs is likely the cross-car mode (direct and far-side-bounce paths differing by ~7 ft). [RARE] Sources: DIYMA 2021-09-16 · DIYMA 2021-12-02

In plain English: In many cars the subwoofer goes in the trunk, behind the rear seat. He calls that the worst place for it, though the easiest to design for. If you measure a deep null (a very deep dip at one frequency) at your seat, he says it can come from where the sub and the listener sit, not from crossover phase (the sub and midbass being out of step). If the dip appears with only one of two subs, it is likely the cross-car mode. He describes it by its paths: the sound going straight from the sub to you and the sound bouncing off the far side of the car travel distances that differ by about 7 ft (2.1 m).

1.6 #18 · Home driver-spacing rules matter less in cars, with two exceptions.

Cars have no flat baffles and use steep active slopes, and mid and midbass sharing a door's air is "no big deal". But stack the tweeter directly above the mid (aligned at every horizontal angle), and never in front of a moving woofer cone, whose motion causes Doppler "warbling" (coaxials below ~8 kHz). [RARE] Sources: DIYMA 2009-06-19 · DIYMA 2021-09-07 · DIYMA 2013-11-29

In plain English: Home speaker builders follow rules about how close together the drivers must sit on the front panel (the baffle). He says those rules matter less in cars, which have no flat baffles and use steep active crossovers (electronic filters that cut off sharply). So a midrange and a midbass sharing the same air inside a door is "no big deal". There are two exceptions. First, stack the tweeter directly above the midrange, so the two stay lined up from every side-to-side angle. Second, never put a tweeter in front of a moving woofer cone, as coaxial speakers do (the tweeter sits in the middle of the woofer). The moving cone causes a Doppler "warbling" in the tweeter's sound, which he says affects coaxials below about 8 kHz.

1.6 #19 · Map candidate locations with a mic and a speaker before cutting anything.

A sound path works the same in both directions (reciprocity), so put a small speaker at the driver's head and a boundary mic at each candidate spot; it works for sub, midbass and mids, not tweeters. By 2023 this "system design by mapping" was his MasterTech seminar, replacing guessed rules like "avoid the windshield corner". [RARE] Sources: OFCAP Ep. 38 @36:00 · DIYMA 2021-11-07 · Absolute Electronix 2023 @1:00:14

In plain English: Before cutting holes, he suggests measuring. Reciprocity is a basic acoustics fact: sound travels the same path in both directions, so a speaker at A heard by a microphone at B measures the same as a speaker at B heard at A. So put a small speaker where the driver's head would be and a boundary microphone (a flat mic made to lie on a surface) at each spot you are considering. Each measurement shows how a speaker in that spot would reach your head. He says this works for subwoofers, midbass and midranges, but not tweeters. By 2023 this "system design by mapping" was the subject of his MasterTech seminar, replacing guessed rules like "avoid the windshield corner".

1.6 #20 · Use the factory locations when they work; experts chose them for a reason.

OEM engineers (many with PhDs) place speakers deliberately, and rearranging a car to install "the thing we've always installed" is a big mistake. Fabricate only for a placement you need, such as a factory-looking mid at the top of the door; "Every installation is a compromise." [RARE] Sources: CMA Connected Oct 2021 @37:25 · CMA Connected Oct 2021 @41:52 · OFCAP Ep. 38 @32:04

In plain English: OEM means the carmaker's own factory engineering. He says these engineers, many with PhDs, place speakers deliberately, and that rearranging a car to install "the thing we've always installed" is a big mistake. So in practice: use the factory locations when they work, and only build custom mounts for a placement you actually need, such as a factory-looking midrange at the top of the door. His reminder: "Every installation is a compromise."

1.6 #21 · No time alignment available? Choose locations that need less of it, and add a real center.

Then the center image doesn't depend on a perfect acoustic sum; if the owner doesn't care about imaging, drivers can go anywhere. Expect car-specific exceptions: a big instrument-cluster bulge can spoil sail tweeters for the passenger. Sources: DIYMA 2014-06-11 · DIYMA 2011-05-12

In plain English: Time alignment is a processor feature that delays the closer speakers so all the sound reaches you together. If you don't have it, he says to choose locations that need less of it and to add a real center speaker. A real center plays the middle of the music itself, so the center image doesn't depend on the left and right sounds adding up perfectly at your ears. If the owner doesn't care about imaging, the speakers can go anywhere. He also warns to expect car-specific exceptions: a big bulge over the instrument cluster can spoil sail-panel tweeters for the passenger.

1.6 #22 · Judge placements by criteria, not trophies.

He had never seen a rebuilt dashboard shaped by anything measured; "Joe Bastardo won IASCA with speakers in the glove box" is not evidence, and a competition win doesn't create new physics. [RARE] Sources: DIYMA 2009-07-24 · DIYMA 2020-10-07

In plain English: He says to judge a speaker location by clear criteria, not by trophies. He had never seen a rebuilt dashboard shaped by anything that had been measured. A line like "Joe Bastardo won IASCA with speakers in the glove box" is not evidence (IASCA is a car-audio sound-quality competition), and a competition win doesn't create new physics. Many people believe that if a winning competition car used a layout, it must be right; this point disagrees.

1.6 #23 · Firing midranges into the glass: he moved against it.

In 2009 his own install fired into the glass and he EQ'd the hole (vocals sat right). From 2011 he said windshield-firing narrows the stage, and in 2021 that firing mids up at the glass or ceiling makes "the most dense comb filter possible". He keeps exceptions for an up-firing center and factory dash "middlers" (1.6 #9). [Changed view: 2009 → 2011/2021] [Conflict: he praises the windshield-aimed middler (Oct 2021–Jan 2022) a few months after condemning up-firing mids (Sep 2021).] Sources: DIYMA 2009-07-10 · DIYMA 2011-03-26 · DIYMA 2021-09-07

In plain English: This point tracks how his view on aiming midranges at the windshield changed. In 2009 his own car fired its mids into the glass, and he used EQ to fill in the resulting hole (a dip in the sound); vocals sat in the right place. From 2011 he said firing at the windshield narrows the stage, and in 2021 that pointing mids up at the glass or the ceiling makes "the most dense comb filter possible" (the densest pattern of reflection dips). He still makes exceptions for a center speaker firing up at the windshield and for factory dash "middlers" (see 1.6 #9). The [Conflict] tag: in September 2021 he condemned up-firing mids, yet from October 2021 to January 2022 he praised the windshield-aimed middler. Both statements are his.

1.7 Stage width, height and depth

What this part is about: The "soundstage" is the imaginary stage you seem to hear in front of you. This part covers what decides how wide, how high and how deep that stage seems, and which popular tricks he says don't work.

1.7 #1 · Stage width is fixed by where the speakers are: the inverted triangle.

With you at the bottom point, the stage can't be wider than the speakers, and a sound in one speaker comes from that speaker (his football-field test). After delays and EQ, pan the balance fully left, then right: those are your real edges. Sail tweeters widen the triangle; A-pillars and windshield reflections narrow it. [RARE] Sources: DIYMA 2008-05-20 · DIYMA 2017-08-18 · DIYMA 2017-08-21

In plain English: Picture an upside-down triangle: you sit at the bottom point and the two speakers sit at the top corners. He says the stage can't be wider than the speakers. His football-field test shows why: put two speakers in the middle of a football field, where nothing reflects, turn the balance all the way left, and the sound comes from the left speaker. A sound that plays from only one speaker comes from that speaker. So in practice: after setting delays and EQ, pan the balance fully left, then fully right; those two positions are your real stage edges. Sail-panel tweeters widen the triangle; A-pillars and windshield reflections narrow it.

1.7 #2 · No two-channel trick reliably widens the stage; "width" from delay games is comb filtering.

"Play with delays for width" is a myth: the out-of-phase bands just expose speaker locations (1.1 #3). Images can't be generated reliably outside the speaker–listener triangle; real extra width needs side or rear speakers and processing. With kick mids and door tweeters, more tweeter level adds a little. [RARE] Sources: DIYMA 2021-08-14 · DIYMA 2017-08-21 · DIYMA 2017-08-18

In plain English: Many people believe you can widen the stage by adjusting the delays between left and right. He calls "play with delays for width" a myth. Those settings create bands of frequencies where left and right are out of phase (working against each other), which is a form of comb filtering, and he says that just exposes where the speakers really are (see 1.1 #3). In his view, sounds can't be placed reliably outside the triangle formed by the speakers and the listener. Real extra width needs side or rear speakers plus processing. One small exception: with mids in the kick panels and tweeters in the doors, more tweeter level adds a little width.

1.7 #3 · In a room, width comes from the side-wall reflection; in a car, crosstalk takes it away.

At home the first side-wall reflection acts like a phantom speaker outside the wall, making the room seem bigger without moving instruments. In a car, reflections mark the car's boundary: side-glass bounces add a little width, windshield reflections narrow it, and pillar mids just make images "big". [RARE] Sources: DIYMA 2011-11-17 · DIYMA 2017-08-21

In plain English: At home, sound from each speaker bounces off the nearby side wall. He says that first side-wall reflection acts like a phantom speaker beyond the wall, much like a mirror image, so the room seems bigger without the instruments moving. A car works differently: its reflections mark the car's own boundary, and crosstalk (sound meant for one side spreading to the other) takes width away. Bounces off the side windows add a little width, windshield reflections narrow it, and midranges in the pillars just make images "big" (spread out) rather than wider.

1.7 #4 · "Stage outside the car" claims rest on one or two demo tracks.

Typical evidence is a single event on a judging disc; his own room-synthesis DSP put the EMMA disc's walking man 10 ft beyond each side, but not on every track. [RARE] Sources: DIYMA 2011-11-17

In plain English: Some people claim their system puts the stage outside the car. He says such claims usually rest on one or two demo tracks, often a single effect on a competition judging disc. His own room-synthesis processor (electronics that add artificial room reflections) placed the "walking man" on the EMMA test disc (EMMA is a car-audio competition) about 10 ft (3 m) beyond each side of the car, but that didn't happen on every track.

1.7 #5 · Depth is not distance to the stage; it comes from ambience in the recording or added by processing.

Judges confuse the two; sitting closer to the speaker line raises width and "depth" but softens focus. Real depth comes from early reflections recorded or synthesized in the track, rears with processing, or room synthesis; not every recording has it, so don't retune the car to find it. [RARE] Sources: DIYMA 2013-04-27 · OFCAP Ep. 19 @3:23:07 · DIYMA 2022-03-03

In plain English: "Depth" means sounds seeming to sit at different distances, some near and some farther back. He says depth is not the same as how far away the stage seems, and that judges confuse the two. Sitting closer to the line between the speakers makes the stage seem wider and "deeper" but less sharply focused. Real depth comes from ambience: early reflections captured in the recording or synthesized by processing, rear speakers fed by processing, or room synthesis. Not every recording has depth, so in practice: don't retune your car trying to find it.

1.7 #6 · Aiming changes tone far more than width or depth.

His home test: move your chair toward the speaker plane, re-aiming on-axis; a stage farther in front of you is a change of seat, not image depth. Doors can beat on-axis mounting for both. [RARE] Sources: DIYMA 2013-04-27

In plain English: He says pointing speakers changes their tone far more than it changes stage width or depth. His home test: move your chair forward until you are almost in line with the speakers, turning them so they keep pointing at you. In his account the depth survives, which suggests speakers beside you, as in the doors, can still give depth and width. A stage that sits farther in front of you is simply a different seat position, not real image depth. He adds that door mounting can beat on-axis (aimed-at-you) mounting for both width and depth.

1.7 #7 · Stage height comes from tweeter location plus crossover phase, not EQ.

Put tweeters high (dash, door tops, sail panels); then correct phase at each crossover pulls the whole range up to tweeter height, so with a proper crossover height stops being a problem and pillar mids are a quick fix, like subs up front. [RARE] [Changed view: 2016 stressed physical location; 2020–21 adds that crossover phase holds everything at tweeter height.] Sources: DIYMA 2016-04-27 · DIYMA 2020-10-07 · DIYMA 2021-09-07

In plain English: Stage height is how high the music seems to sit, for example at dash level rather than down by your knees. He says it comes from where the tweeters are plus correct crossover phase, not from EQ. Put the tweeters high: on the dash, at the tops of the doors or in the sail panels. Then, if each crossover hand-over is in phase (the two speakers sharing those notes move in step), the whole range is pulled up to tweeter height. With a proper crossover, he says, height stops being a problem, and he describes pillar mids as a quick fix, comparable to putting subwoofers up front. The [Changed view] tag: in 2016 he stressed physical location; in 2020–21 he added that crossover phase holds everything at tweeter height. Many people believe EQ can raise the stage; this point disagrees.

1.7 #8 · For height plus width: mids at the top of the doors, tweeters in the sail panels, a center where the gauge pod was.

An up-firing center reflects off the middle of the windshield, giving height and width at once ("Move the gauges to the a-pillar."). With center steering there's no phantom center to build, so L/R tweeters are free to lift and widen the stage. [RARE] Sources: DIYMA 2013-08-31 · DIYMA MS-8 FAQ p2

In plain English: For a stage that is both high and wide, he suggests midranges at the tops of the doors, tweeters in the sail panels, and a center speaker where the gauge pod (instrument cluster) sits. A center speaker firing up at the windshield reflects off the middle of the glass, giving height and width at once. As he put it, "Move the gauges to the a-pillar." With center steering (a processor that sends the middle of the music to the center speaker), the left and right speakers no longer have to create a phantom center (the illusion of a middle sound made by both sides playing it equally). That frees the left and right tweeters to lift and widen the stage. One source is his MS-8 FAQ, so part of this relates to that processor.

1.7 #9 · "Stage base a third up the pillars, vocals slightly above" is one recording's layout, not a rule.

It holds for a dry studio track with the singer panned center. Height-test tracks drove the Asian fad for pillar pods with 6" drivers in boxes barely bigger than the basket: "a super high-Q midrange that doesn't play any bass". [RARE] Sources: DIYMA 2013-04-29 · DIYMA MS-8 FAQ p5 #95

In plain English: The quoted rule says the base of the stage should sit a third of the way up the A-pillars, with vocals slightly above. He says that is just how one kind of recording is laid out, not a rule: it holds for a dry studio track (little room echo) with the singer panned to the center. He adds that height-test tracks drove a fad in Asia for pillar pods holding 6" (about 150 mm) speakers in boxes barely bigger than the speaker's frame (its basket), giving "a super high-Q midrange that doesn't play any bass" (high-Q meaning a peaky response).

1.8 Tweeterless and widebander systems

What this part is about: Some builders skip tweeters and let one small "full-range" or "widebander" speaker play both the midrange and the highs. He argues firmly for using a tweeter, and explains what goes wrong without one and how his own small midranges were designed.

1.8 #1 · Always use a tweeter; there's no credible case for skipping it in a car.

Tweeters are the smallest, cheapest, easiest drivers to fit. A car is near-field listening with road noise (so crossover insertion loss is irrelevant) and reflectors close enough that you hear the power response; a 3" run from 100 Hz to 20 kHz is inefficient and distorts. In 2021 he still said every car sounds better with tweeters. [RARE] Sources: DIYMA 2014-04-13 · DIYMA 2015-05-07 · DIYMA 2021-12-02

In plain English: He says always use a tweeter; there's no credible case for skipping one in a car. Tweeters are the smallest, cheapest and easiest speakers to fit. In a car you sit close to the speakers (near-field listening) with road noise around you, so he says the small loss a crossover adds (its "insertion loss") is irrelevant. Reflecting surfaces are also so close that you hear the power response, meaning all the sound a speaker sends out in every direction, not just straight ahead. A 3" (about 75 mm) speaker made to play from 100 Hz to 20 kHz is inefficient and distorts. In 2021 he still said every car sounds better with tweeters. Many people believe a single full-range speaker sounds purer than a midrange plus a tweeter; this point disagrees.

1.8 #2 · Tweeterless sounds tiny, precise and lifeless.

Images are small and very precise, but the sense of space carried by high-frequency reflections is lost (HF reflections please most listeners). It may score IASCA points; he prefers a believable space. A sail tweeter "won't screw up" full-range drivers; it improves them. [RARE] Sources: DIYMA 2014-10-30 · DIYMA 2013-05-27 · DIYMA 2013-05-23

In plain English: Without a tweeter, he says, images become small and very precise, but the sense of space is lost, because that space is carried by high-frequency reflections, which please most listeners. Such a system may score points at IASCA (a sound-quality competition), but he prefers a believable space. Adding a tweeter in the sail panel "won't screw up" full-range drivers; he says it improves them.

1.8 #3 · Widebanders were designed for desktops, and car builders use them for the wrong job.

They target a listener 18" away with walls 10 ft off, trading sensitivity for low resonance and stretching treble to save a tweeter; cars then pair them with a midbass, removing their reason to exist. No 3" is truly full-range if off-axis response matters; the no-tweeter pitch is "marketing hooey". [RARE] Sources: DIYMA 2013-09-18 · DIYMA 2015-08-23

In plain English: A widebander is a small speaker stretched to cover most of the audio range. He says they were designed for desktops: a listener about 18" (45 cm) away, with walls about 10 ft (3 m) off. They trade sensitivity (how loud they play per watt) for a low resonance (so they reach lower notes), and stretch their treble to save the cost of a tweeter. Car builders then pair them with a midbass, which removes their reason to exist. No 3" (about 75 mm) speaker is truly full-range if its sound off to the side (off-axis) matters, he says, and the no-tweeter sales pitch is "marketing hooey".

1.8 #4 · If you run a widebander without a tweeter, aim it at the windshield.

Aimed at you, highs arrive direct but the reflections carry few: tiny "clinical" images with no space. Aimed elsewhere, highs live in weak reflections: dull and lifeless. Aimed at raked glass from the dash (as OEMs do), the windshield becomes the HF source, "a tweeter without having to pay for a tweeter". A real tweeter still wins. [RARE] [Changed view: 2012–14 he said a tweeterless widebander must be on-axis; 2018–21 he ranks windshield-aimed best.] Sources: DIYMA 2012-09-06 · DIYMA 2018-03-31 · DIYMA 2021-12-02

In plain English: If you do run a widebander without a tweeter, he says to aim it at the windshield. Aimed at you, its highs arrive directly but the reflections carry few highs, so images are tiny and "clinical", with no sense of space. Aimed elsewhere, the highs exist only in weak reflections, so it sounds dull and lifeless. Mounted on the dash and aimed at a sloped windshield, as carmakers do, the glass becomes the source of the highs: "a tweeter without having to pay for a tweeter". A real tweeter still wins. The [Changed view] tag: in 2012–14 he said a tweeterless widebander must be aimed at the listener (on-axis); in 2018–21 he ranked windshield-aimed best.

1.8 #5 · Full-range pillar pods can't be EQ'd for two seats.

Forward imaging needs similar directivity above at least 1 kHz (better, above 100 Hz). A small full-range pillar pod gives midrange spread from three nearby surfaces but pinpoint highs, so an EQ that works in one seat is "horrible" in the other. [RARE] Sources: DIYMA 2013-09-17

In plain English: Directivity describes how widely a speaker spreads its sound at each frequency. He says that for the stage to seem to sit in front of you, the directivity needs to stay similar across all frequencies above at least 1 kHz, and better above 100 Hz. A small full-range speaker in a pillar pod doesn't manage that: its midrange spreads out and bounces off three nearby surfaces, while its highs come from one pinpoint spot. As a result, an EQ that works in one seat is "horrible" in the other.

1.8 #6 · Audiofrog's small mids are midranges, not widebanders.

The GB25 brief: reach 200 Hz in a 0.004 cu ft box at 100 W RMS, efficiency matched to its midbass, and a top-end peak the low-pass removes so a tweeter takes over. Going tweeterless means EQing that peak out ("please, God no!"). Sources: DIYMA 2015-05-07 · DIYMA 2016-03-03

In plain English: Audiofrog's small speakers, like the 2.5" (about 65 mm) GB25, are midranges meant to work with a tweeter, not widebanders. The GB25's design brief: play down to 200 Hz in a tiny 0.004 cu ft box (about 0.11 litres, less than half a cup) at 100 W RMS (continuous power); match the efficiency (loudness per watt) of its midbass partner; and have a peak at the top of its range that the low-pass filter (which cuts the notes above the crossover) removes, so a tweeter takes over. Running it without a tweeter would mean using EQ to flatten that peak, to which he says "please, God no!".

The full series

Download the complete guide as one PDF (351 pages, every point with its explanation).