OE Interface

This is the temporary landing page for OEinterface.com.


As I write this, the car audio industry is changing dramatically. One of the biggest changes has been in the dash of modern cars – new in-dash head units are rarely easily upgraded, and some are practically impossible to replace without significant consequences (broken climate controls, diagnostic trouble codes, etc.)

On the plus side, most OE head units work well and have good potential for sound quality. The shortcomings of OE audio systems usually stem from poor amplifiers and speakers. The amps are underpowered and inexpensively designed, and the speakers are made to be as low-cost as possible. Often, the system may include audio processing which, while claimed as a sonic benefit, is intended to partially mask the mechanical shortcomings of the poor-performing speakers.

So adding better amplifiers and speakers is the path to better sound. Unfortunately, OE integration of aftermarket amplifiers is a field full of misinformation, confusion, surprises, and poor results. Some manufacturers of OE integration hardware have made this situation worse, claiming capabilities for their gear which their gear doesn’t have, and preaching that every OE integration install “needs” their product (without explaining why).

Many professional installers have little experience with OE integration of amplifiers. Sometimes the results are good, sometimes they aren’t, but few installers know the ins and outs of OE integration. There is little usable training from the manufacturers – just lots of fuzzy explanations and sales pitches.

The objective of this page is to help car audio installers – consumers and pros – to get better sound by adding amplifiers (and new speakers) to their OE source units.

The goals:

- Explain the underlying principles so that anyone – professionals and amateurs alike – can get the best results.

- Establish a standard framework for discussing OE integration (the “3×3 Matrix”)

- Share empirical OE integration experience we have gained with real-world vehicles

NOTE: This site isn’t trying to explain electronics theory beyond the basics.

The Three Questions

The Three Questions are about the usable analog signal in your OE audio system. Each question has three possible answers. You need to know the answer to all three questions in order to properly plan an OE interface install. If you don’t know the answers, your installation might turn out fine – or it might turn out terribly – but if you know the answers to The Three Questions, you can make certain that the system design is correct and the gear used is what you need. In order to answer some of the questions, I highly recommend you get a wiring diagram for the OE audio system.

The first question can be answered by referring to your diagram, and you may also need a voltmeter – but the last two can’t be answered practically without an audio analyzer. This capability is lower-cost than ever before, and you can read more about creating your own test setup here. But if your car has already been properly analyzed and described in the forum section, you won’t need your own test setup!

The Three Questions:

What is the signal type? (Answers: Common-ground, Balanced, Speaker-level.)

What is the signal response? (Answers: Flat, Fixed EQ, Dynamic EQ)

What is the signal range? (Answers: Full-range, Usable Crossover Point, Un-usable Crossover Point.)

What is the signal type?

The three types are speaker-level, balanced, and common-ground.

The aftermarket amplifier needs an analog audio signal to amplify. That audio signal needs to be as unmolested and unprocessed as possible, and it usually needs to change voltage when you change the volume setting. The only forms of signal you will work with are the three forms above.

You want to get the usable analog signal from the OE source unit to the OE amp whenever possible.

If your OE system lacks an external amplifier – if all the speakers are connected directly to the OE head unit – then you have a speaker-level signal to work with. I recommend checking the OE wiring diagram. If you don’t already have it, I recommend getting it…

If your system has an external amplifier – and most do in the newest cars – then you could end up using any of the three signal types. Different OE audio system source units send speaker-level signals, balanced signals, and common-grounded preamp signals into OE external amplifiers. (I will explain why you want to use the OE preamp signal to the amp, whenever possible, in a bit).

In addition, some OE systems send digital-databus signals down an optical cable to the OE amplifier. At this writing there are no commercially-available optical converters available, and you will use the speaker-level signals coming out of the amplifier (the most common method), or else you have to rip your OE amp apart and get your hands dirty (much rarer but possible, I haven’t done it). The most common optical system is called MOST – Multimedia Optical Signal Transport. (Some cars used a system called D2B before MOST was introduced). MOST and D2B optical cars should be considered as “speaker-level signals” for the purposes of the 3×3 Matrix. One analogy which may help you is that you are using the OE amp as an external digital-to-analog converter – which happens to have speaker-level outputs.

Some send fixed-voltage analog signals with a databus-carried volume/preamp control (Toy/JBL and some recent GM LAN cars are examples of this). These are usually not usable. The volume, fader, and subwoofer output controls change the output of the amp, but not of the HU, so usually you need to use the ‘speaker-level signal” coming out of the amp in order for those to operate. For the purposes of answering The Three Questions, fixed-output cars should fall under the “speaker-level signal” category – since you’re using the amplifier outputs. The analogy is that you are using the OE amp as an external preamp.

Some people treat balanced signals and speaker-level signals in the exact same way. While there are many practical similarities, the biggest difference is voltage. Many speaker-level inputs and adapters have an attenuation component which is not optimal for low-voltage balanced signals. The signal loses a great deal of voltage, and the noise floor becomes a problem. For this reason, we differentiate between low-voltage, preamp-level balanced and higher-voltage speaker-level signals. Some very high-performance high-power OEM amplifiers paired with higher-impedance woofers have enough voltage on the outputs to overload many input devices. Don’t go by how many watts the interface device can handle, go by how many volts it can handle. Many don’t say on the packaging, unfortunately.

Aftermarket head units use standard RCA connectors to supply outboard amplifiers with a “pre-amplified” signal. An RCA connector has a center pin and an outer ring. The center pin carries AC voltage, and the (-) pin serves as the reference signal ground. The left and right rings are both grounded, so they are “commoned” or common-grounded. This is also called “single-ended” for reasons we will not go into here (Google it).

While the RCA connectors carrying a common-ground preamp output signal have been used for years, it is in many ways an inferior method to balanced. Few OE systems use it, and no OE systems use RCA connectors. If your OE head unit does use common-ground outputs, it can make OE integration very simple, and save you quite a bit of money to boot!

This can often be diagnosed by looking at the wiring diagram. If there is an outboard amplifier with four (+) signals and one (-) signal arriving from the OE head unit, then you have a common-ground system. Aftermarket RCA outputs use a common-ground system, but not all OE systems work instantly with aftermarket amps. Some common-ground systems don’t have AC on the (+) leg – they have significant DC content (called “DC offset”). If your system has DC offset voltage, you may need to find a way to get rid of it if your amplifier doesn’t have bypass capacitors on the inputs designed to do this for you. This can be tested for with a voltmeter (see the test procedures). More on this in another article to follow.

Recap: Is your usable signal speaker-level, balanced, or common-ground?

What is the signal response?

As a general rule, aftermarket head units have the optimal signal response – flat as a pancake. Each note is reproduced at the same amplitude – none are emphasized, none are attenuated.
While some OE systems provide such a response, many don’t – and some do in one point in the signal chain, but not in another. You need to test!

The only practical way to analyze the signal response is to use an audio analyzer and a test track. Methods of doing this are addressed in the test gear section.

Full-range signal, equalized
(Note: Some very good MLS speaker software may not work for testing an OE head unit – mostly it depends on if the OE HU has an aux in jack for the test tone generator. More on this in the article on audio testing.)

If the output of the HU has a slight “happy face” curve at low volumes, but at medium and high volumes the signal is flat as a pancake, you have an “automatic loudness” feature. What’s that?

Fletcher, Munson, and “loudness”

Fletcher and Munson were scientists who determined that humans don’t hear all notes equally well at all volumes. Their test results showed that at low volumes, our ears are most sensitive to midrange notes (such as the human voice), and less sensitive to notes in the bass and treble regions. At higher volumes, this effect gradually decreases. This phenomenon has been described in a graph called the “Fletcher-Munson Loudness Curve”.

Fletcher Munson curve

Many audio systems, for both home and car, have a button labeled “Loudness”. Almost no one knows what it is supposed to do, and it is usually mis-used. I consider “Loudness” to be one of the worst feature names in the consumer electronics world (along with clipping, crossover, valet mode, and shock sensor). On a few better systems, it is called “Loudness Compensation”, which is only slightly less bad a term.

Pressing the Loudness button boosts the bass, and often the treble, by predefined amounts. If you are listening at low volumes, and you want the music to sound more accurate, you should use the Loudness feature. If you are listening at high volumes, you should not use the Loudness feature. Not only do you not need it at higher volume levels, but the boost in the bass may actually force your amplifier into clipping (essentially running it out of power).

Why are we talking about Loudness? Almost no OE head units have a Loudness button, so why talk about it?

Because many OE head units have an automatic, volume-dependent Fletcher-Munson-curve-based loudness equalization feature built into them. See the second graph attached. That is the auto-loudness curve built into a preamplifier IC made for car stereos by a major Japanese consumer electronics manufacturer. As you can see, at low volume settings, that output is not flat at all. As you can also see, at higher volumes it is flat.

Sony preamp IC

The reason for going into this discussion of auto-loudness is to make sure you understand how it works, and to make certain you are not fooled by it.

Auto-loudness mimics how our ears work. Fletcher and Munson proved that at low volumes, our ears need that kind of equalization. It is not a bad thing, and in my professional opinion, auto-loudness requires no de-processing or reversal of any kind. Please don’t waste your money buying OE integration hardware to reverse a problem you don’t have to reverse.

When interpreting the RTA test results, you cannot tell if auto-loudness is present or not unless you test at several volume settings. If the signal becomes flatter as the volume goes up, and becomes completely flat around halfway up, your signal should be classified “Flat” and no de-equalization is needed. Some systems will become less flat as the volume setting is raised – these systems are a different breed (more below).

Deciding if the non-linearities of the signal qualify as “auto-loudness” or as dynamic equalization is an important determination – and it requires some judgement on our part. The only way to make this determination is to play the pink noise track, look a the output using a real-time analyzer, and compare it to the Fletcher-Munson curve.

With some OE systems, the signal is never flat – it is always equalized. Now you need to determine if the signal has a fixed equalization curve, or a dynamic equalization curve. Again, auto-loudness plays a role. If you get auto-loudness effects as your raise and lower the volume, but the other peaks and valleys on the RTA remain fixed at the various volume levels, your signal should be classified as “Fixed EQ”.

If you have a fixed EQ curve – no changes with volume other than auto-loudness – you do not have to install any volume knob that comes with your OE interface, and you can still use your OE head unit’s volume controls – even if you are using an OE interface device with auto-adjustment.

But on the other hand, if your signal changes more dramatically as volume increases – if, for instance, the low bass signal starts to disappear as the volume setting climbs – you should classify your system as having “Volume-Dependant Dynamic EQ”. You will either want to install a new volume control (such as the volume control that comes with many OE processors), while leaving your OE system’s volume adjustment at one specific setting, or you may want to try a processor which is designed to correct such a problem, if you can find one which actually works.

Recap: Is your signal flat, fixed-EQ, or volume-dependent dynamic EQ?

What is the signal range?

When we say “range”, we mean the range between the highest and lowest notes contained in the signal. The audible range is generally considered to be 20 Hertz to 20,000 Hertz. If the signal contains all these notes, we consider it to be “full range”. If some notes are left out, but are contained in another signal dedicated for different speakers (a subwoofer output, for instance) we refer to the system as being “crossover filtered”. The range question really requires a real-time analyzer to answer.

If the amplifier has multiple channels with internal crossover filters preventing any channel from playing “full range”, it becomes critical to determine what the filter “crossover frequency” is, and compare that to your system design.

If the OE system has a crossover point of 80 Hertz between two channels – everything above 80 Hertz goes to the door speakers, and everything below 80 Hertz goes to the OE subs, for example – you need to compare that to your system design. If your system design plans to have a crossover point around 80 Hertz, with a separate set of amp channels playing the notes above 80 Hertz and another set of amp channels playing the notes below 80 hertz, then you can use the OE crossover and leave the individual channels, and the individual signals, separate. This is simpler and easier and less costly, and avoids some potential problems around the overlapping frequencies.

Here is an active three-way system:

LowpassBandpass Highpass

But if the OE crossover point is at 300 Hertz, and your aftermarket system design doesn’t include a crossover at that point (most don’t), then you need to somehow “sum” or add these signals back together into a full-range signal (which you may subsequently crossover filter at a new frequency). There are a few different ways to do this on the market today.

Again, you don’t want to use any hardware you don’t have to, so analyzing the signal is key.

NOTE: If you have an OE crossover point of 100 Hz, and you want one of 80, and you have a processor with a very flexible EQ, you can cheat. EQ the notes just past the crossover point – boost on one side and on one channel, and cut on the other side – and “shift” it from 100 down to 80. No need to sum channels back together for a small shift. Now, from 100 to 300 Hz – that’s too great a shift, and you need to sum the channels back together in most systems.

Recap: Is your signal full-range, or does it have a usable crossover point, or does it have an un-usable crossover point?

MOST or other unusable head-unit/amplifier connections

A usable signal is an analog signal which changes with volume and is as unmodified and unprocessed as possible. There are some connections between the OE head unit and the OE amplifier which are not easily usable for standard OE integration:

- Digital connections, including fiber-optic connections. This includes many late-model Mercedes, BMW, and Audi. With these cars, you must leave the OE amplifier in place, and use the speaker-level outputs. These systems usually qualify as speaker, crossover filtered, and fixed EQ. (Note – while many MOST interfaces are rumored, as of this writing, NONE are commercially available in the US).

- Fixed-level analog signals (where the volume control affects a preamp inside the OE amplifier). This includes many Toyota JBL, GM, Chrysler Boston Acoustics, and BMW DSP systems. The Chrysler and BMW systems can be “reset” into having varying volume voltage on the signal. The Toyota and GM as of this writing can’t, but there are other options (please check the forum section for that vehicle, thanks). These can only be identified in the field with a meter or audio analyzer.


The Three Questions are:

What is the signal type?

- Common-ground
- Balanced
- Speaker-level

This will tell you what you need to do to make the signal usable by an amplifier or processor.

What is the signal response?
- Flat
- Fixed EQ
- Dynamic EQ.

This will tell you if you need to de-EQ the signal, and if you need to use an external volume control or not.

What is the signal range?
- Full-range
- Usable Crossover Point
- Un-usable Crossover Point.

This tells you if you need to worry about summing channels together, or if you can avoid it. Avoiding it is simpler and better-sounding as a rule.