Goldhorn DSP Tuning: Complete Setup and Workflow Guide

A Goldhorn DSP gives you the resolution to tune a car system the way a critical listener actually hears it. That resolution is wasted if the setup sequence is wrong. We're the original US importer for Goldhorn, and the tuning approach below is what we use on builds in our own shop and what we walk customers through over the phone when they call with a system that "sounds off."

This guide is written generically across the Goldhorn DSP line. The exact menu paths and band counts vary by model, but the order of operations does not. Gain structure first, crossovers next, then time alignment, then EQ, then measurement-driven iteration. Skip a step and the next one fights you. Do them in order and the system gets coherent fast.

REW (free measurement software) tuning walkthrough. The sequence applies to a Goldhorn build directly.

What Makes Goldhorn DSP Different From Mass-Market Car Audio Processors?

Goldhorn sits in the audiophile DSP category, not the OEM-replacement category. The difference shows up in how you tune it. An OEM-integration DSP is built to clean up a factory signal and run a budget speaker set with minimal fuss. A Goldhorn unit assumes you're building an active system from the source up and want full control over what each driver receives.

That assumption shapes everything: filter resolution, output channel flexibility, time delay step size, and how the parametric bands behave at the edges of their range. If you're coming from a mass-market processor, expect smaller, more deliberate adjustments to produce audible results. A 1 dB cut on a tight-Q parametric band is a real move, not a placeholder.

The practical takeaway: don't tune a Goldhorn unit the same way you'd tune an entry-level processor. Larger boosts and wider EQ smears that masked problems on a coarser DSP will overshoot here and introduce new problems. Work smaller. Re-measure between moves.

What Should You Check Before You Open the Tuning Software?

Open the laptop last. The most expensive tuning mistakes are made because someone started adjusting parametric EQ before fixing a hardware problem the DSP can't compensate for. A clipped signal at the input is still clipped at the output no matter how good the processor is. Audiofrog's tuning material has made this point for years, and it remains the single biggest cause of "the DSP didn't fix anything" complaints (Audiofrog).

Pre-Tune Checklist

• Source unit output: Set head unit volume to about 75 to 80 percent of maximum and play a 0 dBFS test tone. If you hear distortion or see clipping on a meter, that's your real ceiling. Tune everything else below it.
• Amplifier gains: Use the source-unit clip point you just found. Set amp gains so the amp reaches its rated output at that source level. Match every amp the same way. Do not set gains by ear with music playing.
• Polarity: Pulse each driver with a 9V battery and confirm the cone moves outward on positive voltage. A reversed driver creates cancellation the DSP can reduce but never eliminate.
• Ground loop: System on, no music, ear at the speaker. Any hum or whine is a grounding problem. Fix the wiring, not the EQ.
• Speaker health: Sweep 40 Hz to 200 Hz on each driver individually at moderate volume. Listen for buzzing or rubbing. A torn surround or rubbing voice coil wastes every downstream correction.
• Save a flat baseline: Before changing anything in the Goldhorn software, export the current state and label it with the date. You'll want a clean rollback point.

For the broader background on this sequence (cabin acoustics, hardware fundamentals, measurement targets), the complete car audio DSP tuning guide is the pillar reference. Read it once if you haven't.

How Do You Set Crossover Slopes and Types for Each Driver?

Crossovers come before time alignment and EQ because they decide what each driver actually plays. The slope and type also affect phase behavior at the crossover region, which feeds directly into time alignment in the next step. Linkwitz-Riley at 24 dB/octave is the working default for most active car audio builds because the high-pass and low-pass sections sum flat at the crossover point and stay phase-coherent through the overlap region (Audiofrog, bestcaraudio.com).

LR vs. Butterworth at a Glance

Both filter types are built from cascaded first-order stages. Butterworth is each driver at -3 dB at the crossover, with constructive summing that produces a small bump if the drivers are phase-aligned. Linkwitz-Riley is each driver at -6 dB at the crossover, with summing that produces a flat response. For active multi-way builds, LR is what you want by default. Butterworth shows up more often in passive networks and in specific high-pass applications.

Working Starting Points

• Subwoofer to midbass: LR 24 dB/oct, low-pass on the sub between 70 and 90 Hz. The exact value depends on your midbass and door environment.
• Midbass to midrange: LR 24 dB/oct, crossover between 250 and 400 Hz. Lower if the midbass holds up well, higher if the midrange struggles below 400 Hz.
• Midrange to tweeter: LR 24 dB/oct minimum, often LR 48 dB/oct on small-format tweeters. Crossover between 2.5 and 3.5 kHz for typical 1-inch silk and beryllium tweeters. Steep slopes protect the tweeter from excursion damage.

These are starting points, not final answers. Pull the manufacturer minimum frequency and recommended slope from each driver's spec sheet and stay above their stated minimum by at least a half octave. The Goldhorn's output is only useful if every driver downstream is operating in its safe range.

How Do You Set Time Alignment Correctly?

Time alignment exists because the drivers in a car are at wildly different distances from the listener. The left tweeter might be 22 inches from the driver's ear. The right tweeter might be 42 inches. The subwoofer in the trunk might be 80 inches. Without correction, the sound from each driver arrives at the listener at different times and the soundstage collapses toward the closest speaker. A DSP fixes this by delaying every driver to match the arrival time of the farthest one.

The Reference Driver Is the Farthest One

This is where most first-time tuners get it backwards. The farthest driver from the listening position gets zero delay. Every closer driver gets a delay added to it so its sound arrives at the listener at the same instant as the farthest driver. Closer drivers wait. The farthest driver does not.

The speed of sound at typical cabin temperature is roughly 13.5 in/ms. So one inch of extra path length equals about 0.074 ms of required delay (Sengpielaudio). That number is the foundation of every delay value you'll enter into the Goldhorn.

How to Measure the Distances

Use a metal tape measure. Sit in the driver's seat in your normal listening position. Measure from the acoustic center of each driver (the dustcap on a cone driver, the dome center on a tweeter) to the same point on your head, usually the bridge of your nose. Record each distance to the nearest quarter inch. Repeat for every driver in the system, including each subwoofer if you have more than one.

Calculate and Enter the Delays

1. Identify the farthest driver. That's your reference. Its delay is 0 ms.
2. For every other driver, subtract its distance from the farthest driver's distance. Divide the result by 13.5. That's the delay in milliseconds.
3. Enter each delay value in the Goldhorn time delay menu for the matching output channel.
4. Save the profile under a clear name like "TA-tape-measure-2026-05-18."

Tape-measure delays are an excellent starting point, but acoustic centers don't always sit where the visible center of the driver does, especially with horn-loaded tweeters or off-axis mids. A measurement-based refinement pass is the next step. Our REW car audio beginner-to-advanced guide walks through impulse response time alignment in detail. Do the tape measure first, then refine with REW.

Tape-measure time alignment walkthrough. Apply the same logic to your Goldhorn output delay menu.

How Should You Approach EQ in a Goldhorn DSP?

EQ comes last because nothing you do here is meaningful if the previous steps are wrong. With crossovers set, time alignment dialed in, and a flat baseline measurement in REW, you can finally start shaping the response. The order of operations inside the EQ pass also matters: subtractive cuts first, narrow Q for cancellations and peaks, broad Q for tonal shaping.

Cut First, Then Boost

A 6 dB boost adds 4x the power demand on that band and pushes the driver harder. A 6 dB cut adds nothing and gives you more headroom. Address peaks first by cutting. Most cabin response problems are peaks, not dips, because car cabins concentrate low-frequency energy through resonance. Cut the peaks down to the target curve and you'll often find the dips were never as deep as the plot first suggested.

Q Selection

• Narrow Q (3 to 8): Use for resonant peaks. The width of the cut should match the width of the peak on the RTA. A 200 Hz cabin mode with a 30 Hz wide peak gets a Q of about 6.
• Moderate Q (1.5 to 2.5): Use for typical room mode corrections and broader response anomalies. Most cabin gain corrections in the 60 to 120 Hz range fall here.
• Broad Q (0.5 to 1): Use for tonal shaping above 1 kHz. Small broad cuts or boosts shape character without creating an obvious "tuned" sound.

Work in 1 to 2 dB increments. After every adjustment, run a fresh REW sweep. A move that looked right in your head but pulled neighboring frequencies out of line will show up immediately on the new plot. For the structured EQ sequence (which bands to address in what order, how to set target curves, and how to validate corrections), the step-by-step car audio DSP EQ guide covers the workflow we use for every Goldhorn tune in the shop.

How Do You Verify and Iterate a Tune With Measurements?

A tune is only finished when the measurements match the target and the system holds up at real listening volume across a few different recordings. Three REW sweeps averaged at the primary listening position is the baseline measurement, but it's not the whole story. Verify at moderate volume, verify at near-reference volume, and verify with the doors closed and the cabin in its normal state.

What to Look For on the Sweep

1. Overall tilt: Most reference curves slope downward from low to high by 6 to 10 dB across the audible range. A perfectly flat in-car response sounds bright and thin. Aim for a smooth downward slope.
2. Crossover transitions: Look at the response through each crossover region. A clean LR24 transition shows a smooth handoff with no bump and no dip. A 3 dB suckout at the midbass-to-midrange crossover usually means a polarity or time alignment problem on one side.
3. Modal resonances: Sharp narrow peaks under 200 Hz are cabin modes. Cut them with narrow-Q parametric bands.
4. High-frequency air loss: A roll-off above 10 kHz is normal in cars. Don't try to EQ it back to flat. You'll just stress the tweeter.

When to Stop

You stop iterating when each subsequent adjustment produces a smaller and smaller change in the measured response. The first session usually delivers 80 percent of the audible improvement. A second session a few days later, with fresh ears, catches the remaining issues. Three sessions is usually enough for a daily-driver tune. SQ competition tunes can run a dozen sessions, but the law of diminishing returns kicks in fast after the third.

Save the final profile under a clear name. Export it. Email it to yourself. The Goldhorn holds the tune in memory, but having an external backup means a firmware update or accidental factory reset doesn't cost you a weekend of work.

Frequently Asked Questions