Ohmic Audio

12.7 Complete DSP Setup Walkthrough — Real-World Example

The System

Vehicle: 2019 Honda Civic sedan
Speakers:
- Tweeters: Scan-Speak D2606/9200 (1" dome, OEM locations in sail panels) - Midbass: Focal 165 KRX2 (6.5", front doors) - Subwoofer: Single JL Audio 12W3v3 (sealed, 1.25 ft³, trunk)

Amplification: - Front: 4-channel amp (tweeters + midbass) - Subwoofer: Mono amp

DSP: miniDSP C-DSP 8×12 DL (chosen for Dirac Live capability)

Goals:
1. Flat frequency response ±3 dB from 40Hz-16kHz at driver seat 2. Centered imaging for vocals 3. Natural tonal balance for jazz and classical music

Day 1: Physical Installation and Wiring

DSP mounted: Under driver seat, secured with Velcro (accessible for service)

Input wiring: 2-channel RCA from head unit preamp outputs (front L/R) → DSP inputs 1-2

Output wiring: - DSP Output 1 → Left tweeter (amp channel 1) - DSP Output 2 → Right tweeter (amp channel 2) - DSP Output 3 → Left midbass (amp channel 3) - DSP Output 4 → Right midbass (amp channel 4) - DSP Output 5 (mono summed in software) → Subwoofer amp

Power: DSP powered from switched 12V accessory circuit (turns on with ignition), ground to chassis near mounting location.

Remote turn-on: Head unit → DSP remote in → DSP remote out → both amplifiers (daisy chain)

Day 2: Initial Configuration in Software

Step 1: Input setup

miniDSP software, input configuration: - Input 1 (Left): Gain 0 dB, no EQ - Input 2 (Right): Gain 0 dB, no EQ - Test tone at 0 dBFS, observe input meters peak at -1 dBFS (headroom maintained)

Step 2: Output routing

Matrix mixer configured: - Output 1 (L tweeter): 100% Input 1 - Output 2 (R tweeter): 100% Input 2 - Output 3 (L midbass): 100% Input 1 - Output 4 (R midbass): 100% Input 2 - Output 5 (Subwoofer): 50% Input 1 + 50% Input 2 (mono sum)

Step 3: Initial crossover points

Based on driver specs: - Tweeter HPF: 3000 Hz, LR4 (24 dB/oct) - Midbass: - HPF: 80 Hz, LR4 - LPF: 3000 Hz, LR4 - Subwoofer LPF: 80 Hz, LR4 - Subsonic: 28 Hz, Butterworth 24 dB/oct (driver Fs = 35 Hz, sealed box)

Step 4: Initial levels

All outputs set to 0 dB for baseline. Will adjust after measurement.

Step 5: No EQ yet, no time alignment yet

Save configuration as "BaselineNoTANoEQ_v1.xml"

Day 3: Acoustic Measurements

Equipment: - UMIK-1 calibrated measurement microphone ($75) - Laptop with REW (Room EQ Wizard) software - Test tone generator (REW built-in) - Microphone stand positioned at driver's head position

Measurement 1: Full system, no EQ, no time alignment

Result: - Bass peak +8 dB at 60 Hz (cabin gain) - Dip -6 dB at 90 Hz (cancellation between sub and midbass) - Midrange relatively flat 200 Hz-2 kHz - Peak +5 dB at 3.2 kHz (crossover overlap) - Tweeter rolloff -4 dB at 16 kHz (normal)

Measurement 2: Impulse response

Tweeter arrives first (closest to listener), midbass 1.2 ms later, subwoofer 3.8 ms later.

Measurement 3: Individual drivers

Measured each driver in isolation: - Tweeter alone: smooth from 3 kHz up - Midbass alone: shows resonance at 250 Hz (door panel) - Subwoofer alone: smooth from 35 Hz to 100 Hz with cabin gain lift

Analysis:
Primary issues identified: 1. Bass bloat at 60 Hz (cabin mode) 2. Crossover region 3 kHz needs level adjustment (tweeter 3 dB hot) 3. Time alignment required (3.8 ms spread is audible) 4. Door panel resonance at 250 Hz

Day 4: Time Alignment Implementation

Distance measurements (tape measure from each driver to driver's right ear):

Delay calculation (using subwoofer as reference — furthest driver, 0 ms delay):

Delay formula: Distance difference (inches) / 13,500 inches per second = delay in seconds

Applied in DSP: - Output 1 (R tweeter): +3.70 ms delay - Output 2 (L tweeter): +2.44 ms delay - Output 3 (R midbass): +3.26 ms delay - Output 4 (L midbass): +1.93 ms delay - Output 5 (Sub): 0 ms

Verification measurement:

Impulse response now shows all drivers arriving within 0.3 ms — excellent.

Frequency response: 90 Hz dip reduced from -6 dB to -2 dB (less cancellation due to improved phase alignment).

Day 5: Equalization

Approach: Parametric EQ, cuts only (no boosts except controlled bass rolloff correction)

EQ 1 — Cabin mode at 60 Hz (subwoofer output): - Center: 60 Hz - Gain: -6 dB - Q: 3.0 (narrow — target only the peak)

EQ 2 — Door resonance 250 Hz (midbass outputs): - Center: 250 Hz - Gain: -4 dB - Q: 6.0 (very narrow notch)

EQ 3 — Crossover peak 3.2 kHz (tweeter outputs): Instead of EQ, reduced tweeter output level by -3 dB. Simpler and cleaner.

EQ 4 — Gentle bass shelf to match target curve: Applied low-shelf filter on subwoofer output: - Frequency: 100 Hz - Gain: +2 dB - Q: 0.7 (broad, musical)

Result after EQ:

Measured response now ±3 dB from 40 Hz to 16 kHz. Significantly improved.

Day 6: Fine-Tuning by Ear

Test tracks used: - Diana Krall — "The Look of Love" (vocal imaging) - Steely Dan — "Aja" (bass definition and dynamics) - Patricia Barber — "Code Cool" (high-frequency extension)

Adjustments made:

  1. Subwoofer level: Reduced by 1 dB (measured flat but sounded slightly heavy on Steely Dan)

  2. Tweeter polarity: Tested both polarities. Normal polarity sounded more natural — confirmed correct.

  3. Fine time alignment tweaks:

    • Left tweeter delay adjusted from 2.44 ms to 2.50 ms (+0.06 ms)
    • Resulted in better centered vocal image
    • Distance calculations are starting points — ears have final say

Final listening impressions:

Day 7: Dirac Live Calibration (Optional Advanced Step)

Dirac Live is automated room correction available on the C-DSP platform. After manual tuning established baseline, ran Dirac for comparison.

Process: 1. Measured 9 positions (driver, passenger, rear seats) 2. Dirac analyzed all positions and generated correction filters 3. Applied Dirac corrections on top of existing manual work

Result: Dirac smoothed some remaining response irregularities but reduced dynamic punch slightly. Preferred manual tune for this vehicle/music taste. Dirac saved as alternate preset for reference playback.

Final Configuration Summary

Crossovers: - Subwoofer: 28 Hz subsonic (BW24), 80 Hz LPF (LR4) - Midbass: 80 Hz HPF (LR4), 3000 Hz LPF (LR4) - Tweeter: 3000 Hz HPF (LR4)

Time Alignment: - R Tweeter: 3.70 ms - L Tweeter: 2.50 ms (fine-tuned) - R Midbass: 3.26 ms - L Midbass: 1.93 ms - Subwoofer: 0 ms

Equalization: - Subwoofer: -6dB @ 60 Hz Q=3.0, +2dB shelf @ 100 Hz Q=0.7 - Midbass: -4dB @ 250 Hz Q=6.0 - Tweeter: Output level -3 dB

Measured Performance: - Frequency response: 40 Hz–16 kHz ±3 dB - Impulse response: All drivers aligned within 0.3 ms - Imaging: Centered within 2° of dashboard center

Time invested: 7 days part-time (~15 hours total)
Cost: DSP $650, measurement mic $75, time $0 (DIY)

This is a real, achievable, excellent result for a dedicated installer following systematic process.

END OF CHAPTER 12

Chapter 12 Statistics: - Word count: ~7,600 words - Page equivalent: ~15 pages - Sections: 6 of 6 complete ✅ - Three-tier structure throughout ✅ - Visual placeholders: 12 referenced