FIR And Room-Correction Workflow

Use one sheet per correction goal. Keep the baseline, design assumptions, exported filter, and independent validation together.

1. Problem And Scope

Question this filter must answer:

Date / project:

Operator:

Vehicle / bench setup:

Primary listening position:

Other positions / weights:

Channels / drivers:

Frequency region:

Required decision:

2. Baseline System State

☐ routing and channel identity verified
☐ polarity verified
☐ gain structure verified
☐ protective crossovers active
☐ source, DSP, amplifier, microphone, and ADC remain clean
☐ damaged drivers, rattles, clipping, and severe distortion resolved first
☐ baseline preset and raw measurements saved

Source / playback mode:

DSP / firmware / sample rate:

Amplifier / supply / load:

Measurement method / level:

Existing EQ, delay, dynamics, factory processing, and bass management:

3. Electrical Filter Record

Path	Architecture	Type / target	Settings / coefficients	Peak gain
Reference	bypass / unity
Candidate A	IIR / FIR / delay / all-pass
Candidate B	IIR / FIR / mixed
Final

IIR precision / biquad order:

FIR coefficient format / scaling:

Filter order / taps:

Compiler MIPS / memory:

Block / partition size:

Coefficient update / slew / mute:

4. FIR Design Contract

Sample rate:

Order / coefficient count:

Phase: linear / minimum / mixed / custom:

Time support:

Passband / ripple:

Transition bands:

Stop-band attenuation:

Correction range:

Boost / cut limits:

Regularization / noise threshold:

Expected filter delay:

Total latency budget:

Expected pre-main-event / post-main-event behavior:

5. Measurement And Phase Analysis

Timing reference:

Bulk delay removed / retained:

Mic / calibration / orientation:

Measurement positions:

Impulse windows:

Valid frequency span:

Smoothing / normalization:

Noise-floor limits:

Minimum-phase generation settings / tails / calibration handling:

Minimum-phase regions and evidence:

Excess-phase or excess-group-delay findings:

6. Problem Classification

Feature	Magnitude	Minimum / excess phase	Spatial behavior	Chosen action
1
2
3
4

☐ stable peak considered for cut
☐ pure delay separated from shaped excess phase
☐ deep null checked with individual sources and nearby positions
☐ noise-floor or window-invalid regions excluded
☐ physical, routing, crossover, polarity, or placement fixes considered before inversion

7. Automated Correction Setup

Product / device / license:

Software / firmware version:

Measurement template / position count:

Position geometry / order / weights:

Target curve / target level:

Correction range:

Boost / cut / phase limits:

Expected latency:

Export slot / preset:

Warnings reported by software:

Raw project and prediction file location:

8. Export And Low-Level Check

☐ outputs muted before coefficient or preset load
☐ sample rate and coefficient format match device
☐ channel map and physical output identity confirmed
☐ low-level impulse / sweep matches intended filter
☐ no DC, unexpected gain, noise, pop, clock fault, or reset problem
☐ bypass preset remains available

9. Independent Verification

Trace	Position	Correction state	Magnitude / phase / impulse notes	Pass?
Baseline repeat		bypass
Corrected primary		on
Nearby 1		on
Nearby 2		on
Final repeat		on

Measured total latency:

Channel-to-channel timing:

DSP / DAC output headroom:

Amplifier voltage / clipping:

Distortion / compression:

Excursion / thermal / protection:

Pre-main-event / post-main-event findings:

Listening and bypass comparison:

10. Final Decision

☐ exported filter independently matches the intended response
☐ important nearby positions remain acceptable
☐ latency meets system requirements
☐ headroom, distortion, excursion, thermal, and limiter behavior pass
☐ raw data, coefficients, project, export, and final preset are archived
☐ what the correction does not fix is documented

Final conclusion:

Known limitations:

Next test or revision:

Companion lesson: FIR, IIR, And Room Correction.