Corrective EQ

1. Overview

Corrective EQ is a final correction stage for imported frequency-response curves. It can be used for headphone correction, room correction, or other target-curve workflows where you already have a correction file and want the application to render it into the audio path.

This manual is written for the current desktop standalone Corrective EQ product first, then calls out the small ARIA ONE-specific differences where they matter.

The core Corrective EQ processing is the same in both contexts. The main practical differences are:

• Standalone Corrective EQ opens directly to the dedicated Corrective EQ workspace.
• Standalone Corrective EQ is enabled by default because it is the main processor in the product.
• ARIA ONE embeds Corrective EQ inside the larger advanced workspace and may also refer to the same module as PostDSPEQ or Post-DSP EQ in presets, tooltips, and internal status text.
• ARIA ONE Simple View can label the page HEADPHONE EQ, while the page itself uses CORRECTIVE EQ. Both names refer to the same correction page in the current build.

At a high level, Corrective EQ works like this:

1. Import a correction file.
2. Confirm that the correction curve appears in the preview.
3. Choose how the curve is rendered: FIR or SOS.
4. Adjust phase, size, rate, or spacing options as needed.
5. Use GAIN and MIX to set output level and correction amount.
6. Save a Corrective EQ preset if you want to recall the imported correction later.

Corrective EQ does not take a hearing test and does not replace the hearing-profile system. Hearing profiles belong to the Hearing EQ / Hearing Correction workflow. Corrective EQ is for imported correction curves.

---

2. Standalone Setup

Corrective EQ runs as a plugin and as a standalone application.

In a plugin host, the host supplies audio and there is nothing to configure here. Skip to Section 3.

In the standalone application, you supply the audio. Corrective EQ sits between your source and your headphones: audio leaves the application you are playing, passes through a virtual audio device, enters Corrective EQ, and returns to your headphones.

Setup has two parts. Route audio into the application, then import a correction file. Both are required. Corrective EQ has no factory presets and does not generate a correction of its own, so with no file loaded it passes audio through unchanged no matter how the routing is set.

Before You Start

Install a virtual audio device. A virtual audio device is a software audio path that appears as an output on one side and an input on the other, so one application can send audio into it and another can read that audio back out. The operating system does not otherwise allow one application to read another application's output.

Any stereo virtual audio device works. It does not need to be ours.

Platform	Options
macOS	BlackHole — free, from https://existential.audio/blackhole/ ARIA-Audio-VAD — included with the ARIA Studio installer Loopback by Rogue Amoeba — commercial, advanced routing
Windows	VB-Audio Hi-Fi Cable — free, from https://vb-audio.com/Cable/ VB-Cable — free, basic ASIO4ALL — universal ASIO driver

You also need a correction file. Section 6 covers the supported formats and where correction files come from.

Setup Steps

1. Install a virtual audio device and restart the computer if the installer asks you to.
2. Set the output of the application you are playing to the virtual audio device. Set it either at the system level, which routes everything, or inside that one application, which routes only that application.
3. Expect the audio to go silent at this point. The source is now feeding the virtual device and nothing is reading from it yet.
4. Launch Corrective EQ.
5. Open Audio Settings with the gear button at the top right of the window.
6. Set Input to the virtual audio device.
7. Set Output to your headphones, or to the interface they are connected to.
8. Confirm that Active input channels and Active output channels both have two channels selected.
9. Set Sample rate to the rate the virtual audio device is running at. See Sample Rates below.
10. Leave Audio buffer size at its default. Change it only if you have a reason to.
11. Confirm that Mute Audio Input is not checked.
12. Play audio and watch the input meter beside Input.

Do not reverse Input and Output. The virtual audio device is the input. Your headphones are the output. Reversing them produces silence with no error.

Confirm Audio Is Arriving

The input meter sits beside the Input selector. Play audio and watch it.

• The meter moves. Audio is reaching Corrective EQ. Routing is correct. Any remaining problem is inside the application.
• The meter does not move. Audio never arrived. The fault is in steps 2, 6, or 8, and no correction setting will fix it.

Check the meter before changing anything else. It tells you which half of the chain to work on.

To test the output on its own, press Test beside Output. Corrective EQ sends a tone directly to the selected output device. If you hear the tone but the input meter stays still, the output is correct and the problem is upstream.

Sample Rates

Every device in the chain carries its own sample rate setting: the virtual audio device, Corrective EQ, and your headphone output. They must match.

When they do not match, the chain either resamples badly or fails. The symptoms are clicks, crackles, stuttering, wrong pitch, wrong speed, or silence.

Pick one rate and set it everywhere you can. 44.1 kHz and 48 kHz are both safe. Most of these settings live in the operating system rather than in Corrective EQ.

• macOS — set device formats in Audio MIDI Setup.
• Windows — set formats per device in Sound settings.

Each device is set separately, and a device can carry a different rate for input than for output. Check both directions on the virtual audio device.

If your source changes sample rate between tracks, lock the source application to a fixed output rate so the rest of the chain stays stable.

First Correction

Routing alone produces no audible change. Import a file.

1. Press IMPORT FILE and select a correction file for your headphone model or your room.
2. Confirm that the curve appears in CORRECTION CURVE PREVIEW.
3. Confirm that IR status reads Ready. The correction is not in the audio path until the render finishes.
4. Confirm that ENABLED is on.
5. Leave MIX at 100 for full correction.
6. Use GAIN to match the corrected level to the uncorrected level before you judge anything.
7. Toggle ENABLED off and on and listen to tonal balance, not to width or space.

A correction curve usually changes overall loudness. Whichever side is louder tends to sound better regardless of whether it is better, so match levels first or the comparison is not worth making.

Section 6 covers supported file formats. Section 10 covers what to adjust next.

Setup Inside ARIA ONE

When Corrective EQ runs inside ARIA ONE, ARIA ONE owns the audio device configuration. Open MENU, then Audio Settings. The routing principle is the same: input is the virtual audio device, output is your headphones, and sample rates must match. See the ARIA ONE user manual for its audio routing sections.

Inside ARIA ONE the same module may be labelled PostDSPEQ, Post-DSP EQ, or HEADPHONE EQ depending on where you are in the interface. Import behaves the same way in every case.

---

3. Accessing Corrective EQ

Standalone Corrective EQ

In the standalone desktop Corrective EQ product, Corrective EQ is the main workspace. Open the application and you are already working directly with the dedicated correction processor.

Corrective EQ inside ARIA ONE

In ARIA ONE desktop advanced view:

• Use the Corrective EQ navigation card to open the Corrective EQ workspace.
• Use the card's power toggle to enable or disable Corrective EQ DSP.
• The navigation card may show the short label POST EQ in compact areas.

In ARIA ONE Simple View:

• Open the Sound page for Corrective EQ / Headphone EQ.
• Use the available Simple View controls for enable state, MIX, GAIN, import, clear, and curve preview.

Naming inside ARIA ONE

ARIA ONE currently uses several names for the same module depending on context:

Name	Where It Appears
Corrective EQ	Main product name and advanced workspace title
POST EQ	Compact navigation label
PostDSPEQ	Preset category, import dialogs, and some internal status messages
Post-DSP EQ	Home/status text and technical descriptions
Headphone EQ	Some Simple View navigation labels

These names refer to the same imported-correction stage.

---

4. Signal Chain Placement

Standalone Corrective EQ

In the standalone product, Corrective EQ is the processor you are using. The imported correction is applied as the product's main correction stage before output.

Corrective EQ inside ARIA ONE

Inside ARIA ONE, Corrective EQ is a final post-DSP correction stage for imported frequency-response curves. It runs after the main sound-shaping modules and before output metering.

Use Corrective EQ when you want to apply an external correction or target curve after the rest of the ARIA ONE sound state has been created.

Corrective EQ is separate from:

• Hearing EQ / Hearing Correction - hearing-test-based personalization
• FullScale - the hearing test used to create hearing profiles
• Crossfeed - headphone imaging and stereo blend processing
• Limiter - dynamics, safety, and loudness processing

---

5. Workspace Layout

The Corrective EQ workspace is organized around file import, correction rendering, output blend, and visual confirmation.

Header and power

The top of the panel identifies the module as CORRECTIVE EQ.

The ENABLED control turns the correction stage on or off.

• On - the loaded correction is active.
• Off - Corrective EQ is bypassed.

If no correction curve is loaded, enabling the processor does not add a correction curve by itself.

Correction state

The CORRECTION STATE area reports the loaded correction and processing status.

Common status text includes:

• No correction curve loaded
• Correction curve
• IR status
• Preparing
• Ready

Ready means the correction has been loaded and the rendered processing data is available.

File controls

The file controls are:

• IMPORT FILE
• CLEAR

Use IMPORT FILE to select a correction file. Use CLEAR to remove the imported correction and return the correction stage to passthrough.

Filter controls

The filter controls determine how the imported curve is rendered:

• FILTER TYPE
• PHASE TYPE
• FIR SIZE
• SOS RATE
• FILTER SPACING

Some controls are only relevant for one filter type. For example, PHASE TYPE and FIR SIZE apply to FIR rendering, while SOS RATE and FILTER SPACING apply to SOS rendering.

Output controls

The output controls are:

• GAIN
• MIX

GAIN trims the Corrective EQ output level after correction.

MIX blends between the uncorrected signal and the corrected signal.

Correction curve preview

The CORRECTION CURVE PREVIEW graph shows the loaded correction curve. If no file is loaded, the graph prompts you to import a CSV response or EQ file before previewing the curve.

The embedded graph area can also show target and filter views where available.

---

6. Importing Correction Files

Supported file types

The file picker accepts correction files using these extensions:

• .csv
• .txt
• .eq
• .apo

The extension alone does not guarantee that the file can be loaded. The file content must match one of the supported correction formats below.

CSV frequency response

CSV files should contain frequency and dB values.

Supported row formats:

frequency_hz,magnitude_db
frequency_hz,left_db,right_db

Examples:

20,-1.5
100,-2.0
1000,0.0
10000,1.2

or stereo correction:

20,-1.5,-1.2
100,-2.0,-2.1
1000,0.0,0.0
10000,1.2,1.0

Notes:

• Frequency values must be greater than 0 Hz.
• The current parser accepts correction points up to 24000 Hz.
• CSV validation requires at least two valid frequency/dB points.
• CSV files can contain header rows.
• Duplicate frequencies are replaced by the later valid row.
• If the curve does not cover the full 20 Hz to 20 kHz range, endpoint values are extended.
• CSV files with more than 65,536 valid points are rejected.

AutoEQ / Equalizer APO parametric text

Corrective EQ can load supported AutoEQ / Equalizer APO parametric text exports.

The expected format is the common Equalizer APO style:

Preamp: -6.0 dB
Filter 1: ON PK Fc 105 Hz Gain -2.4 dB Q 0.70
Filter 2: ON LS Fc 120 Hz Gain 1.5 dB Q 0.70
Filter 3: ON HS Fc 8000 Hz Gain -1.0 dB Q 0.70

The current parser uses Preamp: when present and renders supported Filter lines into a correction curve.

Supported filter families include:

• PK / PEQ
• LS / low shelf variants
• HS / high shelf variants

Disabled filters are ignored.

Equalizer APO GraphicEQ

Corrective EQ can also load GraphicEQ: lines.

Example:

GraphicEQ: 20 -1.5; 100 -2.0; 1000 0.0; 10000 1.2

Each pair is interpreted as frequency and gain in dB.

Text frequency/dB rows

For non-CSV text files, Corrective EQ also tries to read simple numeric frequency/dB rows if no supported AutoEQ or GraphicEQ: data is found.

This means a .txt, .eq, or .apo file may still load if it contains valid rows such as:

20 -1.5
100 -2.0
1000 0.0
10000 1.2

or:

20 -1.5 -1.2
100 -2.0 -2.1
1000 0.0 0.0
10000 1.2 1.0

Import feedback

After import, Corrective EQ reports whether the correction loaded successfully.

Possible import messages include:

• Correction file imported successfully.
• The selected correction file could not be found.
• Please select a .csv, .txt, .eq, or .apo correction file.
• No valid CSV correction points were found.
• At least two valid frequency/dB points are required.
• The correction file has too many points.
• Could not load a correction curve from this file.

If a CSV file loads with warnings, the warning dialog can report skipped headers, ignored rows, duplicate frequencies, or range extension.

---

7. Control Reference

ENABLED

Turns Corrective EQ processing on or off.

In standalone Corrective EQ, this is on by default for a fresh state.

In ARIA ONE, Corrective EQ is optional and may be off until enabled from the navigation card or panel.

FILTER TYPE

Chooses the rendering topology for the imported correction curve.

• FIR - renders the correction as FIR convolution.
• SOS - renders the correction as cascaded SOS biquad processing.

Switching filter type can change latency, CPU use, and how closely the correction curve is matched.

PHASE TYPE

Chooses the FIR phase mode.

• MP - minimum phase
• LIN - linear phase

MP is the default FIR behavior. LIN preserves phase more directly but adds latency.

This control applies to FIR rendering.

FIR SIZE

Sets the FIR length used to render the imported correction curve.

Available values:

• 4096
• 8192
• 16384
• 32768
• 65536

Larger FIR sizes can improve low-frequency resolution. They also increase latency and CPU cost.

This control applies to FIR rendering.

SOS RATE

Sets the SOS oversampling family used by SOS / IIR filter processing.

Available values:

• OFF - no SOS oversampling
• 176.4/192 kHz 4X OS - 4X oversampling family
• 352/384 kHz 8X OS - 8X oversampling family

Higher SOS rates can improve IIR filter performance, especially in higher-frequency correction behavior. They also increase CPU cost.

This control applies to SOS rendering.

FILTER SPACING

Chooses the SOS band layout.

• LIN - standard linear spacing
• ADAPT - adaptive spacing

This control applies to SOS rendering.

GAIN

Trims the Corrective EQ output level after correction.

• Range: -12.0 dB to +12.0 dB
• Default: 0.0 dB

Use GAIN to restore level after a correction curve changes loudness, or to reduce level if the imported correction adds too much gain.

MIX

Blends between uncorrected and corrected signal.

• Range: 0% to 100%
• Default: 100%

At 0%, the correction is not heard. At 100%, the imported correction is fully applied.

IMPORT FILE

Opens the file picker for correction files.

The picker accepts .csv, .txt, .eq, and .apo files. The file must still contain supported correction data.

CLEAR

Removes the loaded correction curve and returns the correction stage to passthrough.

---

8. Filter Rendering Notes

FIR

FIR rendering is useful when you want the imported curve to be rendered as a convolution filter.

Use FIR when:

• matching the correction curve closely is more important than lowest latency
• you want access to MP / LIN phase behavior
• you want control over FIR length with FIR SIZE

SOS

SOS rendering uses cascaded biquad / IIR filters.

Use SOS when:

• you want lower-latency correction
• CPU cost is a concern
• the correction can be represented well by the supported filter rendering

SOS RATE selects the oversampling family for SOS processing:

• OFF keeps SOS processing at the current session rate.
• 176.4/192 kHz 4X OS uses the 4X oversampling family.
• 352/384 kHz 8X OS uses the 8X oversampling family.

Higher oversampling families can improve IIR behavior, especially near the top of the audible range, at higher CPU cost.

Changing topology

When you change FILTER TYPE, PHASE TYPE, FIR SIZE, SOS RATE, or FILTER SPACING, Corrective EQ may need to prepare the correction again.

During this time, IR status can show Preparing. When preparation is complete, it changes to Ready.

Latency & delay compensation

Corrective EQ reports its processing delay to the host so the DAW's plugin delay compensation (PDC) keeps this track aligned with the rest of your session. This is automatic — nothing needs to be nudged or time-shifted by hand.

How much delay is reported depends on the rendering choices:

• FIR + LIN (linear phase) — a fixed delay of half the FIR SIZE. For example, a 4096-tap linear-phase filter reports about 2048 samples; a larger FIR SIZE reports proportionally more.
• FIR + MP (minimum phase) — effectively zero added delay.
• SOS — only the small sample-rate-conversion delay of the selected SOS RATE oversampling family (on the order of a few samples). OFF adds none.

Corrective EQ declares exactly the delay it applies and republishes the value automatically whenever a change affects it (FILTER TYPE, PHASE TYPE, FIR SIZE, or SOS RATE). Because the declared figure matches the applied delay, a null or phase comparison against a dry copy stays aligned. Inside ARIA ONE, this contribution appears as the peq term in the OPTIMIZATION panel's LATENCY readout.

Bypassing — with the plugin's own bypass or the host's — does not drop the reported latency to zero. The declared delay is held and the bypassed (dry) signal is delayed to match, so the track stays time‑aligned and an A/B or null comparison never jumps when you toggle bypass.

---

9. Presets

Corrective EQ has its own preset category for the imported correction state.

In ARIA ONE, this category is labeled PostDSPEQ Preset.

A Corrective EQ / PostDSPEQ preset is intended to store the correction state separately from hearing-profile presets, limiter presets, and Crossfeed presets.

Use a Corrective EQ preset when you want to recall:

• the imported correction curve
• filter rendering choices
• GAIN
• MIX
• enabled state

In ARIA ONE, a Meta preset can include the larger sound state along with Corrective EQ state. Use a dedicated Corrective EQ / PostDSPEQ preset when you only want to save or recall the correction stage.

---

10. Suggested Workflow

Quick Start

1. Open Corrective EQ.
2. Press IMPORT FILE.
3. Select a supported correction file.
4. Confirm that the correction curve appears in CORRECTION CURVE PREVIEW.
5. Confirm that IR status reaches Ready.
6. Start with FILTER TYPE set to FIR.
7. Leave PHASE TYPE at MP unless you specifically need linear phase.
8. Leave MIX at 100% for full correction, or lower it if you want a partial correction.
9. Adjust GAIN if the correction changes output level.
10. Save a Corrective EQ / PostDSPEQ preset if you want to recall the correction later.

Importing an AutoEQ file

1. Download or export the AutoEQ / Equalizer APO file for your headphone model.
2. Press IMPORT FILE.
3. Select the parametric EQ text file.
4. Confirm that the curve loads.
5. Adjust GAIN if the file includes strong boosts or cuts.
6. Save a preset if this will be a regular correction setup.

Importing a CSV response

1. Prepare a file with frequency and dB columns.
2. Use either mono rows (frequency_hz,magnitude_db) or stereo rows (frequency_hz,left_db,right_db).
3. Press IMPORT FILE.
4. Select the CSV file.
5. Review any import warnings.
6. Confirm that the correction curve preview matches the expected shape.

Import And Rendering Problems

If the file does not load:

• Confirm that the file uses .csv, .txt, .eq, or .apo.
• Confirm that the file contains supported correction data, not only comments or unsupported filter types.
• For CSV files, confirm that at least two valid frequency/dB points are present.
• For AutoEQ / Equalizer APO files, confirm that filters are enabled with ON.
• Check whether the file contains more than 65,536 valid CSV points.

If the curve loads but does not sound as expected:

• Confirm that ENABLED is on.
• Confirm that MIX is above 0%.
• Check GAIN for large level changes.
• Try switching between FIR and SOS.
• If using FIR, try a different FIR SIZE.
• If using SOS, try a different SOS RATE or FILTER SPACING.

For audio that never reaches the application at all, see Section 11.

---

11. Troubleshooting

Separate the two failure types before you start. If the input meter is not moving, audio is not reaching Corrective EQ and the problem is routing. If the input meter is moving but nothing sounds different, audio is arriving and the problem is the correction.

No Sound

Symptom	Cause	Fix
No sound, input meter still	Audio is not reaching Corrective EQ	Re-check the source output and the Input selection. Setup steps 2 and 6.
No sound, input meter moving	Audio arrives but leaves somewhere you are not listening	Re-check Output and press Test. Setup step 7.
No sound, both meters still	Input is muted, or channels are deselected	Uncheck Mute Audio Input. Confirm Active input channels and Active output channels. Setup steps 8 and 11.
Corrective EQ does not list the virtual audio device	The device installed after Corrective EQ opened, or did not install	Restart Corrective EQ. Restart the computer if it is still missing.

Audio Artifacts

Symptom	Cause	Fix
Clicking, crackling, stuttering	Sample rates do not match, or the buffer is too small	Match sample rates first. Raise Audio buffer size only if the problem remains.
Wrong pitch or speed	Sample rate mismatch	Match every device in the chain.
The same music twice, slightly offset	The source reaches your headphones through Corrective EQ and directly at the same time	Disable the direct path.
Distortion after importing a curve	The correction boosts into clipping	Lower GAIN, or reduce MIX.
Thin or quiet output	Input level is low, or the source applies loudness normalization	Feed Corrective EQ a steady signal. Turn off normalization at the source.

Audio Plays But Nothing Is Corrected

1. Confirm that a file is loaded. With no correction curve, Corrective EQ passes audio through unchanged.
2. Confirm that the curve appears in CORRECTION CURVE PREVIEW.
3. Confirm that IR status reads Ready rather than Preparing.
4. Confirm that ENABLED is on.
5. Confirm that MIX is above 0.
6. Match levels with GAIN, then toggle ENABLED and compare again.

If all six are true and the difference is still inaudible, the curve itself may be small. Check the preview: a curve within roughly a decibel across the spectrum is doing very little by design.

A correction file for the wrong headphone model loads cleanly and makes the sound worse rather than better. If the result is clearly wrong rather than merely subtle, confirm the file matches your equipment.

---