EditPad Aurora ← Back to editpadaurora.com

User guide · Version 1.0

Finishing a vocal.

EditPad Aurora finishes vocal recordings. You give it two things — your vocal, and a reference whose sound you want yours to match. It analyses both, decides how to treat yours, and returns a mixed, mastered vocal ready to drop into your DAW.

What this does

EditPad Aurora finishes vocal recordings.

You give it two things: your vocal, and a reference — a finished vocal whose sound you want yours to match. It analyses both, decides how to treat yours, and returns a mixed, mastered vocal ready to drop into your DAW.

It does not write your song, sing it, or replace your voice. It takes the recording you made and finishes it to the standard of the reference you chose.

Before you start

What your recording needs to be

EditPad Aurora finishes a good recording. It cannot rescue a poor one.

MicrophoneAny decent microphone. A USB condenser is fine. A phone is not.
RoomReasonably controlled. Avoid heavy echo, hard bare rooms, and audible background noise.
EffectsNone. No reverb, delay, autotune, compression or EQ printed into the file. Send it dry.
FormatWAV or AIFF preferred. MP3, FLAC, M4A and OGG also load.
LevelRecorded at a sensible level without clipping.

Dry and clean is what the software needs. Everything else it will add.

Read this one twice

Choosing a reference

This is the single decision that most affects your result.

Use an isolated vocal. A vocal stem, an acapella, or a vocal you have separated from a track. The software analyses the reference to learn what a finished vocal should sound like — so it needs to hear a vocal.

A full song works less well. If the reference contains drums, bass and instruments, the analysis follows the balance of the whole mix rather than the vocal within it. You will get a result, but it will be aiming at the wrong target.

Match the material roughly. A male vocal referenced against a male vocal, a similar genre, a similar style of delivery. You are asking the software to move your voice toward that sound — the closer the starting point, the better the outcome.

AI-generated references work well. If you have produced a version of your song with an AI music tool, that vocal makes an excellent reference — you are matching your real voice to the production you already have.

The three steps

1. Load your vocal

Click Load Source and select your recording. The waveform appears in the Source card. If the file is stereo it is converted to mono — normal for a lead vocal, and the status bar will tell you.

2. Load your reference

Click Load Reference and select the vocal you want to match.

3. Auto Match

Click Auto Match. The software analyses both files and processes yours. This takes a few seconds to a minute or so depending on length. When it finishes, the status bar reports what happened — the loudness reached, the gain applied, and the voice type it detected.

That is the whole workflow. Everything below is refinement.

Listening to the result

Four things can be played, on two cards.

SourceYour original recording, untouched.
RepairedAfter cleanup, before matching. Useful for hearing what was removed.
ReferenceThe track you are matching to.
ResultThe finished vocal.

Click a tab to select it, then the play button. Click or drag anywhere on the waveform to scrub.

Match levels — leave this on when comparing. It evens out the volume difference between what you are comparing, so you hear the tone rather than simply the louder one. Without it, louder always sounds better.

The four main controls

After Auto Match these move to positions the software chose. Adjust any of them and click Auto Match again to hear the change.

PresenceHow far forward the voice sits against the track. Up brings it toward you; down sits it back.
BrightnessAir and top end. Up opens it out; down warms and softens it.
LoudnessThe final level of the vocal. Normally leave it where the software set it, since it is aiming at your chosen loudness target.
ReverbSpace and depth around the voice.

Small movements. These are finishing adjustments, not corrections — if you find yourself moving them a long way, the reference is probably the thing to reconsider.

Delay

Delay is different from everything else in the application. It is not decided by the analysis, it is not applied when you Auto Match, and it changes live while you listen. It is the last thing you add, by ear, once the vocal itself is finished.

How to use it

  1. Get the vocal right first. Auto Match, listen, adjust, listen again.
  2. Select the Result tab and press play.
  3. Bring Mix up slowly while it plays.

You will hear the delay appear in real time. There is no need to re-run Auto Match.

The controls

MixHow much delay you hear. Start low. Delay works best when you notice its absence more than its presence. If you can clearly hear echoes, it is probably too high.
TimeThe spacing between repeats, in milliseconds. Around 100–200 ms gives a tight slap that thickens the voice; 300–500 ms gives distinct echoes that open up the space. Match it roughly to the tempo of your track and let your ear decide.
FeedbackHow many times each repeat comes back. Low gives one or two echoes; high gives a long trail. Above about 60% it becomes an effect in itself rather than a finish.
Hi-CutDarkens the repeats. This is the control that makes delay sound professional. Bright echoes compete with the vocal and clutter the track; darker ones sit behind it and create depth. Lowering this is almost always an improvement.
Lo-CutRemoves low end from the repeats so they do not build up mud underneath the voice. Leave it engaged.
Ping-PongBounces the repeats left and right. Adds width and dimension. On by default, and worth keeping on for most material.
WidthHow wide the delay spreads in stereo.
ModA gentle tape-style movement in the repeats. Keeps the delay from sounding static and mechanical. Subtle by design.

Getting it right

Set Mix higher than you want while you adjust Time, Feedback and Hi-Cut, so you can hear what you are doing. Then bring Mix back down until it almost disappears. That is usually the right amount.

The delay is included when you export. What you hear is what you get.

Exporting

Click Export, choose where to save, and confirm.

You get a 24-bit WAV, dithered, at the same sample rate as your source. The delay is included. The file is time-locked to your original — drop it into your DAW at the same position as your source recording and it will line up exactly.

What to do with it

You now have a finished vocal, not a finished song. Import it into your DAW alongside your instrumental, balance the two, and master the combined track as you normally would.

The target setting

Target determines what loudness the result aims for.

ReferenceMatches the loudness of your reference file. The default, and the right choice when you are matching a specific sound.
PlatformsSpotify, Apple Music, YouTube and Broadcast target that platform's loudness standard instead.

Leave it on Reference unless you have a specific reason. You are exporting a vocal to be mixed with other material, not a finished master, so matching the reference is what keeps the comparison honest.

Simple and Advanced

Simple gives you everything needed for the normal workflow — load, match, listen, adjust, export.

Advanced adds the analysis displays and manual control sections. You do not need them. They are there when you want to see what happened or take over a decision.

The analysis displays

These populate after Auto Match and show what the software actually did.

Spectral balanceThree curves overlaid: your source, the reference, and the result. It shows the tonal shape of each and how the result moved toward the reference. The most useful display in the application if you want to understand a result.
Dynamic rangeThe crest value of the result, in LU. How much difference there is between the loudest and quietest parts.
Stereo imageThe width of the result. A lead vocal is largely centred, so a narrow reading here is normal and correct.

The manual sections

Three sections, one open at a time.

RepairCleanup applied before matching: denoise, de-reverb, de-plosive, soothe, de-nasal, de-ess and air. With Smart enabled the software sets these from what it hears in your recording. Leave it on unless you have a specific problem to address.
MicMicrophone character. The software chooses the one that best bridges the gap between your recording and your reference. You can override it, but its choice is made by measurement rather than preference.
CharacterCompression, reverb and tonal settings applied during matching.

All of these are set automatically. Changing them means overriding a decision made from analysis of your actual files. Worth exploring once you know the software — not the place to start.

The meters

IN and OUT, either side of the display, show live levels while audio plays.

TARGET and RESULT show your loudness target and what the result reached, in LUFS. These hold their values rather than moving — they are measurements of the whole file, not live levels.

Starting again

New / Reset clears everything and returns the application to its opening state. The ✕ on either card clears that file only, leaving the other loaded.

Loading a new file clears the previous result, and Export is disabled until you Auto Match again — this prevents exporting a result that no longer matches the files you have loaded.

If something goes wrong

The result sounds thin, harsh or wrong.

Look at the reference first. A full song rather than an isolated vocal is the most common cause. A reference in a very different style is the second.

The voice type was detected incorrectly.

Set it manually with the Voice control and run Auto Match again.

Nothing happens when I click Auto Match.

Both a source and a reference must be loaded. The status bar will say which is missing.

The export sounds different from what I heard.

Check that you were listening to the Result tab rather than Source or Repaired.

The application will not start.

Confirm your system meets the requirements below. If it was recently installed, the Visual C++ runtime may not have installed correctly — reinstalling usually resolves it.

Something else.

Email support@editpadaurora.com. If the application behaved unexpectedly, attach the log file at %TEMP%\EditPad_debug.log — paste that path into Explorer's address bar to find it.

System requirements

Operating systemWindows 10 or 11, 64-bit
Memory8 GB minimum, 16 GB recommended. Vocals over five minutes benefit from more.
Display1920 × 1080 minimum
Disk50 MB
InternetRequired to activate your licence, and occasionally afterwards to confirm it is still valid. If a check cannot be completed the software continues working normally — you can work offline indefinitely. All audio processing is offline at all times.

Your licence

One licence activates on up to three computers. Deactivate one from within the application to free an activation for another machine.

You own your output. Anything you make with EditPad Aurora is yours — sell it, release it, licence it, with no payment or credit to us.

Support

support@editpadaurora.com — we reply within 48 business hours.

We can help with installation, licensing and faults in the application. We are not able to offer mixing advice or DAW instruction.