Audio preservation is becoming more important in the processing and mixing processes as more recordings are made in less-than-ideal environments. This ensures there are a plethora of directions to boost a poor audio repair signal—but there are still unrivaled chances to muck it up even worse than ever!
I’ve assembled a rundown of the most popular audio restoration blunders. The list is mainly oriented toward music development, although post-production professionals can find some helpful advice as well. If you see yourself in any of these, don’t feel bad: I’ve all been there. What matters is that you understand and pass on. Could this list serve as an example of such a learning opportunity!
1. Improper noise-reduced file marking and maintenance
This is a nasty, potentially serious issue masquerading as a small blip. Be sure you just use an audio-restoration feature on a backup of the audio repair before you commit it to a track. It’s not just a copy in the DAW; it’s a whole new disk. Never use this processing on the initial, and save the new track with a name that specifically identifies the processing you used.
It’s dangerous to make damaging edits to the initial track and you may need to go back to it later: maybe you denoised so many of context and need to start over; maybe the sound design system doesn’t allow for noise suppression at all; maybe some phrases do, but others don’t. Since all scenarios are imaginable, it’s a good idea to plan ahead.
You won’t realize the track is which if the tracks aren’t explicitly numbered, and you’ll waste a lot of time auditioning folders, wondering and second guessing which one is right for the situation.
2. Over de-noising
This, I believe, is the most pressing problem in audio reconstruction, especially when it comes to making speech sound more natural and presentable. Do not de-noise too much, read these terms carefully, and abide by them.
Over-denoising affects many too many podcasts, documentaries, and other pieces of art. I’m not sure if I can mention something so casually. I’ve heard the objects, plain and simple. They’ve been in TV programs, films, podcasts (the biggest offenders), and songs. These objects are difficult to forget until you encounter them. They’re so strange that they’ve gained the nickname “space monkeys” in the industry.
Don’t let space monkeys infect your conversation track!
Here’s how to stop going through it: De-noise until you detect noticeable objects, then back off the reduction slider until the artifacts are no longer audible. Then you should actually go out even more!
You don’t want visible objects in these signals so they show that the processing was done incorrectly. And to the untrained ear, they convey a sense of inferiority—a sense of audio repair in progress. In other terms, those objects render you seem unprofessional. They can also sound critically awful, which can make it difficult for people to enjoy your job.
As a starting point, you should listen to the module’s noise-only output, which will give you a clear sense of what you’re removing. To ensure you’re not inducing objects, check the noise-only quality alongside the process described above.
3. Inappropriate application of background noise in noise control
It’s not enough to reintroduce a noisy vocal (or bass—or, hell, even a noisy dobro, which passed through my workshop recently) until it’s been sufficiently denoised. You’ll have to work the audio repair around it to make it suit the scene.
Maybe we’re discussing a thread of dialogue: If the line is surrounded by a small amount of atmosphere, you’ll need to use a room tone that mixes in with the crowd. You can also get imaginative about this: a well-placed zephyr can be used to obscure the occasional flare-up of noise in an outdoor scene with visible gusts of wind, particularly if the noise couldn’t be minimized without objects in the first place.
4. Processing in the wrong order
In general, separate tonal issues in an area first and rub them, then occasional pops, then crackles, then some leveling issues, and finally de-noising.
A vocal lead that can reach beyond the mix and stand on its own is needed for a track as busy and large as this one. Bright Voice’s additional high-end EQ offers vocal tracks a distinct edge, whereas the room reverb reduces any unnecessary mid-range bite. Increase the microphone input volume and switch the reverb knob all the way up to get a voice tone that brightens up the overall mix for mixes as powerful as the one above.
5. Warm Voice
Spire’s Warm Voice effect is the company’s oldest and most universal vocal effect. It’s a straightforward vocal chain with adjustable long reverb and a warm EQ environment that makes vocals sound direct, welcoming, and competent.
Warm Voice can appeal to a wide range of vocalists, whether they be singers, topliners, or even spoken-word artists.
6. Relying on broad-strokes processing
Multiple passes at various conditions could be more effective at mitigating noise than a single heavy-hitting move. You will fine-tune the procedure each time to reduce objects and achieve a more artistic outcome.
7. Ignorance of problems across specific frequency ranges
Occasionally, a stubborn piece of noise may be seen lurking in the corners of a frequency spectrum. It may be a siren with a frequency range of 500 Hz to 1.5 kHz. It may be a percussive disturbance with a frequency range between 1–7 kHz.
8. Using applications in standalone or AudioSuite modes is not recommended.
Remember that restoration tools may add minor latencies in real-time processing, depending on the DAW. If you’re working on a post-production project where sync is critical, this is a no-no.
9. Working so fast
In the face of a deadline, time is of the essence. Sure, on a loud conversation track, you don’t often have time to go slow. Making the period, though, is extremely beneficial.
Take the Voice De-noise module for example: you might batch it all at once, but you could end up with a stray space monkey. If you do all at once, you might end up having to pay for a slew of issues later. You’ll get a more smooth, artifact-free outcome if you de-noise phrase by phrase, adapting the learn feature to the noise surrounding each phrase and changing the parameters to better fit each phrase.
Final thoughts
When it comes to audio repair, the example of a double-edged sword is a good one to use. It has the ability to eliminate unnecessary sounds with high precision and minimal side effects. If you abuse this app, you might end up with a product that’s even worse than where you started—or perhaps worse: a corrupted original file with no way to go back to the root and start again.
So, when you walk along the happier path of audio-restorative fate, keep in mind what Spiderman said: “With great power comes great responsibility.
