The last two posts explored overlooked field recording errors. The first shared general field recording concepts that are often ignored or missed. The second dove into deeper detail with more specific recording mistakes and ways to fix them.
These are slip-ups that I’ve noticed during my years field recording, mastering sound effects, and curating sound libraries for myself and for others. Often they’re not apparent. Many times they’re errors discovered only well after recordings are finished. So, it’s common for pros and beginners alike to miss them.
In this post we’ll look at more mistakes in another category: sound effects mastering.
NOTE: I’m very detailed. This article should take you about 15 minutes to read. Don’t have time now? No problem. Click the button below to email it to yourself to read later.
What is Sound Effects Mastering?
A field recording’s journey doesn’t end once an audio recorder stops rolling. Sounds of cafes, cars, and chimpanzees reach their full potential later, when they are heard by others.
In most cases, field recordings must change before they reach the point where they are best experienced. This altering process is referred to as sound effects mastering, or just “mastering.”
Sound effects mastering typically involves organizing, editing, and processing and sweetening. I’ll write more about what sound effects mastering is and how it differs from the final step of creating music albums in another post. In the meantime, here’s an overview of sound fx mastering from an earlier article.
How Mastering Errors Emerge
It’s frustrating to notice errors made when field recording. It’s annoying to hear a recordist’s mobile phone ring when a lion has finally decided to growl. It’s a shame to learn that an old church door has been open and closed ten times with identical performances.
However, it’s heart wrenching to hear final field recordings that have missed the opportunity to improve during the mastering stage. Perhaps a sparkling brook has microphone bumps midway. Maybe you’re listening to soft mountain air and are jolted by a recordist speaking at the end. These are listening experiences that have the potential to convey feeling and presence but fall short. The reason? Mastering errors introduce problems that distract listeners and ruin immersion. In some cases, the mistakes make field recordings unusable.
I’d imagine all field recordists want their listeners to hear evocative ambiences. I’m certain every one of them wants their fans to use clean clips simply and easily. So why does this happen?
There are four reasons:
- Work is incomplete. Many field recordists believe the work is finished when recording ends. There are many soundscapes that have merit presented as they are, “flaws” and all. Examples include sound walks and “slice of life” and “in-situ” recordings. For all others, field recordings must be changed for listeners to hear them the best way possible.
- Not picking an audience. Many field recordists capture sound focusing on the present: the environment around them, their performances, and how the sound evolves. There is absolutely nothing wrong with that. Field recording is a lot of fun, and simply being outdoors and capturing audio is an incredible thing to do. However, imagining an audience and recording and finalizing a sound just for those people increases a sound recording’s impact.
- Not knowing how sound will be used. Does a field recordist’s fans need the sound normalized to -10 dBFS? Perhaps -20? Maybe a Web shop needs distinct clips, or multiple takes merged into one file? Presenting field recordings one way when fans expect another creates problems. The best mastering techs have experience working in the field where their sound effects are used. They understand the context of how they will be used and presented. That takes years of experience, though. Not every pro knows the best way to prepare a clip for how it will be used.
- New skill set. Sound recording properly takes years of practice. Editing clips? Zapping audio like a plug-in wizard? Writing evocative sound file descriptions? Those are separate skills in their own right. Not every field recordist has the skill – or inclination – to diversify their skill set. The result? Field recordings that are stronger in one way and weaker in another.
Learning the value of sound fx mastering comes from experience. There isn’t a pill you can take to instantly know the best way field recordings will be used. Without that knowledge, learning how to spot mastering errors and knowing the best way to fix them is much more difficult.
Let’s see if a few tips from four categories can help fans use and experience field recordings better:
Imagine sitting in the dawn wilderness as the forest wakes around you. It’s an evocative experience. Scanning through those half hour morning ambiences to use in a feature film? Not so much.
It’s vital to remember that field recordings must be accessible. If they’re difficult to use they won’t be auditioned and will never be selected. Having usable sounds is important for almost every step beyond capturing sound in the field: finding, browsing, selecting, auditioning, copying, uploading, and editing sound clips. Long and large sound files are cumbersome. Instead, focus on making them easy to preview and use. Unless you’re creating an art installation, sound walk, or other specific immersive experience, limit the duration of ambiences to around three minutes. Why three minutes? It’s more than long enough for any film scene. It also has enough length for editors to find variation and cut multiple takes if they need it.
Being mindful of duration doesn’t just apply to atmospheres, either. Avoid gathering specific effects into excessively long files. Half an hour of bullet shell drops is more of a pain to use than it is worth. Chop that 30 minute epic into shorter clips. Divide them by timbre, speed, pitch, and so on.
The more “digestible” field recordings are, the more likely they will be previewed, selected, and used.
Read A Blueprint for Mastering Long Field Recordings for other ideas.
I’ve often seen sound effects presented in dual mono: both left and right channels contain the same content. Avoid this. Why? Dual mono ambiences don’t convey space and depth. Specifics sound clips shared in dual mono doubles size for no reason.
Present specific sounds in single-channel mono.
Have a mono ambience? It won’t be give the sense of immersion. Consider trashing it, or stereoize it.
Lack of Take Variation
The last post shared tips for recording diverse performances. Sometimes field recordists can forget to alter their technique and end up capturing a helicopter door shutting with the same performance ten times in a row. They all sound the same. It’s not easy to be inspired every time you punch in. Don’t worry, though. There’s an opportunity to salvage these sound fx during mastering.
Be unforgiving: delete a performance that sounds the same as the last. There is no value having identical takes. Listeners use diversity to make their choices. Takes that sound the same slow down a fan’s decisions, complicate choices, and make browsing tedious. Instead, find the most evocative, distinct versions of a sound and use only those.
You’ve exhausted yourself bashing panes of glass with a baseball bat. Now you have dozens of field recordings of shattering glass. Is it best to collect them into a single file with each burst spaced a second apart? Another approach is to share each hit as its own, distinct file.
The community is split on the answer. Game audio pros and film editors may like it one way. Video editors and podcast creators may prefer it another. Find out which style your audience prefers and deliver for their needs.
Leaving Errors in the Track
This is one of the most common mastering problems. Often problem sounds remain in sound library clips. The most common are hearing a field recordist speaking, microphone bumps, and clothing movement. Unfortunately, these field recordings can’t be used if these problems remain. So, if you don’t remove the, some editor down the line will be forced to do your work for you.
This is why the first step when editing is remove all sounds that do not contribute to a field recording’s expression. Think of this in two ways.
The simplest task is to remove a recordist’s speaking, slating, and breathing. Edit out of their movement (footsteps, jacket rustles, shoe squeaks). Cut away microphone bumps and wind blow-outs. Get rid of all incidental sounds.
The second approach is conceptual. The idea is to slice away sounds that do not contribute to a field recording’s intended expression. So, while a motorcycle passing amidst a market ambience may not be an error itself, it isn’t the point of the recording; crowd sounds and shopkeeper activity are instead. Remove the motorcycle and weave the remaining clips back together to create a complete, uninterrupted expression: a market atmosphere. Other examples: bird chirps while smashing rocks, a drain trickling while recording drizzle, seat creaks during a jumbo jet flight.
Need to chop out a sliver of sound and seal the remaining clips together? Ensure the transition from the first to the second are smooth. Otherwise, sudden changes from one piece of audio to second can create distracting jump cuts or, in the case of repeating files, bad loops.
Generous cross-fades are often the best solution. For the truly dedicated, matching waveforms at the sample level avoid pops in looping files.
The last post mentioned that a sound’s most fragile places are its beginning and end. That described distracting errors that can occur before and after the core of the sound.
This can happen when mastering, too. Draw fades at the beginning and end of sounds so they don’t blast suddenly, or end jarringly. This also helps avoid a pop at the beginning and the end of a sound created when a waveform isn’t “zeroed.”
Don’t overdo it, though. Excessively fades take too long to get to the point. They also rob valuable seconds of principal sound that is lost while a field recording ramps up to full gain.
What’s the best fade duration? For ambiences, find a spot between loud sounds and apply a fade of less than a second. Specific sounds are trickier; the best method often depends on the subject itself. Using your ear to determine what’s natural is the best idea.
This post has some suggestions about creating good fades.
It’s not easy to know the perfect levels for capturing field recordings. Often scouting and test recordings are needed to find the right gain to capture a sound properly. That’s not always possible though. It’s common to begin recording, realize that levels are improperly set, and then adjust them while recording. Just the same, these level adjustments cannot remain in field recordings. They sound like a sudden surge or dip of sound that distracts the listener.
The easiest way to fix level changes is to divide these takes into two separate clips: before and after the level change. With a bit of practice, volume graphing and iZotope RX’s clip gain can undo the damage. This technique creates “nodes” on a “gain line” that can be adjusted with pinpoint precision. With care, this will reverse the level changes and make sound appear uniform.
I did this myself just yesterday: my Sony PCM-D100’s -20 dB pad was accidentally switched when I pulled it out of my bag. As a result my skytrain platform atmosphere was captured too quietly. Not every quiet take happens by mistake. Many field recordists set levels conservatively to avoid accidental peaking. At other times, quiet sounds may be recorded too loudly. Soft wind in pine trees mastered to -5 dBFS will sound unbearable. Instead, it’s best to diminish the gain to convey the soothing sound of wind through the needles. Whichever the case, these files must be normalized so listeners hear them properly.
What’s the best level for field recordings? There isn’t one perfect answer. Some prefer sounds presented loudly so they can select them easier. Others like them quieter so they fit into projects with little work. You can choose what’s best for your own fans from the two most common approaches to sound fx normalization.
Regardless of your approach, all sounds must mastered proportionally: make crashing glass louder than cafeteria crowds, which will be louder than a dribbling stream. Of course, most of the time field recordings will be recorded with proportional levels like we saw in the last post. Master then with the same approach.
Processing and Audio Restoration
I’m old enough that I remember when noise reduction plug-ins didn’t exist. Most audio restoration was done with expensive outboard gear, if it was done at all. Now fixing clips, pops, and noise takes only a few seconds of work. Restoration apps from Accusonus, iZotope, and others can produce miraculous results. Field recordings that were useless before can be reclaimed.
However, there are two unfortunate truths to using restoration plug-ins:
- Using the apps takes a lot of skill. With just a few clicks anyone can strip a hissy forest atmosphere of noise. The simplicity of these apps and plug-ins lead beginners to believe that results are easy.
- It is difficult to hear problems. When used without care, restoration plug-ins damage field recordings in subtle ways. It’s often difficult to hear these errors without a lot of critical listening experience.
Other than leaving slating and movement in field recordings, I find excessive restoration processing is the most common mastering error I hear. I’d go so far to say that restoration apps have cause far more problems than they have solved.
I can relate, though. No one wants to trash ambiences they’ve spent hours working for in the field. There’s a temptation to use whatever methods possible to reclaim that time and effort. It’s always a good idea to try restoration tools before deleting a field recording. Just be mindful of these common, overlooked errors:
- Artifacts. When noise reduction is pushed too aggressively, artifacts can occur. These sound differently depending on noise. Listen for watery or slurring noise and sparking digital tinkling (like “singing robots”).
- Pumping. Some noise removal apps try to sculpt away noise as closely as possible to a sound’s crests and dips. However, often they don’t follow the peaks and chasms fast enough. The result is sudden surges or dropouts of noise that seem like the audio is pumping or pulsing.
- Pushing too far. Excessive reduction levels can pull the natural airiness from sounds, leaving them sounding thin and lifeless.
- Slurring. Performing spectral repair on too broad of a selection can add a combing or slurring aspect. This is often fixed by choosing more gentle settings or fixing a smaller error.
- Sculpting. Spectral repair can remove entire areas of the spectrum. I’ve seen this done with airplanes flying overhead: the high frequencies are completely removed around the swell of the pass, leaving the low end behind. While this is a quick fix for removing noise and wind, only part of the spectrum remains. The result? The remaining audio sounds unnatural.
- Gaps. Incorrect settings can remove a click but leave behind a very slight gap or “hiccup” in the sound.
- Thumps. This happens when settings are too conservative. The click isn’t removed entirely and a slight thump remains.
- Combing. Hums usually appear in harmonics. Improper settings leave behind strips of audio that sound similar to combing. Dig in deeper and narrow your bands to avoid this mistake.
Restoration apps are valuable tools. It’s important to learn them and develop skills de-noising, de-clicking, and removing hum. Just don’t rely on them as a single-fix band aid for every error in the field. Train your ear to hear restoration errors you may mistakenly commit. If in doubt, avoid processing and trash the sound.
EQ can be used to enhance a sound to pull out deep, resonant frequencies, or reveal crisp, sparkling higher tones. As with audio restoration, be careful when using EQ to repair a track. One common error I see is using EQ to remove wind problems.
Wind is difficult to remove from field recordings. Why? It overlaps many valuable frequencies. Removing it risks trimming away the good audio along with the bad. It’s common to hear low shelf filtering used to EQ wind rumble away from a field recording. In most cases it’s ineffective. Wind is often so strong and invasive that extreme EQ is needed to reduce it to the point where it doesn’t interfere with the sound. However, this usually strips the heft from a recording in the process.
Earlier we saw the problem with providing dual mono sounds. You may be tempted to use a plug-in to transform a mono sound into a stereo one. They can also change the breadth of the stereo field.
Be careful. These plug-ins can very easily introduce phase problems. Watch a phase scope while working a field recording’s stereo field.
Upsampling is the term used when converting a field recording from a lower sampling rate to a higher one. It’s a bit of an artificial practice. Why? A sound recorded at 44.1 kHz captures up to 22050 kHz of sound. Converting this 96 kHz? The new clip will still only have 22050 kHz of audio, with the rest of the frequencies filled with “dead air.” Upsampling doesn’t provide any benefit. It just adds to the size of the file. Avoid this.
Share field recordings as interleaved files. These files provide all channels – left, right, stereo, mono, surround, whatever – combined in one file. Avoid mastering left and right channels are separate sound files. This adds to library bulk, risks losing a channel, and makes collections harder to search. Most metadata apps can select, audition, and transfer single channels from interleaved files anyway with “leg pickers.”
Mastering for Everyone
Usually field recordists remove some technical errors and then consider mastering complete. Maybe some troublesome frequencies are notched out. Perhaps a siren is removed from a room tone. Maybe wind in winter wheat is adjusted so the volume sits calmly in the ear.
This is a great start. It eliminates the most obvious problems. It tweaks field recordings so they’re heard more pleasantly. What is often overlooked is that mastered field recordings have great potential to benefit from context, too.
How is context important when mastering sound clips? Context considers how a field recording will be used, who will use it, and why.
Of course, this approach is quite different from mastering music. In that case, songs are mastered to sweeten the sound. They’re also mastered so albums can be heard well by people on many different systems: on mobile phones, in cars, shopping malls, and so on.
Mastering for field recordings is different. The beauty of sound fx mastering is that it provides an opportunity to alter the sound not only for corrective purposes, but to manipulate a field recording so it meets the demands the people that need them.
Some of this is technical. For instance, a rainstorm will be mastered differently for:
- Film editors: realistic, quiet, relative levels
- VR designers: immediate and immersive performance
- Radio spots: choosing a classic sound for quick recognition
There are others. Each of these groups master to recreate reality with a slightly different approach.
It’s possible to master field recordings with a creative approach, too, such as:
- Abstract soundscape field recording albums: present and interpretive, focusing on sounds that tell a story or convey a sense of space
- Sound design: highlighting frequencies to compliment other sounds
- Art installations: manipulating rhythm and tones to invoke emotion
These groups use mastering to transform a listener’s perception of the original event, to present an idea expressed as sound.
As you can imagine, it’s not possible to master a sound so that each of these groups will be completely happy. This is why context is so important. Without it, a field recording will be unfocused. It tries to be too much to everyone. It’s far better to pick how a field recording will be mastered and stay true to that path. While it does make the sound useless to some groups, it enhances its power for the others.
How can you do this? Ask yourself these questions:
- How will the soundscapes be used?
- Who will use them?
- Why will they choose your field recordings?
Once you answer these questions and apply them in mastering, you’l have the opportunity to inject intentionality. With this sense of purpose, mastering can highlight technical features or change the interpretation and experience of the field recording itself.
There are many ways to master sound. It’s always better to choose a mastering style best suited to your listeners.
Next post: overlooked sound library curation mistakes and tips and tricks to fix them.