When we consider capturing video and audio, it’s common to assume that the audio aspect is more straightforward. In most cases, you might attach your shotgun microphone to the camera, ensure the audio levels are good, and then move on to setting up the lights or camera angle. However, why is sound often regarded as more important than visuals?
Documentary filmmaker Michael Moore once said, "I tell students that sound is more important than pictures. The audience will forgive you if an image is a little blurry, but not if they can’t hear or understand what’s going on. Sound carries the story."
This is a sentiment shared across the filmmaking industry.
Think about it — how often do we see and forgive visual slip-ups? Perhaps the cameraman is visible in a reflection, or maybe the extra in the background is standing dead still, or the visual effects are a little wonky. It happens all the time, and usually, you don’t catch it until an eagle-eyed content creator makes a blooper reel on YouTube about the mishaps.
However, the moment we hear audio that doesn’t sound right, we are automatically withdrawn from the viewing experience. This is because our brain processes auditory information differently than visual information. Auditory information is crucial for communication, and as such, we are highly susceptible to inconsistencies.
To ensure there are no inconsistencies with your audio, this article will discuss the core principles for setting audio levels correctly to maintain the expected quality.
Setting the correct audio levels in production
Video has become increasingly salvageable when mistakes are made during filming. When shooting in RAW, especially with a high color bitrate, you can often recover blown highlights in overexposed footage and details in shadows in underexposed shots. There was even a time when a camera was in production, and you could change the focus after the recording!
Sound, on the other hand, has a limit to how much it can be adjusted. Unless you're recording in 32-bit float, it doesn’t take much to push sound beyond the point of saving. Therefore, whether it's capturing dialogue on set, ADR, or ambient recordings, the sound must be captured perfectly.
First and foremost, you need to ensure that the microphone is positioned appropriately. Simply put, the closer the microphone, the better the sound. The video below highlights the importance of choosing the correct microphone.
Video via Videvo.
However, in this article, we want to focus on what happens after the microphone setup — the recording and editing of sound.
While the microphone captures the audio, sound crews use audio recording devices like field mixers or handheld recorders to record it. The video below does an excellent job at explaining the differences between the two.
Video via Videvo.
From inexpensive field mixers to devices that can cost up to $3,000, one thing they have in common is a ‘gain knob.’ This gain knob allows sound recordists to control the input level of the audio signal, ensuring that the sound is neither too quiet nor too distorted.
Not sure what gain is? Well, think of it like volume, but instead of controlling how loud or quiet the sound is, it controls the amplification of the audio signal. So, while turning up the speakers on your PC will make your music louder, it won't degrade the audio quality — it will just annoy the neighbors. However, increasing the gain amplifies the signal and may lead to distortion as you are 'amplifying the signal.'
When you adjust the knobs, your sound recording device — or even your camera monitor if the mic is plugged into the camera — will display audio levels.
These are visual representations of the audio signal, and it's important to note that the level meter uses decibels as its unit of measurement.
While your ears are invaluable for judging how the audio sounds, the level meter on your device provides precise technical feedback. It helps you identify where the audio falls on the spectrum and can highlight hidden issues you might not initially notice, such as low rumbles.
The height of the level meter is divided into distinct, often color-coded sections that represent different signal ranges. As the meter responds to the audio signal, you'll see the colors shift as the levels rise.
- The green signal is safe and represents the optimal range.
- The yellow signal indicates the signal is hot and approaching the peak.
- The red signal means the audio levels have been clipped.
Before we continue, what is clipping? This is when the audio signal has surpassed 0 dB, resulting in distortion that renders the sound harsh, fuzzy, and often unusable. This happens when the signal is so loud that it overwhelms the microphone's amplifier, resulting in distortion.
Video via Alicia Butterworth.
Unfortunately, quite like blown highlights from an 8-bit video, clipping is rarely salvageable. Some devices feature auto-clipping, which safeguards your audio from clipping if the source becomes too loud.
To warn you of this, the audio levels will appear red on your recording device, or if you have a device that has a backlit LCD screen (no color information), like the Tascam DR-40X, there is usually a dedicated warning light.
Is there an exact level for capturing audio?
Not exactly. The goal is to capture the cleanest signal possible, which, as previously mentioned, requires getting your microphone as close as possible but adjusting your gain control. Hence, the signal stays between -12 dB and -10 dB on the meter is a good practice. This also allows for an essential concept in audio recording known as headroom.
As the name suggests, just like leaving enough headroom in a video composition to ensure your subject’s head doesn’t hit the top of the frame, headroom in audio refers to the space between 0 dB and the peak levels of your audio. It gives the audio editor (or yourself) enough room to add effects without risking exceeding 0 dB.
But how do you stay between -12 dB and -10 dB with a source that fluctuates in sound intensity? Well, that’s the primary challenge with recording audio is predicting signal fluctuations.
Take this scene (at 01:30) from Martin Scorsese's “Silence.”
Via via RR PLANB.
In a rare case of poor audio in a feature film, there's a noticeable hiss in the audio track during the prisoner's dialogue. This was likely caused by the sound recordist setting the audio levels when the tide was out without considering the intense sound of the crashing waves. As such, when they had to raise the audio in post-production, it raised the digital noise floor, which introduced hiss.
While it seems like a small detail to pick on, just this one scene sparked several social media and forum posts discussing the poor quality.
“But that’s not the end of the audio mix issues. In some of the most important and emotionally resonant scenes, you’re drastically taken out of the experience by distorted sounds and inexplicable dips in volume. These technical issues, when present in such astounding numbers, stack up and make for a very disjointed and irritating viewing.” — Michał Żak
Ouch.
It's unlikely you'll find yourself filming a scene like that, and in most cases, such as a corporate interview, it's safe to assume the interviewee won’t have a wide range of volume. However, in a narrative short where a character pleads with their partner to stay, the audio might shift from loud shouting to quiet moments of desperation. In these scenarios, you can’t keep the gain the same—otherwise, the levels will clip during loud parts and be too quiet during soft ones. Adjusting the gain becomes necessary, but this task typically falls to the sound recordist, as the boom operator will be focused on holding the microphone. This explains why there are so many crew members on a film set!
However, don’t get too disheartened if you’re a novice filmmaker without a dedicated sound crew to manage your audio levels. As mentioned earlier, much of today’s audio gear has auto-clipping features to prevent distortion. Similarly, many modern cameras with audio inputs offer auto-leveling, an audio function that adjusts gain based on the signal from the scene. While this feature is helpful, like autofocus in cinema cameras, it’s not a substitute for the dedicated role. There are noticeable differences between auto-leveling and manual adjustment.
If your camera doesn’t have this feature, your best approach is to rehearse the scene with your actors and mark the moments when sound levels will shift. This will give you the cues needed to adjust the gain while filming.
Audio levels in post-production
Now that we understand the basics of audio levels, the settings on a field recorder, and the gain knobs that control the input of an audio signal, let’s look at some recommended audio levels for different types of sound sources within your NLE.
While recording sound involves specific recommended dB levels, these aren’t the levels you want to stick with when it comes to mixing within your project. Imagine if the dialogue, music, and sound effects were all set to the same level on the timeline—it would sound chaotic, with nothing standing out clearly.
Below are some recommended levels for various audio components in video production. These are guidelines, not strict broadcast rules, but they provide a solid foundation to build on as you gain experience:
- Overall audio mix: -10 dB to -14 dB
- Dialogue: -12 dB to -15 dB
- Music: -18 dB to -22 dB
- Sound effects: -10 dB to -20 dB (with peaks up to -8 dB)
An ideal approach is to assign each type of audio to its own dedicated track (e.g., one for music, another for sound effects) and adjust the overall track levels, rather than lowering each individual audio clip separately.
Now, it's important to remember that these are guidelines. For instance, if you're editing a conversation scene with an underscore, and the music feels too loud, even if it's within the recommended -18 dB to -22 dB range, you don't have to adhere strictly to that. You can lower the music further to ensure the dialogue is more audible. Using these levels as a starting point helps you avoid distortion while maintaining flexibility
Additionally, while the audio levels may be within legal standards, it's always a good idea to render your video and play it back on a variety of devices — computer speakers, cell phones, tablets — or even send it to a friend to check if they can hear it clearly. This helps you gauge whether the audio is set correctly, as the room you’re editing in can directly affect how you perceive the audio quality, even if the levels are technically correct.
If you're looking for an easy way to adjust audio levels without complicated software, try WeVideo!
Unorthodox uses of bad audio
What about creative applications? You can do many things to the camera, film, or lens to create specific in-camera effects. For instance, you can apply a filter to the lens to create a hazy, dream-like glow. You can even detach the lens and manually hold it in front of the camera body to let light leak into the image, and the list goes on. These methods physically alter the image to elicit a specific visual reaction from the audience.
However, with audio, it’s less common to degrade the sound quality intentionally. Of course, creative elements can be introduced into a film’s soundscape. A good example is the D-Day sequence in “Saving Private Ryan.”
Video via MovieClip.
At 04:50, when the lead character, Captain Miller, experiences shell shock, the sequence features a soft ringing sound with a low-pass filter over the diegetic audio, making everything sound slightly muffled. However, the sound eventually returns to normal. That effect helps convey a specific moment or feeling within the story. This, nonetheless, is a creative choice and one that serves the story and, in particular, the moment.
Yet, when filmmakers deliberately keep the audio unclear for extended periods, even highly respected directors can face some criticism. Take, for example, the esteemed director Christopher Nolan. While he’s often regarded as one of the favorite filmmakers of today, even his fans will admit that many of his films feature muffled dialogue. Think of Tom Hardy as Bane in “The Dark Knight Rises,” some moments in “Interstellar” and “Dunkirk,” and most notably, “Tenet.” Numerous articles have discussed how difficult it is to hear the dialogue in these films.
Video via Thomas Flight.
Now, Nolan has offered explanations for this creative choice though the specifics of his reasoning aren't always embraced by the audience. While his intention may be pure, viewers still need to understand the characters' dialogue to fully engage with the story.
Conclusion
Throughout this article, you should now understand the principles behind setting the correct audio levels when recording and how to set the right audio levels during editing. However, remember that this is just one aspect of audio editing.
Like a cinema camera may produce gorgeous images, they don't truly come to life until color correction enhances the detail. It's the same with audio. Setting the correct audio levels is only one part of the process. There are deeper layers where the audio can be polished with tools such as EQ, processing, and compression.