Voiceover Tech: How Loud Do They Want? – Tuesday Tech Tip

Among the variables we can control in our home voiceover studios, setting audition level can be tricky. How loud should that audition be? It’s actually tough to know.

When it comes to levels, Input is not the same as Output. This is a key point in recording – especially when it comes to our voiceover performances. As lifelong consumers of audio, we are all used to adjusting only the playback volume. As I mentioned in the last article, when stepping up to the mic in our own studios, we have to control our Input, or we could end up with unusable results. 

Once we capture things appropriately, what next? Is there an appropriate volume level for delivering our auditions? (As I’ve discussed before, once we have the job there should be no questions about those deliverables).

First – what exactly do we mean by “Levels”?

In voiceover recordings, we’re most frequently talking about the “Peak” – this is the loudest single point in the recording. We use the decibel (dB) scale for this. Whatever hits the highest value – even if it’s just for a microsecond – is the number we get for Peak. There could simply be one random, tall wave in the entire recording providing the value. Remember, we use a negative scale in the studio, so -1 dB is a higher value than -9 dB.  

In any recording, the Momentary Peak changes all the time unless we are recording a static test tone. Because we don’t want to hit 0 dB and distort our recordings, we tend to use certain maximum Peak values as a kind of shorthand reference. It just doesn’t tell the whole story.

Loudness is different than Peak

That’s where Loudness comes in. Loudness is always a measurement over time. It’s a weighted average that puts a number on the entire recording, or a selection of the recording. Since we don’t actually tend to hear a single peak as “loud”, a Loudness scale is designed to mimic the way human ears perceive sound. It’s similar to the dB Peak scale as it’s also basically logarithmic.

There are two commonly referenced scales for Loudness (kind of like Centigrade and Fahrenheit – two temperature scales measuring the same thing). These Loudness scales are RMS, which stands for “Root Mean Squared” and LUFS  or “Loudness Units Full Scale” (there’s also LKFS, which is equivalent to LUFS).

(I touched on this subject back in “How Loud Is Loud?”, discussing the differences between Loudness and Peak values).

The challenge facing us when we go to send off our auditions is that there’s no real consensus about what is appropriate for Peak or Loudness. 

One place we agree: Audiobook recordings

Let’s start with one place where there is pretty strong agreement – delivery of audiobook recordings. ACX (the Audiobook Creation Exchange) has the benefit of being a major player (helped by the fact that they are a division of Audible, which is a division of Amazon). That means they get to set de facto standards. They have clear guidelines on their project deliverables – Peaks no higher than -3.0 dB and Loudness between -18 and -23 dB RMS. 

Commonly, when audiobook narrators deliver their auditions, they use a “Mastering” process to adjust their Loudness levels and Peak values to meet those specifications. Meeting delivery spec is actually not required on auditions, but not doing so runs the risk of being discarded. Over the years, potential clients (the Rights Holders or Authors) have learned what properly prepared audio should sound like. The safest bet here is to master your audition to those final project delivery specs.

Why should we worry about delivery Loudness?

Let’s think about things from the audiobook client’s end. A narration project gets posted and auditions come streaming in from voice actors. Perhaps the client limits things to only 50 submissions. Even with that fairly small number of auditions, listening through all of them becomes a significant investment of time and attention.

Their task at that point is likely to be eliminating the non-competitive reads. That’s where poor recordings – background noise, distortions, reflections in the space, etc. – immediately get nixed. After all, each submission tossed out simplifies the task of choosing the “right” narrator from the ones remaining. 

Even if the person who posted the project has never evaluated auditions, by the time they are about five or six files into the process, they likely have an expectation of the playback volume. Just as you don’t expect your radio or streaming service to suddenly double or halve the volume, most of the potential narrators have likely aimed at the same delivery spec (that Loudness of -18 to -23 dB RMS). If the next one starts playing at a super low volume, the client may be less likely to reach over to turn things up. It becomes a quick litmus test whether the narrator knows their stuff.

Putting it simply: there’s an established Loudness delivery standard within that niche of the voiceover business. It’s best to meet those when auditioning

It becomes much more of a wild west when it gets to non-audiobook auditions. There simply is no standard delivery value for Peak or Loudness. What we are left with are odd clues and general practices. I’ll talk about that in the next article.

This information recently went out directly to my email community. Each week I send out a new article about using technology in the home voiceover studio, and how to balance those challenges with the need to be creative behind the microphone.
If you would like to join in to receive those emails the day they publish, please take a moment to share your contact information through this sign up form.
Thank you!