Audio Limiters Explained: How to Control Loudness & Prevent Clipping (BGAE Ep.5)

Audio Limiters Explained: How to Control Loudness & Prevent Clipping (BGAE Ep.5)

Audio Limiters are our last line of defense. Just like a compressor, a limiter has a threshold. But instead of being set really low in the dB range, we normally set it around -1dB to -2dB. A limiter is typically found on the master channel, to control the absolute limit of everything all together at the same time. Its primary goal is to prevent clipping on the master channel and enable us to increase the loudness of our tracks to the level of a final product ready to be released into the world of streaming platforms.

We don't really have to go too deep into Limiters after our lesson on compressors because they are essentially the same. The biggest difference between a compressor and a limiter is that a compressor allows audio through, while a limiter does not! They often share the same controls, and sometimes even less control than a compressor.

We can look at the stock Ableton live limiter which gives you Gain, Threshold, and release time for control. 

Or we could look at something like the Shadow Hills Mastering Compressor which gives you so many controls and types of internal circuits that we could have a whole separate lesson on this limiter alone.

---

 

Since we’ve already covered Attack, Release, and Threshold. We’ll focus on the controls that a limiter has that differentiates a limiter from a compressor. I’ll also get into the Gain levels for master but only slightly, Mastering is an absolute art form in its own right and will have its own dedicated lesson. It’s only fair that I cover some of the statistics for loudness because its relevant to limiters.

Let's Dive it!

---

 

The point of Limiting

Whereas a compressor has a threshold, for a limiter, it's called a ceiling. The goal of a limiter is to prevent any audio from passing this set level. A compressor has a threshold which indicates where the commission will start. A limiter has a ceiling which indicates the level at which nothing passes. This is a fundamental property of limiters.

If you recall from our lesson on what the dB is, we discussed dBFS or dB full Scale. This is a level of -100dB to 0dB. Where anything over 0dB is considered clipped because the computer starts to distort the audio due to its inability to reproduce audio at this level. 

The limiter is our last line of defense when it comes to keeping the audio from hitting 0dB/ clipping. It’s safe to say that a limiter should always be on your master channel and it should always be the last effect in the chain. If you were to put any effect or tool after a limiter it will increase the gain past your limiter's ceiling thus defeating the purpose of having the limiter on your master channel.

By the end of this lesson you’ll have a clear understanding of how to set up a limiter, and why they are important. My preferred Limiter is the Fabfilter Pro-L2 which is what I will be using to show you how to limit your tracks, but it’s also important to know that these methods apply to every limiter just like everything else I teach in these lessons. The tool isn't what matters, it's how you use them that matters!

---

 

Limiting Controls - Look ahead

The look ahead feature is the amount of time the Limiter has to “Look ahead” at the incoming audio. This allows the limiter to buffer the incoming audio to apply a smoother level of processing to it. If set to 0ms this could induce some distortion by actually allowing the audio to clip because not enough time was given to the limiter to actually catch all the peaks.

By giving the limiter a few milliseconds of look ahead time we can ensure that the limiter sees all the incoming peaks and will act appropriately to the audio, accurately limiting and preventing clipping from happening. Giving the limiter some look ahead time allows the limiter to buffer (delay) the incoming audio, analyze the waveform, and proactively apply limiting as opposed to having too short of a look ahead and forcing the limiter to act reactively. 3-5ms is normally a default timing.

---

 

Limiting Controls - Attack & Release

Just like a compressor, we’ll also see attack and release. If you want to get into how to set these, go check out the lesson on compressors where we get into calculating the timing for these. Essentially they are the delay in the response time, and how long it takes for the tool will reset back to 0.

---

 

Limiting Controls - Channel Linking 

Channel linking has to do with unifying or separating the limiter's response between your left and right channel.

When the channels are 100% linked, no matter which channel triggers the limiter, both channels left and right will be affected.

If the channels are 0% Linked, the left channel and right channel will be treated as individual channels. Essentially active as 2 separate limiters, though this sounds like a good way to keep your stereo image wide, it tends to sound fairly strange when your left and right speakers are acting in separate manners.

Using 80-100% linked will allow you to maintain your stereo integrity when it comes to mastering.

Using 0-80% linked is more so recommended when dealing with specific applications such as vocals, drums, or instruments.

By using unlinked / partial links for specific things like vocals or drums, this allows character and better transient response for individualism and uniqueness. While applying linked limiting on your master will help unify or tie together all the individual pieces of your song to sound as one piece.

---

 

Limiting Controls - Gain

Unlike a compressor where we set the threshold to match the level of the audio we wish to compress. With a limiter having a ceiling. The ceiling is equal to the limiters the output level. Most often we’ll set the Limiter’s output level or, ceiling, to -1dB or -2dB (i’ll explain this deeper in a few).

Setting your ceiling/ output to -1dB will give us a 1dB clipping buffer between -1dB and 0dB, remembering that anything over 0dB is clipping. Giving this little bit of breathing room will not be audibly quieter so dont worry about not limiting to absolute 0dB.

Now that our ceiling is set to -1dB. We use the gain (left of image) to push our audio into the ceiling. By pushing the audio level into the ceiling we then begin to limit the audio while also increasing the loudness. As you can see in the image below we aren't pushing any gain into the ceiling, the limiter is just on. Though our audio is still being limited (top right arrow) by 3.5dB because our audio in general is already louder than -1dB. This shows the importance of a limiter even before the mastering stage. It's preventing our general mix from clipping period by just being there.

The bottom right arrow is pointing to the LUFS level which we’ll get into a little more in a little bit. LUFS stands for Loudness Units relative to Full Scale and is how streaming services as well as mastering engineers measure a track's overall loudness level.

By applying a lot of gain, we push our audio into the ceiling, forcing the limiter to keep all of the audio that would be clipping. Though it keeps the audio from clipping, over-limiting such as the image below can cause audible distortion.

As you can see we’ve applied +13dB of input gain (left arrow)

You can also see that the limiter is now limiting by 17dB (upper right arrow), we’re now absolutely crushing out audio.

The LUFS level is now at a staggering -4.5 LUFS which is insanely loud and an absurd level for reasons i’ll clarify when discussing streaming.

A very important aspect to understand about limiting is that by increasing loudness we are reducing dynamics. Dynamics in audio are very important. We don’t want to remove too much dynamic range, but we also want to achieve a competitive loudness level. Though, the more dynamic a tack is the more you can push it through a system in a live event scenario. If your track is maxed out during mastering, it can actually sound quieter than a dynamic track because the dynamic track has “more to squeeze” than a track that's already had all the dynamics squeezed out of it. 

For example, The image below is at -11 LUFS which is still technically 3 LUFS above streaming standard. We have not applied any any gain and have maintained all of our dynamics (Center arrows highlight dynamic range and we can see that there is a lot of dark space where audio is not being played)

If we apply a heavy amount of gain, we drastically reduce the amount of dynamic range (center arrows). Not only do we reduce dynamic range, we are flattening our audio (red bar top left). You can see that there are almost no peaks and valleys and its practically a flat line across the top leaving no character or movement to the audio.

We’re also now at -4.5 LUFS again which will inevitably make our music sound quieter than the more dynamic song because it's already been squeezed to the absolute max.

What makes music sound larger than life is the distance between its quiet moments and its loudest moments (Dynamics)

Think about jump scares in horror movies. The audio starts to get really quiet and the suspense builds and builds and right before the jump scare, the audio is almost silent. Then all of a sudden, there's a huge loud noise alongside a frightening scene that makes you jump out of your seat!

Now imagine if that same scene had the same level of volume the entire time. The audio getting quieter wouldn't exist, and the loud jump scare sound that forces you out of your seat would have absolutely no impact because it wasn't any louder than “normal”. 

The reason why that jump scare scene was so impactful was because of its dynamic range between absolute silence and then an abrupt loud noise that jars your system provoking the jump scare! Dynamic range is very important.

Without a quiet sound a loud sound has no relative point to base your judgment on.

Without a loud sound, we wouldn't be able to say that something sounds quiet, because we wouldn’t have a relative reference.

Over-limiting reduces dynamic range making the audio less impactful and just “loud all the time”.

By limiting and maintaining proper dynamics, our music can cause more impact to the listener due to the difference in the quiet and loud parts because it provides the listener with relativity.

Someone yelling “HEY!” in a library where it's nice and quiet would probably cause you to jump up and turn quickly towards the person.

But if everyone was already yelling in the library you wouldn't know the difference and it wouldn't have any impact.

Get the picture?

---

 

Streaming services also limit your audio!

In our examples we had a level of -11 LUFS and another level of -4.5 LUFS. With our -11 LUFS track being full of dynamic range, and our -4.5 LUFS tack having almost no dynamic range.

Well, Spotify is coming to reduce your audio's loudness to -14 LUFS (integrated). With integrated being a collective average over the course of your full track. What does this mean? 

If your tack has an average (integrated) level of -4.5 LUFS across the length of the track (not dynamic at all) your track will be turned DOWN by 9.5 LUFS!!! Making your audio sound quiet all around. This is a HUGE level cut. 9.5 LUFS is a massive level change. I almost can't even stress how significant this change in level is.

If your track has an average (integrated) level of -11 LUFS across the length of the track. Your track will only be turned down by -3 LUFS. Even though your audio will be turned down, we have dynamics on our side.

Most audio playback systems such as Home radios, Car stereo systems, HiFi systems, and especially live events and concert system’s have limiters as well!

This is not only important to protect the sound system from blowing up, but important for use to understand that we still have another limiter to go through!

Take a look at these Streaming platform statistics on loudness level

Below are examples of why over-limiting with less dynamics prior to release makes your audio quieter than a less limited more dynamic audio.

-11.4 LUFS pre spotify limiting

-11.4 LUFS post spotify limiting to -14 LUFS

-4.5 LUFS pre Spotify limiting

-4.5 LUFS post Spotify limiting to -14 LUFS

-4.5 LUFS post Spotify (left) and  -11.4 LUFS post Spotify (right)

The difference in peak volume is extremely apparent when they are placed side by side with the -4.5 LUFS peaking at -11dB. While the -11.4 LUFS track is peaking at -3dB.

This makes the less limited track sound a whopping 8dB louder post uploading to Spotify!!! With the Laws of human hearing stating that 10dB is equivalent to twice as loud. 8dB louder is almost insultingly louder. Not only is it 8dB louder it maintains all of its dynamics!! 

Let's take this a step further and play this track through another limiter as though we were playing it at a concert through a system that has a limiter in place before it goes out to the speakers.

-11.4 LUFS post Spotify limiting to -14 LUFS and replayed through a limited speaker system. 

You can see here we are actually louder than the original -11.4 LUFS which is low -9.9 LUFS while STILL maintaining some level of dynamics. Remembering that dynamics allow for music impact on the listener.

Not only can we make it louder and still maintain dynamics, we can apply an unholy amount of limiting achieving -8.5 LUFS and STILL Maintain Dynamics.

-4.5 LUFS post Spotify limiting to -14 LUFS and replayed through a limited speaker system. 

You can see that if we were to slam this audio into the concert system limiter we’re actually only 0.1 LUFS louder but we’ve lost all of our dynamics making the music impactful to the listener and actually annoying due to the constant level of pure noise. 

If we applied an unholy amount of limiting to this track, we’re no louder than where we started and actually more flat lined than ever. Absolutely no dynamics Just like everyone yelling in a library, someone randomly yelling “HEY!” wouldn't make you think twice. It provokes no emotional response and desensitizes the listener.

---

 

Let this be a stark indicator that louder is NOT better. More Dynamic is better. Over-Limiting can make your audio sound quieter in the long run, make it less impactful, and out right annoying to listen to because it just becomes a constant level of noise!

Dynamic music is highly impactful, enjoyable, and engaging to the listener.

Limiters allow us to increase the loudness of our audio while also preventing the audio from clipping. On the other hand, abusing the limiter can damage your audio.

It is just as much of an art as it is a science.