SHARE

Over the years I’ve seen a lot of confusion out there about levels and clipping during mixing – but only recently I came to realise that the confusion was so deep that people were altering their mixes to avoid clipping that wasn’t even really happening!

Once and for all, I thought I’d lay out for people in highly technical terms, but hopefully still keeping it fun too, everything they need to know about digital audio and clipping so that they can finally realize how little they really need to know.

If your the type of person who has found themselves worriedly watching the meters more than your listening to the sound, this tutorial will put your mind at rest and your concentration back to the fun part: making music.

Transcript

Intro: why this video exists

I’ll say this upfront: this wasn’t a topic I really planned on covering. Yet every time I put up a video, some eagle-eyed scout looks around on my screen and comments with some mixture of concern, confusion, or critique about what they see in my Reason Mixer. And it wasn’t until recently that I realized how deep this confusion went. In fact, I came to realize that these misunderstandings are actually causing some people to compromise their mixes. And so I thought maybe we should discuss ideas like clipping, bit depths, and how to know when a red LED is your enemy and when it’s your friend. But my biggest wish is that once you understand this, you mostly forget it and go back to the real task, which is making music.

The one simple idea: don’t sweat it

Because everything we’re going to talk about today amounts to one simple idea: don’t sweat it. That red LED phobia that so many of us have from years of conditioning? Yeah, for the most part, don’t sweat it. To understand how I could suggest something so careless and reckless, we have to discuss audio bit depths inside of Reason.

Why we fear zero dB

First of all, why is there fear in the hearts of musicians and producers the world over when anyone mentions zero dB? The answer lies in how we all consume music.

16-bit fixed-point audio

Ever since music went digital, music is almost always listened to as what’s called 16-bit fixed-point audio. To record a sound digitally, thousands of times a second, the amplitude of a signal is measured and documented. And if we looked at those measurements over time, we’d see the very familiar audio waveforms we’ve grown used to. But digital audio pioneers were faced with a problem: real-world dynamic range is virtually infinite. This has a lot more volume than this, which comparatively has infinitely more volume than this.

To represent every possible dynamic value in between those extremes was impossible, not least because the more dynamic values we need to represent, the more storage it takes. And back when they were working this stuff out, this was a 250 megabyte hard drive. That’s megabyte, not gigabyte. So they defined a range of not the quietest sound possible and not the loudest sound possible, but a workable range in the middle, with over 65,000 possible values in between.

What clipping actually is

Their method worked amazingly well when a signal fell within that range. But when it went over, they had no choice but to represent the higher signal with their highest number possible. So instead of accurately plotting where the waveform is at, we end up hitting the ceiling of the numbering system and storing the maximum value over and over. The end result is a waveform that looks like it’s gotten a haircut. It’s been clipped. And that’s why we call it clipping distortion. It looks wrong and it sounds wrong.

More bits, wider dynamic range

But what if we had more numbers to store more signal value possibilities? We could accommodate a wider dynamic range of sound before we had to clip the values. All it takes is more bits, and more storage to store those bits. But thankfully, now we can put 500 of these onto one of these, so space is less of an issue.

Without going too far down this rabbit hole, just understand that higher bit depth audio means a wider dynamic range of sound. 24 bits can handle more values before clipping than 16 bits, 32 bits even more, and by the time we get to 64 bits, we’re pretty close to representing every value that’s measurable.

How Reason handles bit depth

So how does this all relate to Reason? Well, I’m glad you asked. Reason uses high-bit-depth audio for all of its internal signal routing. The audio that flows from synths to effects and down the mixer channels is 32-bit floating-point digital audio, offering enormous dynamic range. When we start summing those signals together into group output buses and the master output, Reason jumps up to 64 bits for effectively limitless signal headroom.

There is no “too loud” inside Reason

So what does this mean for you? It means what I said at the start: don’t sweat it. It means that for as long as signals are traveling around inside of Reason, they won’t clip. There is no “too loud” inside of Reason’s mixer or Reason’s rack. In fact, the meters on Reason’s mix channel aren’t even concerned with representing a clipping point.

The VU offset meter

And here lies one of the key points of confusion. The VU offset metering on the channels is its own scale, bouncing up and down, showing a smoothed moving average representation of channel levels. It was never meant for you to make decisions about clipping.

Demo: a -1 dB signal with red LEDs

Look at this sine wave I have. I’ve wired the Selig Gain Rack Extension to it on its way to the mixer, and it’s telling me that it leaves Thor and goes into my mix channel at -1 dB on the peak dBFS scale. And sure enough, it’s passing through my mixer and out the master fader at the same -1 dB. But look at the channel meter and you’ll see a visual reading of +8 dB on Reason’s VU offset scale, and those dreaded red LEDs. Remember what I said, though: don’t sweat it.

Why you shouldn’t chase the red LEDs

And this is what led me to make this video. I’ve seen people that treat every mix channel like it’s at risk of clipping and adjust their faders until the red LEDs go away. All that does is complicate your mixing process, and it might even force you to make compromises in your music based on what you think you’re seeing. And I certainly don’t want that to happen.

Leaving the clipless world

Now, unfortunately, until our cracked team of R&D engineers creates some neural jack that connects our brain directly to Reason in order to actually hear our sounds, we eventually have to leave the clipless world of 32-bit and 64-bit floating-point audio and go out to our sound cards and speakers, or we export files for consumer formats like iTunes and YouTube. It’s in that final external output stage that we are thrust back into the limitations of fixed-point audio, and suddenly we have reason to fear 0 dB.

Setting the master fader

One obvious way to make sure that we’re not exceeding 0 dB on our master output is to simply turn down the master fader while checking either the master meter or the big meter in Reason’s rack. When set to peak mode, which is the standard reference for digital audio, we can simply pull our fader down until our master signal is not going above 0 dB. Easy enough.

Why the master meter looks so hot

But this brings us full circle to how we got to this topic in the first place. People keep seeing my videos and remarking on the red LEDs, both the channel faders, which we all know now are fine, but also the red LEDs they’re seeing in my master fader. From the looks of it, I’m trying to send my music out above 0 dB and I’m clipping. And I’ll concede that they’re right: that is how it looks. But that’s not actually true, of course. So why do my meters look so hot, but without harsh clipped audio?

The loudness wars

The answer lies in something called the loudness wars. Basically, commercial releases are in a competition to see who can sound the loudest. It’s a war they’ve been waging for more than 20 years. But we all know by now that there is no way to be louder than 0 dB in consumer 16-bit formats, but there are ways to sound louder than 0 dB. I won’t get into philosophical discussions on the merits or the evils of the loudness war, except to say that, love it or hate it, it’s the world we live in.

How maximizers work

That’s why, when I’m not having my tracks professionally mastered by a mastering engineer, I at least do the loudness work myself with any one of Reason’s maximizers. Maximizers are surprisingly transparent, highly effective look-ahead digital brick-wall limiters. Most follow in the footsteps of popular hardware limiters like the Waves L2, which became a mastering-house mainstay. Any maximizer, whether it’s hardware or Reason Rack devices, works by a clever method of turning down audio peaks in a signal in such a way that you can drive your music louder while knowing that none of the peaks are getting above 0 dB.

Reason comes with a maximizer as part of the MClass Maximizer mastering suite of devices. Today, though, I’ll demonstrate loudness mastering using iZotope’s Ozone Maximizer, but I encourage you to try out all of the options available in Reason to see which one works for your style.

Keep limiters off while you mix

The first thing I should say about mastering limiters is that you don’t want to have them on while you’re creating or mixing your music. If you do, you won’t get a clear picture of your mix and how instruments relate to each other, because the limiter is altering your sound. Turn up the bass guitar too loud and suddenly your vocal might seem more buried than it once was.

So during mixing and production, use the previous simple method of setting your master fader so that the output to your sound card isn’t clipping. When you’re done, you can snap your master fader back to its default, or what’s called unity position, by Command-clicking on a Mac or Control-clicking on Windows, and then drag your maximizer into the master section inserts. If your Reason song template has the insert FX bypass button activated by default, you’ll want to deactivate that for illustration purposes. I’m also going to drag in a Selig Gain so we can monitor the peak output level.

Maximizing loudness with the threshold

Maximizing the perceived volume of your mix mainly happens with one control. On Ozone, we adjust the threshold fader to taste. As I move the fader down, you can hear the perceived volume of my mix goes up. But notice on Selig Gain, the peak volume isn’t going above 0 dB no matter how low I set the threshold. Ozone just processes the signal more and more, and the perceived volume goes up as a result. There will come a point where the result becomes very audible and undesirable. I’ll leave that point up to your taste to decide. For me, something around here is working.

Bypassing the maximizer

Now, watch what happens if we bypass our maximizer. Not only does the perceived volume of our song go down, but the peak level goes up above 0 dB and into clipping. So when people are noticing my master fader’s LED hitting 0 dB, it’s not because it’s going over 0 dB, which you don’t want. It’s because it’s getting right up to 0 dB without going over, which is what you do want.

Wrap-up

So, I hope I could help clear up some of the confusion and anxiety out there about clipping and digital audio. But like I said upfront, the most important thing you could take away from this tutorial was actually to pay far more attention to your melodies and having fun than you do to how much headroom there is in 32-bit floating-point audio streams. Sure, when it comes to the final output, it wouldn’t hurt to keep an eye on your peak meter level. But before that, don’t sweat it. So good luck, and I’ll see you again soon.