Updated: Jan 15, 2021
If you were to ask someone what might be one of the most popular things emphasized within recording right now, they might respond with the term, "gain staging." I've seen it littered all over YouTube, and rightfully so. It's one of the most important aspects of recording and mixing that is violated all of the time. And mind you, most of the time, it's not deliberate. So, to help contextualize a rather complex topic, I aim to help by breaking down gain staging during the recording phase, into a more applicable topic to help you get from point A to point B quickly and effectively.
What is gain staging?
Well, you can approach gain staging in two different ways: in an analog sense and a digital sense. Both offer similarities and slight differences for their rules and how to handle gain staging according to each domain. However, one encompassing definition can still be derived. Here's a nice precise definition of gain staging holistically:
"In the analog world especially, gain staging refers to adjusting the level at each point of amplification to ensure an optimal signal-to-noise ratio, without (unusable) distortion" (https://www.izotope.com/en/learn/gain-staging-what-it-is-and-how-to-do-it.html).
Gain staging is the act of ensuring 'healthy' levels throughout each stage of the recording and mixing process - too low and you're not utilizing the full resolution of the recording medium and you might run into noise issues, too high and you run the risk of causing overloads. (https://support.focusrite.com/hc/en-gb/articles/208958089-How-high-should-I-set-the-gain-control-on-my-interface-)
Sure, this all makes plenty of sense, right? Don't record too loud to prevent clipping and don't record too quiet to prevent noise & resolution issues. And again, this is one of the most violated elements in recording and mixing. Without getting gain staging right on the front end of the recording, you will likely run into problems later down the line in your song; whether your favorite vocal take is clipped beyond repair or there's so much noise within the take, it stands out in the mix once compression is added. Sounds terrible, right? Well, it doesn't have to be, so let's get into it!
Gain Staging in the Recording Stage
Generally, a good way to think about gain staging is the ability to attenuate a signal's gain (or volume) throughout any step of the recording or mixing process. This can include: when you're recording into your DAW, when mixing (usually boosting) with an EQ plugin, or when applying make-up gain after use of compression, etc. Using this framework, one can derive the importance of maintaining an appropriate and consistent volume while making adjustments to prevent undesirable outcomes from occurring.
Now, for this blog post, I'm assuming we're working mostly within the digital domain. I'm also going to assume there are not instances of analog gear present within your chain for now, not that you need any anyways. There are just subsets of rules, additional considerations, and ways to break these rules to work in your favor and for creativity purposes (with even-order harmonic distortion, etc.), but we're going to keep it relatively simple for now, setting our sights only on the digital domain.
Typically, for the home recording chain, it looks something along the lines of this:
Artist > (Instrument) > Microphone > Interface > Computer & DAW > Output
Each point within the chain is crucial for the overall fidelity of the recording. However, the particular point we'll be examining here is the microphone into the interface.
Before I begin any serious recording, I ALWAYS ask the performer to do a practice run (or two) (Pro Tip: always record during practice runs too). And if you're recording yourself, you should be doing multiple practice runs to ensure you have the correct gain set for yourself. Sometimes that means running back and forth to and from your mic and computer. While the performer is practicing, you should be setting the gain on your interface or preamp according to their performance. This not only warms up the performer, but also allows the engineer to set their gain levels accurately. When you ask someone to just simply "talk into the mic" or "sing something for me," it actually doesn't help all that much. In fact, you're doing more of a disservice to yourself. Why?
They're likely going to be quieter (or louder) without music playing, thus causing you to set the gain too high (or too low) on your interface, introducing clipping (or noise) during the proceeding takes.
This is one of the biggest pitfalls during gain staging one can fall into, but luckily, it's easily remedied by just doing a couple practices takes.
Next, while the performer is doing a practice run, simply adjust the gain until signal doesn't clip. BUT WAIT, there's more! You can easily over-compensate during your primary prevention against clipping, where your gain won't be set high enough to prevent against recordings being captured near the noise floor.
You may be asking, "what the hell is a noise floor?" Excellent question!
Essentially, your noise floor refers to the equipments self-inducing noise, usually pre-amp hiss. You'll want to avoid this at all costs as there is no great way of maintaining clear fidelity while compensating for induced hiss during a take.
How do you avoid noise issues? Well, record loud enough to be substantially louder than the noise floor. This is basically optimizing your signal-to-noise ratio. You can read more about the signal-to-noise ratio here! If you've noticed, there is a dilemma between preventing clipping vs. preventing induced noise. Subsequently, that medium occurs somewhere within the midpoint of the recording level.
NOISE FLOOR < HAPPY LEVEL > CLIPPING
Where that level is, however, is completely dependent upon your source material, the performance, your microphone, microphone placement, etc. For example, when you record a more dynamic instrument such as drums, you're likely going to be gain staging a bit lower than you would for an acoustic guitar. So, the question becomes...
How do you find the perfect level?
Glad you asked! Before you begin, it's good to have a target for your recordings and knowing your levels on both your interface and your DAW. Your target is going to be an average level, otherwise known as your RMS level. However, instead of using an RMS meter, we're going to be visualizing this level based off of your peaks. You could gain stage according to an RMS meter, but might be way more cumbersome since RMS doesn't show you your peaks, only an average level. Therefore, you may clip during peaks without even knowing.
Overall, these averages will vary by instrument so feel free to experiment with different levels to find where your equipment's sweet spot is. A lot of determining your gain's target has to do with experience and knowing your gear. So, this is just another case of practice makes perfect. A key thing to remember is that gain staging is a part of your workflow and it's ultimately beneficial to become so efficient at gain staging quickly and effectively. Therefore, it takes serious effort and practice. Just remember, when you're gain staging...
It's All About Your Levels!
The visuals of your meter may vary. This meter is specifically from Pro Tools, but the concept remains the same. In some DAWs, metering can be set to different measuring aspects so make sure you check your metering settings (I keep mine set to Pro Tools Classic). Otherwise, take a second to search and become familiar with your DAW's metering and how to read them. It will pay dividends in the future for your workflow and audio production. More about metering here.
Taking into account the diagram above, one can see that there is a pretty large area of a target range we can aim for. You'll want to avoid the blue area in order to record higher than the noise floor. You'll also want to avoid the red area to prevent clipping. Anywhere else is fair game, but remember, it's all about averaging out your performance while preventing peaks from clipping, so be sure to not average near the yellow. Intermittent peaks are okay in the yellow, but even then, you may consider backing down on your gain a bit. Peaks are expected throughout a performance, which is why we aim to give enough headroom before clipping occurs.
A key thing to note, this is done without plugins initiated on the channel. This is a 100% dry signal, therefore your metering is completely dependent upon your gain settings on your interface. If you had a plugin that attenuated the channel's gain (like a compressor with make-up gain enacted), you'd have an inaccurate meter reading since the meter defaults to reading your channel's signal post-fader. Meaning, the meter is dictated by both your input gain and plugin inserts. So, to make this simple, make sure you do this without plugins initiated on the channel to ensure you're being completely accurate. You can change your DAW's meters to read pre-fader, but I recommend sticking to this simplified method first. Also, if your interface has a meter, then you can gain stage based off of the interface's metering while using plugins, but if not, I recommend sticking to a no plugin method for now.
Once you set your gain accordingly, you're good to leave it be for the duration of the recording (of course, within reason; making adjustments if you're averaging too high or low depending on the performance). Remember, your intentions are not to react to all peaks. That's not the proper use of gain staging. You're not trying to level out the performance's level. That can be done later on with some dynamics processing. Instead, you're merely capturing the artist's performance at its fullest integrity so you can manipulate it later on (if needed). Yes, you can adjust the gain post-recording (clip gain, trim plugin, etc.), but it's not useful if you clipped or introduced noise during the initial recording. Why? Because the clipping and/or noise is permanently baked into the recording. Consequentially, there's no perfect way of fixing it to maintain its full resolution. Therefore, post-recording gain attenuation won't fix any of what has already been captured. Undoubtedly, adjusting gain post-recording is useful for a multitude of reasons but that's beyond the scope of this blog post.
To see this in action, watch this video below for a nice tutorial!
How to Look at Gain Staging through the lens of a Photographer
Gain staging may seem like more of a difficult thing to picture, especially when you're first starting out recording yourself. However, an interesting way to frame it is through photography. Disclaimer: I'm by no means a professional photographer.
Take for instance this photo:
To the left, you have an underexposed photo. Notice how dark the original photo is. Could you turn up the exposure in photoshop? Definitely! But you're going to have a fair amount of noise (grain) introduced into the photo that you can only attempt to remove with de-noising or color de-noising in photoshop, which, in turn, ruins the sharpness and resolution of your photo. This is similar to setting your gain too low for a take; where it's rather quiet and/or barely even peaks in the green on your DAW's meter. Can you turn it up post-recording? Of course! But, when you do, you'll also turn up the recorded noise in your take that you likely didn't hear while recording; just like the noise you don't see before post in the underexposed photo. It's a tradeoff and arguably better than clipping (or overexposing), but still should be avoided.
In the middle, you have a properly-exposed photo. There isn't any visible noise and isn't overexposed, maintaining full color and resolution. Since the fidelity of the photo is maintained, the photo is able to be manipulated in post-processing without much hassle. This is similar to a well set gain staged take, where the fidelity in the take is fully
On the right, you have an overexposed photo where the brightness washes out all of the color and contrast, effectively ruining the photo. Working with this would be an absolute nightmare in post. You can attempt to remove the overexposure and white wash, but the original color will likely never be able to be brought back. You can think of this as clipping, where you drive your gain too much causing clipping and non-linearities, effectively ruining the take's fidelity. Could you save some of it? Maybe with iZotope RX's de-clipper, but it's never going to be as good as it could have been with better gain staging. Besides, the non-linear distortion will still be baked into the sound of the take itself.
We'll Leave It At That... For Now
This blog post definitely bit off a lot in a short amount of explanation. You could easily spend way more time going over gain staging, which I suggest you do. However, if you take what we put in place today and try to apply it to your recordings/workflow, you're going to find your recordings turning out a lot better and a lot easier to work with later. If you still need help, you can always hit us up to record at our studio and let us handle it so you can be more creative and focus on your performance.
Thanks for reading!
Follow us on social media!