Record Quiet And Loud Voice-Overs Simultaneously Without Distorting

If you record any type of voice-over, you are familiar with the recording level struggle. Setting a good level is easy for consistent conversational speech. However, setting a good recording level while capturing dynamic performances can be challenging. A recording level that captures soft or quiet speech is not a good setting for loud, blaring speech and vice versa. Several solutions exist to this problem, ranging from studio techniques to special equipment. The following will explore different ways to capture ultra-dynamic speech while retaining the best recording quality.

The RX Solution

One solution is to record your performance and try to correct the distorted portions of your audio after you record (post-production). This might be accomplished by using RX or restoration plugins to correct overs (or distorted peaks) in your audio. I don’t suggest this method, but plugins like iZotope’s RX10 De Clip can sometimes work well.

The “Record Twice” Solution

Another way to approach this is to record twice at independent level settings and edit out the less-than-optimal sections. This is probably the most common option. However, it takes extra time and can disturb your creative performance flow.  You can accomplish this by setting your preamp gain first to achieve a proper level for your quiet sections while ignoring the bad level on your loud sections. Then, record your entire script again after resetting your preamp gain to achieve a proper level for your loud sections while ignoring the insufficient level during your quiet sections. Then, you select the takes with the best levels for each section and edit it all together. This is a good option because it doesn’t require any extra equipment. But, the downside is that it requires extra time. And, the worst part is that you can’t correctly capture one cohesive performance. You might ask, “Can overly loud distorted levels harm my equipment?” Not really. You could harm the mic capsule if it's an extremely sensitive mic. However, most mics can handle pretty high sound pressure levels (SPL). Most people yell at an average of 82db. The world record is 129 dB. Most microphones have an SPL rating of > 130 db. 

Splitting The Microphone Signal Solution

This method uses “Multi-Tracking” to capture ultra-dynamic voice content. Two tracks are captured from a single microphone source, but each track uses an independent recording level.  For example, Track 1 can be ideal for quiet speech. Track 2 can be for loud speech. This leaves you with twice as much footage. However, it can be captured in one cohesive performance. The question is, how do we accomplish this? How exactly do we split the signal WITHOUT losing quality or causing other sound problems? It’s common to think it can be done with a simple, inexpensive cable splitter. However, BEWARE! A simple Y-cable may function, but it can open the door to potential phantom power issues and cause microphone impedance problems. You may notice a small change in your microphone's tonal character or transient response, and an increased noise floor might appear.

Externam Mic Preamp Solution.

A similar, better approach is to use an additional external microphone preamp and then split the line-level signal output to separate tracks on your interface. This can be achieved with a simple Y-cable to split the line-level signal (not the mic signal leval). External preamps can be expensive, but there is another, more affordable option that can yield pro-quality sound.

Microphone Splitter Solution

The next best solution might be a passive splitter like the one I have here. It uses a transformer to generate a second isolated mic-level output. Sound quality is generally related to the quality of the transformer. My first choice for a passive balanced mic splitter is the Radial Pro MS2. The onboard high-quality transformer delivers a great sound. It’s a great way to accomplish this for less than the cost of a legit external preamp. 

The link to this unit is in the description.

Other Methods

You may run across several other ways to split your mic signal to a second track. However, options like software loopback patching or physical loopback patching can have some unintended consequences. Your two tracks aren’t truly independent. You can experience noise, distortion, or even feedback in some situations using this method. You might ask, “Where does 32-bit float fall in all this?” The approach is slightly different if you want to take advantage of the 32-bit Float format to capture ultra-dynamic voice recordings. You would first ensure, 100%, that your loudest peak never goes above 0 dB.  Then, even though your quiet recording appears much lower, 32Bit FLOAT allows you to use digital gain to raise the softer speech sections without losing any of the resolution. 32bit Float will enable you to retain the dynamic range and resolution even after making substantial digital gain changes after you record.