This is a very important tutorial for beginners in audio recording and mixing. Gain staging is a technique to maximize signal clarity and minimizing noise or distortion. So how you are going to achieved this in your recording and mixing projects?
Tip #1: Always record at 24-bits to MAXIMIZE signal to noise ratio
The higher the signal to noise ratio, the better will be the clarity/fidelity of the recorded signal and the lesser will be the noise. By recording at 24-bits, you maximize the capacity of your audio interface (provided it supports 24-bit recording) when it comes to signal to noise ratio. Since a pure 32-bit audio interface still does not exist commercially as of 2011; then always use 24-bit in your recordings. If you like to know the details why a 24-bit recording is more accurate than 16-bit recording, etc; then read this pulse code modulation tutorial.
In some digital audio workstation software, there is no 24-bit recording option but only 32-bit float. In this post on 32-bit float recording depth vs. 24-bit, it shows that 32-bit float has 24-bits accuracy. If your recording software does not support 24-bit recording, then you can select 32-bit float recording.
The theoretical dynamic range of 24-bit recording is around 144dB while for 16-bit is only 96dB. Dynamic range is the maximum difference between the quietest signal possible in your recording and the loudest signal. However there is no such thing as a perfect recording with absolutely zero noise. In practical recordings, there is always a noise floor and a recorded signal level. The ratio of the noise floor to the recorded signal is the signal to noise ratio. To easily understand these concepts; take a look at the chart below:
Dynamic range 24-bit chart
In the above chart, it depicts a typical 24-bit recording. The dynamic range is around 144dB and the noise floor is around 120dB. Therefore the effective dynamic range or signal to noise ratio is around 120dB. Human ear dynamic range is around 120dB so if the noise floor is around at 120dB, you have optimal signal to noise ratio for a 24-bit recording. It is because the recorded signal has a LOT of room to occupy and expand (120dB range); making the sound clear, transparent and true to the original sound source.
If the recording is made at 16-bit; the effective signal to noise ratio after noise would be around 80dB to 90dB which is substantially smaller than 24-bit recordings. The resulting recording is not as clear and “true to the source” as compared with the 24-bit digital signals. Also since 80dB to 90dB is less than 120dB human ear dynamic range, you are not optimizing the signal to noise to ratio and dynamic range that the human ear is capable of listening. To get a VERY LOW noise floor, some suggestions:
1.) Use high-end and professional recording gears as they have very lowered noise output compared to consumer devices. This applies to your audio recording interface, mixer, etc. Research about the noise floor contribution of the device before you will purchase them.
2.) Use quality microphones. For example for vocal recording, you might be using high-end vocal condenser microphones than simple dynamic microphones because they have higher dynamic range output and lower noise floor.
3.) Use balanced inputs and outputs as possible. Read this important tutorial on balanced and unbalanced audio guide in home recording to learn some theory as well as the best practices during recording. Identify the common causes of noise in recordings and correct them.
Tip #2: Avoid distortion by leaving some headroom
Now that you know how to deal with the noise floor and getting great S/N ratio; the next step is to leave some headroom to avoid the possibility of distortion. In recording, this is around -18dBFS to -6dBFS maximum. See below:
Headroom application
When you are recording, you should leave a maximum of -6dBFS headroom. This will decrease the chances of signal distortion and clipping; an essential element in gain staging. However setting at -6dBFS probably is too much, most of actual recordings would settle at around -42dBFS to -18dBFS with -18dBFS as the most common peak zones.
In gain staging, you should maximize the recorded signal in such a way it is within those peak zones. The recommended is around -18dBFS to -6dBFS maximum or when you get a good green to slight orange (but never red) in your audio interface recording meters. This will give you the best signal to noise ratio particularly if you are recording at 24-bits.
Adjusting for good recording signal is a trial and error process with your audio interface preamp gain. For proper gain staging, start at 50% gain applied and check its response with respect to headroom. Never assume you get a good recorded signal at full gain adjustment of preamp because you will end up with no headroom and increase the possibility of distortion and clipping.
For example, below is a screenshot of a Saffire Pro 40 audio interface level meters with peak level at -6dB.
peak level during recording Saffire Pro 40
Tip#3: Limit -3dBFS in master fader level out in mixing and every channel
This tip only applies when you are finished with your recording and you are now about to mix the recorded tracks. You are now ready to mix and adjust your faders. Every channel or track in the mix has its assigned fader and level meters. And all of these are summed up in the master channel which also has its own level meters (known as the master level meters).
For proper gain staging, you should not clip either of these channels; leave around -3dBFS maximum. See screenshot below (for example if you are using Reaper DAW):
Proper gain staging guide in Reaper
The yellow box is the master fader and its level meters, while the red box are the tracks fader and its corresponding level meters. To avoid distortion and get the best proper gain staging, observe a -3dBFS maximum in all channels. The following are some suggestions on how to adjust:
1.) If one track sounds very weak and buried in the mix, do not further increase its fader volume. Instead pull down the volume of the other tracks until that problematic weak track will now sit properly in the mix. This technique reduces the possibility of clipping for each channel and for the master fader output.
2.) If the master fader clips beyond 0dBFS. You can either reduce the fader on each channel by the same amount (e.g. if you reduce -1dBFS in channel 1, you reduce -1dBFS also in the other tracks). This will make the leveling assignment on each track unaltered while you effectively lower the master fader output volume to less than -3dBFS. However if you are sure nothing in each channel is clipping, you can safely reduce the master fader volume to less -3dBFS directly without adjusting the volume in each channel.
3.) To prevent ear damage while doing gain staging in the mix; monitor at moderate volume.
4.) During mix down, it is recommended that the peaks of the wave should be around -6dBFS to -3dBFS. This leaves a lot of headroom for mastering engineers to work with. For more details, you can read this tutorial on correct audio mixing levels and headroom in preparation for mastering.
Content last updated on June 20, 2012