Professional Analog Mastering.
Professional Analog Mastering.
226K+ subscribers
Breaking up a signal’s frequency response is pretty common at this point. Most of us are familiar with the bass range, the low, mid-range, and so on. However, the dynamic range is rarely conceptualized in this way, which is a shame since understanding and controlling these ranges is just as important as controlling the frequency response - so let’s change that today, as we
Let’s start by organizing and conceptualizing the dynamic range.
Although the overall dynamic range could arguably be broken up into more ranges, let’s start with 4.
From lowest to highest in amplitude, we have our noise floor, which is comprised of the noise generated from our recording equipment, such as a microphone and amplifier’s self-noise, and quantization noise or error, which is the difference between the original analog signal’s amplitude and the digital or quantized signal’s amplitude.
With today’s equipment, this noise is typically imperceivable, especially if no processing has been introduced to amplify it yet.
Above this, we have what often gets masked, or what I call a signal’s low-level details. This might be the tail end of a reverb’s reflections, or a tiny buzz in an acoustic guitar’s string - basically anything that’s difficult to hear and often covered up by higher amplitude signals.
Above this we have the tonal aspects of an instrument - so, the ring or sustain of a snare, or maybe the sustain and release of a piano note or chord. These tonal elements are the biggest contributors to the musical aspects of a recording - it’s how we perceive important musical building blocks like notes, chords, melodies, harmonies, and so on.
Lastly, we have our high amplitude signals, like the onset of a note, the transient, the attack - basically anything recorded right when the instrument or signal begins. Going back to our snare example, think of this as the initial hit of the snare. Or, with our piano example, it’s right when the hammer hits the string.
In other words, it’s the signal’s peak, and it’s what we typically control when we use a compressor.
Now, these ranges can be controlled with traditional dynamics processors, but let’s use this Dynamic Grading plugin since it gives us a lot more control over each one of these ranges and is really helpful for me in showing these ideas.
Let’s take a listen to it, as I adjust the levels to generally amplify and attenuate these ranges.
Watch the video to learn more >
As I said earlier, compressors are the most common way to affect dynamics, especially peak-down compressors. By attenuating peaks, these processors increase headroom, or the distance between the peak level and the ceiling or cutoff for the allowed amplitude in a digital system.
Controlling peaks is more or less a stepping stone to increasing lower-level details - since we increase the headroom, we can use post-compression make-up gain to increase quieter details, and help them become easier to perceive.
RMS compression is a form of downward compression, but instead of being triggered by peaks, the compressor is triggered when the average loudness of a signal crosses the threshold. This isn’t the best for controlling peaks, but it is a good way to control the tonal aspects of a signal.
Upward Compression is similar to waveshaping - it’s the amplification of quieter details while leaving peaks alone. In this regard, upward compression skips peak attenuation and gets right to amplifying low-level signals.
Using a processor to amplify the noise floor isn’t very common - instead, downward expansion or gating is used to attenuate the noise floor whenever a signal becomes low enough in amplitude or falls below the threshold.
Lastly, we have upward expansion, which amplifies the peaks. This is typically used to expand transients, like if we want the snare to cut through a complex mix. Amplification to the peak occurs once it crosses a processor’s determined threshold.
The Dynamic Grading plugin I was using earlier is a combination of all of these processors, with the exception of RMS compression.
The signal’s dynamics are broken into controllable bands, similar to how an EQ can break up the frequency response.
The left side is the input, and the right is the output.
So, if I want to attenuate the peaks, I’d isolate the range or have the plugin find the range, and then drag down the bars on the right to reduce the dynamic range of this region.
Or, if I want to amplify quieter details, I would drag the range’s top marker on the right up to expand the range, or drag both markers up to amplify the full range, or a combination of the 2.
Let’s work on a drum track - I’ll attempt to do 4 things. I’ll first amplify the noise floor until it’s noticeable, before downward expanding or gating it.
Then, I’ll amplify low level signals to achieve more detail from the drum bus. Then, I’ll compress then amplify the tonal aspects to increase the sustain and decay of the kit, before slightly controlling the peaks.
Let’s listen, and notice the role that each range plays, and how they come together to create our overall perception of the drums.
Watch the video to learn more >
The dynamic ranges we’ve discussed could easily be seen as separate from the frequency response, but the 2 concepts are inseparable.
Remember earlier when I mentioned that tonal aspects play the biggest role in our perception of musical building blocks? Well, the simplest of these building blocks is a note - which is a specific fundamental frequency, with overtones or multiples of this fundamental frequency, also called harmonics.
So, say the sustain and release of a piano note is 440Hz and a series of harmonics - this would be the note A4. If I took this dynamic grading plugin and amplified the tonal range, I’d amplify these frequencies. Similarly, if I reduced the range, I’d attenuate these frequencies.
Following this idea, say we have a snare tuned to A3, or its fundamental frequency is 220Hz. The peak or attack of the snare will definitely have some of A3, but it will also have frequencies that don’t relate to the fundamental. The point being, the attack and peak of an instrument can have a different frequency response than the sustain and release or the tonal range.
By altering one and not the other, or altering both but in different ways, we both control dynamics and equalize the signal.
Let’s take a listen to both a snare and a piano note being affected with this dynamic grading plugin - notice how the frequency response changes as the dynamic ranges are altered.