Spectral Dynamic Processing for Guitars
By Frank Gryner (as published in Recording Magazine)
Up until the last decade, spectral dynamic processing or multi-band compression was largely reserved for mastering and broadcast applications. The concept of allowing an engineer individual dynamic control over the separate sectors of the frequency spectrum is often looked upon as a corrective measure for an entire mix and less as a recording or mixing staple for separate instruments in the studio. It doesn’t help that multi-band limiting often gets implicated for crimes against audio in the ongoing Loudness War, and that condescending recording know-it-alls like to claim that multi-band compression is too much power for our inferior minds to implement practically or tastefully. Yes, there seems to be a lot of polarizing opinions on the use of multi-band compressors. Like a lot of tools out there, they can be misused and wreck a perfectly good recording. I contend that they can also be a cure for the most disorderly audio and be instrumental in sculpting individual elements in unconventional ways. This month we’ll dive into one aspect of multi-band compression as it relates to correcting anomalies in your guitar tracks.
Multi-Band Aid
Guitars can be a sonic challenge to record. How it comes out of the amp or sound hole doesn’t always translate through the microphone and into your session. Inconsistencies are often exaggerated, whether it is boomy palm muted rhythm chugs or piercing string squeaks in between chords, under the recording microscope, guitars are usually riddled with problematic anomalies that need special treatment.
I remember my eighth grade teacher would threaten to punish the whole class for the actions of one or two kids as a ploy to get everyone to turn on the guilty party and give me…uh, I mean…them up. This strategy doesn’t work with audio and certainly not when dealing with guitars. Using an eq to globally fix the issue would compromise the overall sound of the guitar including places in the song where that frequency isn’t a problem. I propose a more surgical strike on the offending frequencies, separating them from the group, and applying appropriate discipline in the form of corrective compression.
Sawzall for Audio
There are a lot of multiband compression options out there in both hardware and software. I’ll focus on plug-ins since it’s much easier to get free trial versions for them than it is to have Drawmer orTube Tech let you test drive a unit for 14 days. For the most part, software offers more flexibility to manipulate your source anyway. Waves C4, McDSP MC2000 and UAD Precision Multiband are just a few of the decent ones out there. Basically we’re going to be slicing up the frequency spectrum into 3 different pieces (low, mid, and high). Where the sonic problems occur will determine where the crossover points are that separate the segments and what kind of compression—if any-- we apply to each segment.
Resonant Evil
A lot of times, after tracking some heavy, distorted rhythm guitars, you’ll notice that there are certain chords that have an excess of low frequency. I find that it is harder to place it in the mix without infringing on the space where the bass guitar lives. Simply rolling off or dipping anything below 180hz may solve the issue but would negatively impact the overall sound, making it sound thin and brittle everywhere else. We can eliminate this trade-off by taking the mid band on the plug in and defining the crossover points to surround this boominess from 120hz-300hz. You can isolate the problem and set the compressor to only apply gain reduction to this frequency range when it crosses a threshold. I find much easier to solo the band in question to initially dial in the appropriate ratio, threshold, attack and release settings rather than to go at it blindly in context.
Because the typical result of over-compression usually is heard in higher frequencies that we’re not touching, we can be a little more heavy-handed with our ratio settings at around 8:1. A fairly slow attack and moderate release setting will work well here. Make sure that the threshold is set such that the offending anomalies trigger the compression and rest of the performance enacts no gain reduction. In this case, no gain make up is necessary. When you unsolo the band and allow it rejoin the rest of the signal, you’ll hear how much more even and contained your guitar track is with very little compromise to its spectral balance.
Now consider what other issues to which you can apply this type of frequency-targeted treatment like string squeaks and other chord specific resonance. You will have to re-adjust your compression settings appropriately in each case.
Don’t Act Like You’re Not Compressed
While multi-band compressors excel at corrective processing, they can also be incredible tone sculpting devices for some advanced guitar mixing techniques. Next month, despite objection from the self-proclaimed experts, we’ll explore the more creative side of spectral dynamic processing and breathe some new life into those old guitar tracks of yours.