The way most commercial recordings are made today is not that different than the way that we made them in the 60s through 90s. Recording studios and engineers would use multitrack recorders to successively capture the rhythm instruments (and possibly scratch vocals) followed by a number of overdubbing sessions. Once all of the instrumental and vocal parts were just as the artist and producer wanted, the tracks would be mixed down into a stereo rendition.
Signal processors were used during each step in the process. During the tracking sessions, engineers would capture tracks using compressors to create specific tonal characteristics AND to protect the track from distortion. Noise gates (the infamous Kepex) would remove the obnoxious crap that came through pedal boards and other electronic equipment. These processors would be recorded on the tracks because they were part of the “sound” that the producers and engineers wanted.
Other signal processors like the ubiquitous digital reverbs run all the time but are not locked onto tape until the very final mixdown. You can always add reverb but it’s virtually impossible to remove it later. These processors were individual pieces of equipment that populated an “outboard” rack and were looped into the signal flow using patch cords and a patch bay (a series of holes that represent the inputs and outputs of every piece of gear in the studio).
Things changed when PCM digital system began to dominate music production. The general process these days is very similar to forty years ago but the equipment and capabilities of that equipment have dramatically changed. We struggled to get 8 then 16 then 24 tracks on the ever wider analog tapes. The standard became a 2-inch 24-track machine. As a joke, I once saw an 8-inch high reel of tape at an AES event…thankfully we never got there. It must have weighed 25 pounds.
Nowadays, Pro Tools and other digital audio workstations can easily accommodate 60-100 tracks (remember Sgt. Peppers was produced on a 4-track deck)! The final mixes that are done at places like the Warner Brother rooms I visited last week routinely have 150 plus tracks…and consoles with at least that many inputs. To be accurate, sometimes a track contains only a single gunshot or horn stab…but the idea of limiting the number of tracks is a thing of the past (at least if you stay in the standard definition sample rate range…meaning 48 kHz). Moving to 96 kHz means the digital processors have to struggle with the added data. Some engineers are now recording at 96 kHz and even 192 but not many.
So has there been an attendant increase in the size of the outboard rack? Do engineers have to have dozens of individual signal processors to serve the large number of tracks? Not really. Teams of clever DSP engineers figured out that they could “model” the characteristics of any piece of analog or digital electronics in software. Thus the world of digital plug-ins was born.
Figure 1 – An amplifier modeling plug-in for Pro Tools.
One of my first experiences with modeling came when I purchased a Spider guitar amplifier by Line 6. Rather than get one sound from a Fender or Musicman amp, I got an “acoustic modeling” amp that allows you to switch from a variety of amp types. My wife loved the heavy grunge Marshall sound!
These days if you want to use an Orban parametric EQ or UREI 1176 compressor/limiter on a particular track, all you have to do is have the right add-on software plug-in and presto…your track is routed through a software processor that emulates the real thing. The patching is done “in the box”. There is no need for a physical patch bay or the physical processor. Hardware makers had to find new ways to sell their boxes.
Figure 2 – An analog 1176 compressor/limiter and a plug-in version
The world of plug-ins includes emulators for existing “classic” pieces of gear AND processors that can only be implemented in software. Designers of these software signal processors sell them to studios that have the host Pro Tools rigs to support them. And they aren’t cheap. As I said yesterday, I have many thousands of dollars worth of software licenses for my existing PT rig to support my film mixing clients. They require a digital locking mechanism that must be attached to the CPU via a USB stick.
So track counts have risen, the use of processors has dramatically increased (think vocal processors especially) and the thick music layering that is all the rage. But has the sound of the music improved? I’ll let you decide.