Old School Vs. Nu Skool: Was the “classic” way to record better?

By Craig Anderton | Thu, 01 Oct 2009

There’s a fascination with everything vintage— from tubes, to optical compressors, to analog tape. Yet there’s also no question that digital audio technology has come of age. Digital has always offered convenience, but now the sound quality of even budget digital systems has increased dramatically since those little ones and zeroes first entered our lives.

With vinyl making a comeback (!), and many musicians having the nagging feeling that maybe those classic records held a production or gear secret that we’ve lost along the UAD-2 Solo way, let’s take a look at how recordings used to be made, how they’re made now, where there may be differences, and how we can reconcile any significant differences to combine the best of the old and new. We’ll also review some products (all prices are MSRP) that “cross over” the old school/nu skool worlds.

OUTBOARD SIGNAL PROCESSING

Old School. As consoles typically had limited processing (EQ and sometimes dynamics control), engineers relied on racks of outboard gear for processing. Some producers and engineers even had that rack gear in flight cases, so they could ship their “trademark” processors to sessions anywhere.

Nu Skool. Native plug-ins are actually more like a variation on the concept of building processors into your console— just as console processing was constrained by size and cost, native plug-ins are limited by computer power. A more accurate Nu Skool analogy would be the hardware processing options on cards that go into your computer, or in outboard boxes that connect to your computer via FireWire. Examples include the Sonic Core/Creamware cards, early E-mu interfaces, TC PowerCore, UA Powered Plug-Ins, and SSL’s Duende system.

Nu Skool problems/solutions. Hardware gear had real knobs, and could be “played” for more creative processing. Although plug-ins, whether native or DSP-assisted, don’t have controls, the industry recognizes this limitation and is addressing it. Native Instruments’ Kore provides a slick hardware controller for programming not only NI processors and instruments, but ones from other companies. Novation’s AutoMap protocol maps plug-in and effects parameters to their line of hardware controllers, and Propellerhead Software’s ReMote protocol provides “hooks” for hardware to control Reason. Cakewalk’s ACT takes an approach that’s similar to Kore and Automap, as it exposes all VST parameters for control via control surfaces (including their own VS- 700C control surface).

Another issue is portability. With dongle-based processors, you can install software processors in multiple machines, then use your dongle to bring them to life. However, perhaps the ultimate example of this is Universal Audio’s UAD 2 Solo/Laptop (see review below) as it can “link” to UA plug-ins installed on your “home” computer, obviating the need to buy the same plug-ins again for your laptop.

REVERB

Old School. Pro studios used an actual acoustic space or plate reverb, which consisted of a large metal plate (about 1 x 2 yards), with driver transducers at one end and pickup transducers at the other end. Although a plate reverb had only one basic sound, it was very natural, and you could damp the plate to shorten the reverb time.

Concrete, reflective rooms were favored as acoustic spaces. Probably the most famous acoustic reverb setup is the eight underground reverberant spaces built 30 feet underground at the Capitol Records building in Los Angeles, as part of a studio designed by Les Paul. Other famous spaces include the hallway at Headley Grange, where the drums were set up for Led Zeppelin’s “When the Levee Breaks,” known for its Olympian drum sound.

Nu Skool. Digital reverbs have taken over project studios, whether as outboard gear like Lexicon’s hardware reverbs, or software plug-ins. Digital reverbs sub-divide into synthesis and convolution types. Synthesis reverbs use algorithms that synthesize a room, and include many adjustable parameters— room size, reverb density, etc. Convolution reverbs, like Audio Ease’s Altiverb, capture a “sample” of a room’s characteristics and apply that to a signal. Although CPUintensive and with fewer adjustable parameters than algorithmic types, convolution reverbs provide sonic accuracy.

Nu Skool problems/solutions. Sorry, but the complexity and richness of early reflections in an acoustic space is something that today’s technology still can’t duplicate. However, you can make a digital reverb sound more like the desirable old school sound by combining it with a real space—bathroom, garage, or other room with hard surfaces. Set up a loudspeaker and a mic, send the reverb aux send to the speaker, pick it up with the mic, and mix this with your digital reverb (getting the balance right) with the digital reverb’s “tail.”

Lacking that, use a multi-tap delay to simulate additional early reflections. I set the taps for short, mostly prime number delays to avoid harmonic buildups (e.g., 3, 5, 7, 11, and 13ms) and a few longer delays (21, 23, and 25ms). Try a little feedback, but keep it moderate (around 20%) and pan the reflections around the stereo field to “open up” the sound. Mix this in with your main reverb.

A final option: Patch different reverbs in parallel or series. Use different reverb algorithms, but set for approximately the same delay time. One reverb tends to “fill in the cracks” of the other reverb, creating a smoother, more natural sound.

MASTERING

Old School. Veteran mastering engineers used expensive, complex analog equipment. Much of their work involved accommodating the limitations of vinyl and 8-track or cassette tape; low frequencies were often attenuated to keep a record player’s stylus from jumping out of the groove, and there was always a tradeoff between album length, the music’s volume, and audibility of surface noise.

Nu Skool. Analog signal paths are still used, but the requirements of mastering for digital audio are much more relaxed than for vinyl. Computer-based digital audio editing programs do a lot of mastering these days, using plug-ins to provide signal processing.

Nu Skool problems/solutions. Proper acoustics are essential for mastering, and home studios may not have the same rigorous treatment as old school mastering suites. However, there are now many relatively inexpensive tools to improve acoustics, such as bass traps and diffusors from companies like Auralex, Primacoustic, Real Traps, HFS Acoustics, etc. When applied properly, and used with near-field monitors, room acoustics can approach being a non-issue.

Also consider the sound quality of the plug-ins themselves. Some are designed for use in multitrack DAWs where there may be many instances, and therefore, trade off audio quality for lower CPU power. Often this is not a problem, as processing an individual track is generally less critical than working with complex program material.

However, some companies make “CPU consumption be damned” plug-ins intended for mastering-level applications. Examples include plug-ins from Waves, IK Multimedia’s TRackS, Cakewalk’s LP series processors (bundled with Sonar), mastering-oriented plug-ins bundled with Magix Samplitude and Sequoia, PSP Audioware’s superb MasterQ and MasterComp, the URS line of plug-ins, BIAS Master Perfection Suite, McDSP ML4000HD, iZotope’s Ozone 4 mastering suite, encoders from Minnetonka, Steinberg’s bundled plug-ins for Wavelab, as well as plug-ins from Sony, Sonnox, WaveArts, TC Electronic, and others.

A good acoustic space, accurate reference speakers, and a collection of quality plug-ins can do decent “project mastering”— but only if you have the ears. Otherwise, use the old school approach of taking your precious project to a veteran mastering engineer with a good track record.

COMPOSITE RECORDING

Old School. In the days of limited tape recorder track counts, composite recording (i.e., recording multiple takes, then picking the best parts of each one to create a “perfect” take) wasn’t an option, so edits had to be punched in. Punching was a problem because the engineer had to make sure nothing was erased accidentally, and artists didn’t like interrupting the flow of a take.

As the number of tracks increased, you could record multiple takes and pick the best parts. In tape’s twilight, recorders could rewind automatically at the end of a take back to the beginning, and while rewinding, the engineer could record-enable the next track.

Nu Skool. DAWs now take advantage of unlimited track counts to include composite recording as a feature. Of the various DAWs, Apple’s Logic 9 (Figure 1) has arguably the most advanced comping options (especially combined with the new Flex Time feature) because of how the program simplifies the process of editing all the good bits together.

Nu Skool problems/solutions. Stitching together disparate parts under any circumstances can lead to a performance that doesn’t “feel” quite right because it doesn’t flow. One solution: Limit the number of takes used for composite recording. If you can’t get a good vocal in less than a 8–12 takes, re-think the vocals—or change the vocalist! Fewer takes also means less editing time spent deciding which parts to keep. Also, use the longest sections possible. Compositing two or three long phrases will generally produce a better-sounding vocal than doing fixes on a word-by-word basis.

Another issue is that the tape’s rewind time gave any vocalists a chance to relax for a few seconds, and get ready for the next take. With DAWs, rewinding is instant, which causes some vocalists to feel rushed. So, consider a long pre-roll time before a take starts recording. This gives the artist a chance to get into the feel of the track, and can lead to a more relaxed performance.

MIXING/INTERFACING

Old School. Mixing involves some of the biggest differences between old school and nu skool techniques. Mixing used to be done manually, and with large consoles, often involved multiple people assigned to different faders. Automation was rare and costly; interfacing was only about having mixer inputs to accept signal sources, and outputs to send to your multitrack tape recorder.

Nu Skool. Console automation is on the endangered list, as host DAWs invariably include automation within the program. In fact many musicians spring for control surfaces (which don’t pass audio) for their DAWs instead of using mixers; these may include moving faders, thanks to ever-declining prices.

In fact, I’m starting to think we need a new term to describe boxes that are part mixer, part control surface, and part interface—or some combination of those three. At the other extreme, some people mix with a mouse, a track at a time, and don’t even use control surfaces.

Nu Skool problems/solutions. With old school mixing, the mix was more like an extension of performance, with faders, EQ, and switches being played in real time. That’s impossible to do if mixing with a mouse, or with control surfaces that have a limited number of faders (i.e., less than one per track). One option is to add additional “sidecar” surfaces if the design allows (e.g., you can expand the number of channels with the Euphonix Artist Series), and another is to “perform” on the most important parts while leaving the others at a constant level, then go back and automate background vocals, percussion parts, and the like.

SESSION MUSICIANS

Old School. You’d hire the musicians, and pay them based on union pay scales. The most important factor was finding a pro who could not only play what you wanted, but also put some emotion into the part—not just hit the notes and go home.

Some session musicians were so proficient they played the parts in recordings that were normally played by band members onstage, and there were also élite session musicians, like drummer Hal Blaine, bassist Carol Kaye, Motown’s mighty James Jamerson (bass), and too many others to mention—even Jimmy Page got his start as a session player.

Nu Skool. You can still hire session musicians, but now there are two more options: Online services where you send in files, musicians play on them for a fee, and then the files come back to you with the parts; and loop/sample libraries.

Of the online services, eSession is well-regarded, and serves as a clearinghouse for many musicians who have a long résumé of session work. However, a Google search for online+session+musicians brings forth a variety of services, including relatively well-established ones like Studio Pros and Live Studio Drums, as well as specialized options like String Section.

Loop and sample libraries provide “virtual session musicians,” but many drum programs include patterns as well—probably the most “session musician”- oriented one is DrumCore, which has libraries from drummers like Alan White (Yes) and Matt Sorum (Guns ’n’ Rose, Velvet Revolver, etc.). Loop libraries are an economical option; some DVD-ROMs offer thousands of loops for under $100, and you can also buy loops for bass, keyboards, and even ethnic instruments.

Nu Skool problems/solutions. The obvious limitation is no face-to-face interaction—opportunities for giveand- take are limited with online sessions, and non-existent with sample libraries. Some websites are more into back-and-forth and approvals, but still, those genius interactive moments where a song goes off into an entirely new direction are difficult to pull off in cyberspace.

Loop libraries are even more rigid, but there are some ways to make them more malleable. For example, many drum libraries include individual hits of the drums used to create the library, so you can add parts to augment an existing loop, or even make custom loops. You can also cut up loops to customize the part somewhat.

Sony has taken a novel approach with their Artist Integrated line of four CDs (drums, bass, keys, and guitar). The drum loop library by Siggi Baldursson was recorded first, then bassist Tony Franklin recorded bass grooves on top of the drums, and the bass parts became available as a separate library. Former Prince keyboardist Matt Fink added keyboards, and Parthenon Huxley did a library of guitar parts. Although each library stands on its own, they also combine to form more of a “groove.”

There’s an enormous variety of sound libraries available—whether you need orchestral strings, a jazz saxophone part, or even Bollywood-style percussion, you’re covered.