Recording has become ridiculously easy in the 21st century, thanks to the steady progression of technology currently available to any level and budget of practitioner. We can now produce professional quality results in a home studio that weren’t possible even 10 to 15 years ago. But as much as the technology has changed to facilitate the process for the domestic sound engineer, the principles and techniques required to produce euphonious results when recording with microphones have remained the same, and it is still paramount that one aspiring to release exemplary recordings assimilates and applies these sonic axioms.
Recording real acoustic instruments certainly seems in danger of becoming a lost art, what with the proliferation of virtual instruments (or VSTs) that can provide just about any sound desired with a digitally sampled, astonishingly true-to-life result. However, just as CGI can’t quite yet simulate the look of a live human being in a movie (Rogue One’s Tarkin and Princess Leia were, to say the least, off-putting), there aren’t any natural-sounding VSTs or simulations of acoustic guitars. Though many products have come tantalizingly close to duplicating or sampling the traditional instruments of the orchestra, the acoustic nylon and steel-stringed guitars have been as of yet resistant to those imitative efforts. Capturing the sound of these string boxes with microphones still seems like the best––if not the only––option.
One needs to have a dedicated DAW (Digital Audio Workstation) with at least a two-input interface to accomplish this. The capability of producing a stereo––or two-channeled––recording is preferable to a single mono track, as it provides a fuller sound and more options during mixing. It also helps to have some kind of compressor one can run the signals through to curb clipping and distortion; I use FMR Audio’s very affordable and highly functional RNC (or Really Nice Compressor, naturally) which, as a multiple-options-averse engineer I tend to leave on a single general setting as a limiter so as not to affect the overall timbre or quality of the instrument (Threshold: -10 dB, Ratio: anywhere between 1:1 and 3:1, fast Attack and slower Release, Make-Up/Gain at 0.00).
Then a decision has to be made as to how best to capture the sound emanating from the guitar. This usually entails one of a few options, depending on whether or not the instrument is fitted with an electronic pickup and a jack for a ¼” instrument cable. This kind of signal, run through a direct box and balanced via conversion to an XLR-cable transmission, can provide not only increased control of the output but also a pleasing clarity, especially in the upper frequencies, that can bestow an augmented range on the end result. Then a microphone can be utilized to complement the direct-line signal.
Another option is to employ a 2-mic array to capture the stereo field. It’s been said many times, but as with so much else of what has and will appear in these paragraphs it bears repeating: though it certainly does help to have high quality microphones, a good sound can be garnered from most any mic if it is placed properly. I recently began recording an album of classical guitar transcriptions of Domenico Scarlatti sonatas after mulling over microphone selection and placement for several months prior; from the beginning, my instincts urged me to put Earthworks TC20 matched-pair overheads in an x/y pattern about a foot away from the guitar’s sound hole during the initial tests. I wanted a full sound that still placed the guitar more or less dead center in the stereo field, and which would completely negate any phasing issues that can arise when recording with multiple mics. And the Earthworks didn’t disappoint, delivering a full swath of frequency information, no phasing issues (incongruent loudness or disappearance of audio due to the blending of sound waves) and a focused stereo image.
Though it is certainly a strong recommendation to experiment with microphone placement in terms of sampling the sound the mic captures in different positions and aimed at various parts of the guitar, there are some caveats. Common sense would seem to dictate placing and aiming a mic right in front of and directly at the sound hole, but this can be perilous; one needs to consider the physics of all that sound bellowing out of a substa
ntially sized wooden box through such a comparatively small aperture. Like an opera singer belting out an aria directly into a sensitive microphone, it can cause an engineer any number of headaches. One can also capture a little too much of the undesirable noises, like the scabrous scrape of fingernails or a plectrum across the guitar’s strings. An ideal solution to all of this––and the one I used in my successful initial experiments––is to place one of the mics in front of the sound hole a foot or so away but pointed in a 45 degree angle toward the bottom of the fretboard, just above where the neck connects with the aperture. This way, the sound waves hit the microphone’s diaphragm off-axis, providing the fullest signal without the unwanted noise and boominess, and you also capture the sound of the strings resonating without the concomitant din of tone production. The other microphone, placed next to and at a 90-degree angle to the first (which explains the x/y nomenclature, as it pertains to a graph’s axes), points across the sound hole and past the bridge to the lower bout of the guitar, yielding complementary tones and no phase issues.
Another approach to mic positioning, depending on the desired application (for instance, if one wanted to be able to pan the individual tracks hard left and right and not have the blended signal aggregate in the middle of the stereo field), is to perpendicularly place and point the mics directly in front of the same aforementioned locations and follow the 1:3 rule (whatever distance the mics are from the guitar, make sure the microphones themselves are at least three times that distance apart from each other) to negate phasing issues.
Once the classical guitar tracks have been properly recorded, they can then be improved via equalization (eq), or the adding and subtracting of decibels at various frequencies to make the soundscape consistent. As a general rule, it is best to use predominantly subtractive eq, especially in the lower frequencies (20 – 250 Hertz), because the guitar’s lowest pitch––6th string E––is situated at roughly 82 Hz, and leaving any frequency information below that range is not only superfluous to the instrument but can also muddy up the sound (a low shelf or high-pass filter placed around 50 Hz at -12 dB works nicely). The key is to even out the audio panorama by sanding down the edges; for instance, my guitar––a Hirade H15––tends to be a bit “wolfy” (i.e. produces a louder volume on certain notes) from G# to B below middle C, so I usually make a mild bell cut (-3 dB) in that general vicinity (100 – 250 Hz).
One can pinpoint other unwanted frequencies by narrowing the “q” (or width) of a bell-pattern eq channel, setting the gain on maximum, and sweeping it slowly through the entire audible range. If a particular frequency confuses the sound picture, there should probably be a subtraction made (I have found that the main culprits tend to reside in the lower mid and mid ranges, or 250 – 1000 Hz). However, if upon executing the eq sweep there is a particular frequency that lends a sonorous clarity to the overall sound (most often in the upper mid and/or high ranges––1,500 – 16000 Hz), perhaps this should be gently boosted with a bell or high shelf-pattern eq. Keep in mind that the end goal is to create an even sound where notes in all ranges can be heard clearly at close to the same volume.
The music industry, once comprised mostly of dedicated specialists, has transformed into an individual-based farmer’s market where most everyone has to wear multiple hats out of necessity. Many recording artists not signed to whatever labels remain have become the sole curators not only of their artistic output and refinement but often also of self-management, promotion, social media presence, booking, and yes, recording, mixing, engineering, and even mastering their own albums and singles. With more time spent on adjunct concerns and less and less on practicing, it is only natural that mistakes happen in the nerve-wracking environment of the recording studio. Faced with the daunting prospect of recording an album of difficult Scarlatti guitar transcriptions, and suffering from the same vocational schizophrenia and subsequent dearth of rehearsal described above, I am reconciled to my mistake-rife reality. So I have geared my studio performances of this challenging material not only around the inevitability of those mistakes occurring but also with the thought of what works best during the editing process in mind.
I try to maintain rhythm and dynamic feel when I err, playing the scuffed-up phrase through again directly after, as many times as needed to get it right, and with the same energy and tempo. I know I can go back later with the editing hat on and cut and reassemble the tracks properly as long as the correct performance is consistent with what was played before and after the mistake. It is also crucial that one keeps phrases in mind when making an edit; it is nigh unto impossible to make a clean cut in the midst of something sustaining, so make an acute mental note if an open string or sympathetic vibration creeps into the next phrase or chord change during an undesirable passage, as editing in the midst of these is highly problematic.
When making an edit incision, look for the start of the tracks’ waveforms and cut in the middle of the largest gap behind it, removing the unwanted material from the right side and bringing in the correct version. If it’s in the midst of a quick or sustained phrase, it is advisable to take out some of the material to the left of where the main attack occurs, just to match the faster feel of the surrounding phrases. It might take a playback or two––and some lateral movements of the replacement clip––to get the rhythm right before connecting the two segments, preferably where there is no discernible sound material on either side. Many engineers recommend overlapping and subsequently cross-fading edited tracks, but I have found that having them abut with no overlap or gap between and with a short fade-out and fade-in on the left and right sides respectively works well in most instances.
A well mixed and edited classical guitar track, with all its edges sanded down and sounding rhythmically organic and clear, is now ready for whatever reverb ambiance is required. It is important to let the music speak for itself in that regard, that whatever reverb and/or delay is implemented doesn’t affect the emotional impact of the piece. If it’s something slow and intimate like Scarlatti’s K. 208, it probably doesn’t need much, if anything. If it’s the gothic K. 32, crank the knob up on a large hall preset and watch the goosebumps rise. Just bend the knee to the music—and to the solid principles that helped successfully record it––and you won’t go wrong. If you build it right from the ground up, the listeners will come.