With the modern era of digital capabilities to re-tune recorded material, almost all music you hear nowadays has “perfect” tuning on all of the instruments and vocals. But what does this mean, exactly?
Generally, this tuning is based on equal temperament. But is equal temperament the most “perfect” tuning for absolutely all musical situations? I say… nope.
Imperfect? Or just Different?
There’s lots of styles of music wherein notes tuned differently than equal temperament provide their characteristic flavor and personality.
Singers recorded a cappella tend to naturally shift their notes towards just intonation, with its more perfectly mathematical interval ratios and cleaner, purer harmonics. But of course, it doesn’t work as well with most instruments, non-diatonic scales, and shifting keys.
Each and every note on most guitar necks will not hit perfectly to 0 cents in equal temperament, due to intonation issues… but this is one of the things i love about guitar. Beyond the timbral differences, playing the same lick or chord in two different positions, or on two different guitars, or with slightly different miro-tunings, can lead to subtle differences in the ghost web of generated harmonics, giving them slightly different emotional qualities, therefore providing an open world of additional creative options to select between.
Another example of perfect equal temperament being not ideal is for electric keyboards with metal wires, as tuning them with perfect octaves leads to unpleasant inharmonicity, which is alleviated by giving them stretch tunings, pulling the octaves a touch wider.
Just like the visual spectrum of colors is a continuous range that blends and fades between each color, rather than abruptly shifting between them, so are musical tones a range of cents values that our ears will hear as distinct notes.
If we automatically assume that making sure every single note is perfectly “in-tune” (according to A440 equal temperament) is the ideal route in all cases, we… might be wrong. Making sure that every single note conforms to a specific rung-like interval, while avoiding nearby shadings and highlights, leads to overly stiff, robotic- sounding music.
Variations from this conformity can come in the form of randomized pitch blur (using such things as vibrato, the chorus effect, or by blending multiple imperfect takes), or in the form of a deliberately different tuning standard, wherein certain notes and/or octaves are tuned to different targets than to the typical equalized intervals. Using such variations can sound exotic, weird, downright bad, or quite interesting.
The soundcloud example below demonstrates the Nahwand scale, which actually has different tunings when it ascends vs when it descends, so i composed a line that highlights this. I think it has a mystical, ritualistic sort of sound. If you’re listening with western ears, some of the notes may feel like they are “off-tune”, but if you zen out with it for a while, you can change your headspace and groove with the feel.
Some songs, in particular orchestral string instruments and a capella vocals, might modulate their specific intonation as they progress, using different cents values for particular notes based on which harmonic relationships are present in the chord they reside within at any given moment. For example, a Major 3rd in equal temperament has a ratio of 1.25992, while a Major 3rd in just intonation has a ratio of 1.2500 (exactly 5/4)… which a string section or a choir of human voices matching together are more likely to gravitate towards naturally.
So, for example, a C note might have a certain cents value for most of a piece, but later on the key modulates and that C becomes the major third instead of the root; at that point the alternate, more harmonically pure but less rigidly rung-like cents value might naturally take over. Guitarists, though they have fixed fret intervals, can accomplish this to a degree by slightly bending notes as they play (something most guitarists have gotten used to doing occasionally as damage control at some point when halfway through a song live a string goes a bit flat). Another trick you’ll sometimes see is bending the neck to add vibrato to an entire chord.
You can certainly experiment using these techniques with non-acoustic music as well, for example by tweaking the cents value of only particular notes, either permanently across the track, or only during select chords or riffs. Many synth instruments have cents adjustment built-in, and there are various micronotonal tuning Max for Live effects you could try out. If automating a microtonal pitch shift doesn’t work due to overlapping notes in chords (or whatever reason), you might resort to sneaky tactics like duplicating instruments and splitting the midi clips for various differing cents values. For example, say you wanted standard tuning except all E and B♭ notes in a line set to +17 cents. You might need to copy the instrument, and in the original, deactivate all E and B♭ notes. In the duplicated instrument, set the tuning to +17 cents and deactivate all notes except for E and B♭. After this, i would group the tracks and consider them as one instrument, moving existing and applying subsequent effects to the group.
Teraser: we are working on an Ableton pack that incorporates a vast collection of both equal tempered and microtonal tunings (such as the Nahwand tuning from the example above) that you can drop on to synth instruments.
Nonstandard tunings can sound weird at first when we’re not used to them, but denying them as a whole reduces the sum variety of adventure we can experience as listeners and crafters. And that doesn’t sound fun.
Micro-Fluctuations for Modulatory Girth
When recording acoustic instruments, there will be minuscule tuning variations, and often these themselves give the music a certain magic and charm. I’m not talking about unpleasantly off-tune bad singing and badly tuned guitars, but rather about natural subtle variations in tuning from note to note on the order of single digit cents. These microtonal variations will result in a natural “chorusing” which can be very nice… for example, two notes played by guitarists, one on the left and one on the right, are unlikely to come out at the exact same cents value, and this tiny difference will enhance the stereo width effect. Chorusing, vibrato, etc. added deliberately as effects can also serve to “smear” the target pitch standard, helping to both avoid an overly rigid sense of stiffness and also to help “disguise” notes that are slightly off tune.
This can be an appeal factor of some types of analog synths: since their oscillators are analog, there is semi-random fluctuation going on with the tuning (particularly just after being switched on) which some people covet for its unpredictability and natural warble.
In a vocal chorale, the micro-variations between each singer can make the result sound much more rich and full than if every singer was hitting every note absolutely identically. Say you’ve got a multi-miked group of backing vocal tracks to mix… should you meticulously tune each one so every singer is hitting every note perfectly at +/- 0 cents in equal temperament? Fuck no. Doing so would strip away the natural harmonic contour the singers were themselves feeling and generating by reacting to each other in realtime, leading to a hollow, lifeless inhumanity. For unobtrusive, transparent, classy processing, save the tuning adjustments for those occasional off notes which stand out unpleasantly to the ear.
Then we also have microtonal tunings which deliberately use non-standard notes, or experimental music based on mathematical tuning algorithms which defy pitch convention altogether.
I recently experimented with a 5-part harmony layer stack for a melody line based on harmonic tuning, with each layer tuned to its corresponding harmonic interval’s pitch offset in cents (and also successively dropped in level). Some of the layers sound “off-tune” technically when played in isolation, but when combined with the underlying melody and chords, it all sounds surprisingly harmonious in an exotic, almost mystical manner. The idea is that those higher lines are hitting harmonic tones that we are already subtly hearing as the instrument timbre is generating them—hearing them not so much as distinct notes but rather as part of the overall quality. This experiment to me was enlightening and i plan to incorporate some of these types of tunings into my upcoming album.
I snipped some of the parts from the song i was just talking about, so you can hear what i’m talking about.
First plays the main rhythm guitar by itself, to provide the context of the chord progression. Next, the harmelody stack plays along with a bassline to keep a grounded sense of key. Finally, the chords and melody are combined together so you can hear how they mesh.
What’s interesting to me about it is that the harmelody stack sounds more in-tune when it’s played along with the rhythm guitar than it does by its lonesome.
Sometimes, a weird note that’s not even in a scale can happen to work for particular moment. I remember this one time mixing a song with Charlie Milo and we were fixing some off-notes from a recording. There was this one note that he wanted exactly halfway between an E and an F. I was like “what!??!?”. Turns out his ear did not lie and it happened to work perfectly for the song.
So should you tune your vocals to aim for absolutely perfect equal temperament concert pitch at all times?
Maybe if you’ve making a shiny bubble gum pop gem for crystal androids from a homogenized and boring utopia, but otherwise… i say nah. Instead, record lots of takes, listen to them each carefully in context, and make a comp out of the select moments that make your spine the most tingly on a visceral level, whether or not they are “technically perfect”.
Some Vocal Tuning Tips
•if you don’t own any pitch correction software, you can still correct small notes that are slightly off in Ableton Live. This is actually the most high-quality and natural pitch correction available. Split the note you need to tune (select that area and press CMD/CTRL-E), make sure the clip is not warped, and adjust the detune value in the clip view sample controls. Make sure the split section fades smoothly to connect. If you go down in pitch, the sample becomes longer; if you go up in pitch, it gets shorter, so you may need to adjust the timing backward or forward if need be to situate it ideally. Alternately, you could have the split clip section warped so its timing doesn’t change, but that comes at the cost of signal degradation according to the warp mode, so choose wisely.
•if you do use pitch correction software, avoid the urge to correct things based on your eyes. Use your ears because sometimes notes look off in wavestune’s or melodyne’s or whatever’s display, but they are actually at the pitch they should be for the song. When you fix a note, try to avoid quantized pitch jumps, keeping smooth adjustment curves to retain a natural sound.
•if you can’t manage to keep a fixed note sounding 100% human-sounding but it’s too far off-tune to not fix, consider dropping it in level until it’s barely inaudible. This is more a trick for layered backing vocals; if your lead vocal is that much off, it’s probably best to punch in a re-record on that section instead.