Does it really make difference to play a song in a different key?

Question

I hear sometimes players and composers say that this song won't sound good in F Major for example, and you should play it or write it in G Major (this was just an example). In other words, it seems that they assign aesthetic attributes to keys.

My question is that, imagine a baby growing up in an isolated island, and hear all the songs of his entire life in F Major (all re-written for F Major), while another baby grow up in another isolated island and hear all the songs of his entire life in G Major. Now, is there an objective, real difference in the amount of pleasure they can get from same songs written for different keys?

This has really obsessed my mind, as I'm not sure if great composers really feel difference in songs being played in different keys. Did Beethoven for example felt Moonlight more sad in key X rather than being played in key Y?

Answer 1

Music, as an art, is in the ear of the listener. As a musician, I can say there are definitely times when a song sounds "better" in one key than another.

The primary reason this is so is when the key fits the "natural" range of a singer or instrument. A song may sound perfect when sung by a female alto, but as those notes sung verbatim would be at the top of a male vocalist's range or, if transposed, the very bottom, it would very likely sound "off" (regardless of subject matter) if sung by a male in the original key, because you will hear the strain in his voice, or the timbre changes as he moves from his chest voice into either his head voice or to pedal tones. Transposed down a fifth into a new key, the song would sound much better when sung by a male singer. The same applies in reverse.

In other cases, such as stringed instruments like guitars, there is a mixture of convenience and of chord voicing. Guitars, because of the tuning of their open strings, have a few "natural" keys, like G and D, in which the most common chords heard in a song of that key (I,ii,iii,IV,V,vi) are easy to play. Also, because of the tuning intervals between strings and the subsequent natural spacing of notes in guitar chords, the playing of a song in a different key often necessitates the use of chord fingerings which produce alternate voicings; a different note in the chord will be on bottom and on top, and different notes will be next to each other. This can definitely make a song sound better or worse depending on key.

Even in cases of instruments like pianos, where theoretically the instrument has the same timbre across the entire instrument and the same chord voicing can be played just as easily in any key, there are changes in timbre as the pitch changes. This is due to differences in construction across the same instrument (a piano for example goes from one to two and three wound strings, and then to three monofilament strings, for each key of the keyboard as you move from left to right) and due to simple physics (for instance, as notes increase in pitch, we hear less of the harmonic overtones produced as they pass beyond 40kHz, and conversely, as pitch decreases and heads toward the subsonic range, the frequency difference in notes and thus our ability to perceive said difference diminishes). This may lead to a song sounding better in a lower (or higher) key for reasons that may not be explainable at the time.

Answer 2

In modern Western music, we use equal temperament where all keys are basically equivalent. Notes are based on 2 ^n/12. Using A440 as a base, you get the following:

• A = 440 Hz * 2 ^0/12 = 440 Hz
• B♭ = 440 Hz * 2 ^1/12 = ~466 Hz
• B = 440 Hz * 2 ^2/12 = ~494 Hz
• etc.

Historically this was not the case, however. Just intonation ruled the world, where notes are based on ratios:

• A = 440 Hz * 1/1 = 440 Hz
• B♭ = 440 Hz * 16/15 = ~469 Hz
• B = 440 Hz * 9/8 = 495 Hz

So the short story is that different keys are / were fundamentally different due to the application of ratios not preserving the exact frequency values of notes. This is apart from the the sound of higher/lower frequencies, which of course affect the aesthetics as well. If you want all the details, we have several related questions with great answers:

• What's the difference between a G♭ and an F#?
• Why are there twelve notes in an octave?
• (On Math.SE) Mathematical difference between white and black notes in a piano?

And on this question there are actually examples of songs played in different temperaments:

• Examples of songs or phrases played in different temperaments

Answer 3

Assuming that the intonation and tambre of the instruments/voices were maintained, there is one main difference that is dependent on the way you listen to music. People with strong relative pitch listen to a piece as a progression of intervals. The arpeggio C-E-G would be heard as M3-m3. This is a very effective way to think of music, but is also not the only way. People who can identify individual notes more strongly, or have perfect pitch, listen to this arpeggio just as it is played: C-E-G. When the arpeggio is changed to another key, like F-A-C, there is a more defined difference because the notes are changed. If you are more dependent on relative pitch, you would not hear as much as a difference because the intervals remain unchanged.

Answer 4

I like your baby/island question, and it actually has some interesting answers. Not counting for timbre, range, or tuning system, there is nothing inherent in one key center or another that makes them more or less enjoyable.

However, if you took those babies on their islands and switched them for a day, they would probably think that the other island's music was entirely crap! This is because musical tastes, even down to key center and tuning, is socioculturally constructed.

As people brought up in the culture of western classical music (for the purposes of this illustration), we have a number of preferences for key center since that is what composers have written over the centuries. C minor, especially written by Beethoven, is often considered to be reminiscent of "heroic struggle." We could also look for historical musical links between, say, G major and "bright" sounding music, or C# minor as "sad and brooding". It should go without saying that there's no objective list for key centers and their related "feelings"; this is music at its most subjective. But suffice it to say, a classical composer with broad experience in the classical repertoire will usually have an opinion on whether a piece he's written sounds better in this or that key.

Answer 5

I recently bought a recording of Schubert's "Die Schöne Müllerin" sung by a baritone. This song cycle has always been published in two versions, high-key (for sopranos and tenors) and low-key (for altos, baritones and basses). The recording has the low-key version.

Throughout, the piano writing sounds too thick and heavy. The high-key version does not. There is an ideal spacing of notes in piano chords. Four notes all within the octave starting on Middle C sound OK; those same four notes transposed down a fifth sound darker and thicker.

I saw the same effect when I took a song and transposed it up a fifth to be easier for flute and piano. The whole thing sounded too high, as if a record were played at the wrong speed.

In both cases it was not a mismatch of the range of the voice or instrument. The baritone was singing in his natural range, and the flute was in its best range.

Answer 6

I'm currently playing bass in a band that is doing 60-70s covers. Key changes have become an issue. The most popular shift seems to be to the 5th below, depending on who is doing the lead singing. For even just my bass, it spoils the lines that I have heard and have possibly worked out.

For me, dropping a cover by 5 semitones makes it too different a song.

Answer 7

it's called transpose, of course it have a difference in detail.
let say your first note in C4 then goes to D4 (Frequency Difference from C4 to D4 is 32 Hz),
then you want to transpose it into D4 then goes to E4 (Frequency Difference from D4 to E4 is 36 Hz)
although both of note interval difference is same (2 Semitones)