Mixtures can make the organ vocal. I don't mean to literally make the organ "speak" words. Instead, by mimicking the formant structure of vowels, mixtures can add a vocal-like color that has powerful musical effects.
Think of it this way: A choir singing a complex canon. If everyone sings on "Mmm", you can't tell the parts apart. But if Sopranos sing "Ee", Altos "Ah", Tenors "Oh", and Basses "Oo", each voice is instantly distinguishable. In the same way, a subtle vocal-like color from the mixtures clarifies the polyphony.
When you speak, your mouth emits a complex mix of sound (vowels and consonants), which are assembled into syllables, words and finally speech. Speech actually is made of two simultaneous tones (or more correctly, partials): a low pitched fundamental (about 250-500 hz) plus a higher pitched partial (900-2,300 hz). There are actually more than two partials, but for what we are doing, we only care about the first two partials (in the drawing, its the left two, lower-pitched partials).
Vowels are distinguished not by fundamental pitch, but by higher pitched partial. The fundamental pitch is the note that you sing. A vowel upper partial is not an exact pitch; rather, it is a band of sound.
Approximate center frequencies of vowel upper partial
You can understand speech, because we all have a sort of "perfect Pitch" ability. Under the right conditions, we hear a sound within certain frequency bands, as a vowel. If you want to hear vowels, try slowly whispering these vowels in this order: "U, O, A, E, I". You can hear the vowel, but you also hear a hiss-sound rise in pitch. It's subtle, but it's all you brain needs translate the pitch to a vowel. Our brains are hard-wired for this.
Speech starts with fundamental sound from your vibrating vocal cords (about 250-500 hz), travelling up your vocal tract (throat and mouth cavity), and out your mouth. Of course the fundamental isn't a sine wave; it includes rich background broadband noise. Your vocal tract (throat, mouth cavity, lips and tongue) is a resonator which acoustically selects, filters and amplifies some of that noise. By moving our lips, jaw and tongue, we tune the vocal tract, creating the partials of the vowel we want.
The pipes in a Gothic organ were all principals, with identical voicing and scale. Typically they were a chorus of 8', 4', 3', 2', 1-1/3', 1' and 1/2' ranks. Because there was no stop control, all ranks spoke, all the time. This made a powerful, brilliant sound that filled the room.
But the sound was pure and brilliant chorus. There was no partials, so no vowel colouration. Vowels consist of a lower and upper partial with a gap between them. You need the gaps, to allow the vowel format stand out. But the blockwerk always had all ranks speaking, there were no gaps, so no vowel effect.
With the newly invented stop controls, Renaissance Organs broke apart the Blockwerk. The 8', 4, 2-2/3' & 2' were given their own stop knobs. All the higher pitched ranks were gathered together to form a Hintersatz (blockwerk of upper pipes), simular to a mixture without breaks. The pipes still were all principals, with identical voicing and scale. Despite stop controls, the chorus was essentually a blockwerk decompose with no mixture breaks, no vowel sounds.
These "vocal" organ effects are most noticable on old North German/Dutch organs. Modern "neo-baroque" organs with low cutups, stringy mixtures, and loud, bright 4', 3' & 2' chorus stops fill the entire spectrum. You end up with a brilliant chorus, but no vowel. With all that brilliance, the vowel format is buried. To hear the vowel, you need space around the Mixture. For a vocal organ the mixture wants to be bold, but fluty (cut it up). And the 4, 2-2/3' & 2' should not be assertive.
Mixtures can be difficult to visualize, so I'm going to share the chart (hypothetical blockwerk- shown on right) that I graph mixture breaks on. The horizontal scale, from left to right represents the notes of the keyboard, starting a C1 to C61. The vertical scale represents the actual pitch the pipe speaks. The diaginal lines represents each rank of pipes. (Click on the chart to enlarge).
When pipes of a mixture rank get too small, they jump back to a bigger size. That is called a break and is shown on the graph by jumping to a lower diaginal. This graph is a hypothetical blockwerk so there are no breaks.
Mixture Files
The frequency range of vowels almost exactly concides with the range of a 2' stop. Not coincidently, it is also the range of maximum human hearing sensitivity. Sit down at an organ console and put on an 8' and 2' stop. Now, starting at bottom C, slowly play up chromatically to the top. Besides the 2' sound, you may faintly hear a vowel sound, starting at "ooh" at bottom "c" and sliding vowels, to "ee" at top "c". It can be hard to hear; try whispering the vowels to inform your ear as to what you're listening for. Also, try different stops.
Look at the "Typical Great Mixture" chart to the upper-left. The mixture starts at "C1" as 1-1/3', 1', 2/3' and 1/2' ranks at the left side of the chart. It breaks about every octave so that at the right side of the page at topnote "c61" it is 8', 4', 2-2/3' and 2'.
Almost all mixtures end up at "c61" as 8', 4', 2-2/3' and 2'. It cannot be higher than 2' at "c61", because the pipes get too small to tune. You shouldn't use the 5-1/3' or anything bigger than 8' or you introduce a 16' resultant tone. A higher pitched Posttiv mixture (e.g. C1 = 1', 2/3', 1/2' and 1/3') would need an extra breaks so to end-up at "c61" as 8', 4', 2-2/3' and 2'. You can judge a mixture by noting when the 2-2/3' enters the mixture.
A mixture stop artificially creates a fixed set of reinforced harmonics, independent of which key you play. That is exactly what a formant does: a fixed frequency region that emphasizes certain harmonics of a variable fundamental. If you design a mixture to strongly reinforce harmonics around, say, 500 Hz and 1500 Hz, it will make any note played on it sound somewhat like a particular vowel, relative to that pitch.
Your brain selects the mid-point of the four ranks of a mixture, to determine the vowel sound. On the "Typical Great Mixture" chart, I've drawn the "mid-point vowel line" as a row of "XXXX". I've also written the associated vowels onto that line. So you can see that this great mixture will start at bottom C1 with an "oh", vowel and the tone will gradually transition to "a" at top C61.
Okay, a division with a Mixture can have a vowel colouration, which changes from dark to bright (Oh - ee) as you play up the keyboard. How does that help? You gain two advantages from vowel colour. Please note that the vowel effect can be subtle; if you voice your chorus like a blockwerk, the vowel effect disappears.
This is the most critical musical function in contrapuntal music (Bach, Buxtehude). When several independent melodies (voices) play at once in the same register, they blur together. A mixture solves this by giving each part of the keyboard a different harmonic fingerprint.
Think of it like this: A choir singing a complex canon. If everyone sings on "Mmm", you can't tell the parts apart. But if Sopranos sing "Ee", Altos "Ah", Tenors "Oh", and Basses "Oo", each voice is instantly distinguishable.
The organ mixture does this pitch-dependently. The left hand playing a low C will have its mixture reinforce harmonics around 1100 Hz (an "Ah" quality). The right hand playing high C on the same mixture stop will have its reinforced harmonics around 2,250 Hz (an "Ee" or hiss-like quality).
In summary: The mixture acts like an automatic intelligibility filter. It vocalizes each pitch range differently, so the listener's ear can effortlessly follow intertwining melodic lines simultaneously. Without this vowel-formant effect, complex fugues become muddy soup.
Because pipes in Blockwerks (and Renaissance organs) have the same scaling and voicing, and they do not break, the organ sounds almost the same as you play up the keyboard from bass to treble. The left hand sounds like it is playing on top of the right hand, muddying the polyphony.
The "magic line" on my mixture sheet is an arbitrary curve, on which the vowels spread out evenly, over the compass of the keyboard. A Great mixture will typically fall below and touch it, Positiv will be on it and Brustwerk be above it. If you draw an unknown mixture onto a blank mixture sheet, the magic line instantly tells you how the mixture compares.
If you look at the "master sheet" to the right, you will see curved lines labeled "Ped, Gt, Pos, Bw". These lines indicate the center points of Mixtures for keyboard of the organ. A different mixture curve means that each keyboard has a different range of vowels.
Playing a plenum chord on the Great will have a subtle "ah" sound, Pos an "eh", and Bw an "ee" sound. The vowels give each keyboard a characteristic sound, so a listener immediately recognizes a keyboard change. The Mixtures are the most important stop, to provide differentiation of character between the various keyboards.The mixtures make the Positiv brighter sound than the Great, and the Brustwerk sounding the brightest.
In summary mixtures don't just make an organ louder. They contribute a vowel character to the chorus that 1) helps identify which keyboard you are playing on, and 2) where on the keyboard you are playing. The pitch of the mixture determines which vowel you hear. The voicing and scales of the chorus determine how obvious the vowel is.
I've already described my approach for Mixtures in Classic organs. But what about Romantic Swells? In smaller organs I often treat the Swell as a Positive division that also has a string and Celeste pair, plus appropriate reeds.
But in larger organs, with a 3rd manual Positiv, I let it stand as a Romantic Swell, without Positiv ambitions. I give it an 8' chorus (like a Great) with a lower pitched Great style mixture. If the swell lacks a 2' Principal, I'll use a 2' Mixture, to fill in the Swell chorus.
An English stop I like to use in big Swells is a Harmonics III, often as a second Swell mixture. I call it "Harmonics" to emphasize that it is not a chorus mixture; it is used only to reinforce the reed chorus (though the English will sometimes call it "Mixture"). Note that it has a Tierce, which meshes well with the strong thirds found in reeds. This wants generious cutups to blend well with the reeds.
Shown here is the new Harmonics added by Mander, to their 1998 rebuild of the huge Skinner at Christ Church, Cranbrook, near Detroit, Mi. In the Mander example, all ranks had the same scale, halving on the 18th.
I love Cornets. But they are not a mixture. They are a solo stop made up of a collection of open flute ranks that are harmonically related.
There are mixtures with tierce ranks, originally intended for Meantone organs with pure thirds. I don't recommend using tierce ranks in equally tempered chorus mixtures. It's tuning clashes with tempered thirds, muddying the chorus. Tierces should only be used in solo combinations (like Cornets), so they don't clash with the temperment.