Musical Offering BWV 1079
Jordi Savall
Review: Savall Musical Offering

Review: Savall Musical Offering

Kirk McElhearn wrote (September 9, 2001):
One is occasionally fortunate enough to come across a new recording of a great work, and a personal favourite, that is so good that all other recordings pale in comparison. A recording that is not only played and recorded impeccably, but also one that gives the listener an entirely new outlook on the work in question. This is the case with this new recording of Bach's Musical Offering by Jordi Savall and Le Concert des Nations.

The Musical Offering is the name of a set of pieces that Bach wrote for king Frederick II of Prussia. Bach's son, Carl Phillip Emmanuel, was harpsichordist for the young king, who was an avid music lover. He begged the younger Bach to have his father come and play for him. One evening, Johann Sebastian showed up, and the king immediately ushered him to a pianoforte, where he played a theme for a fugue. "Old" Bach improvised a fugue to this theme, but was so impressed by it that he wrote a much larger set of pieces around this theme, and dedicated it to the king, hence, this
Musical Offering.

Like Bach's Art of Fugue, this work shows the many possible ways that a single theme can be elaborated on to make a large, varied work composed of fugues and canons. The two works do indeed have many similarities. The themes are related, and the manner of treating them is similar.

This new recording is brilliant, and makes all previous recordings redundant. Savall and his ensemble have managed to take this music and give it the attention and creativity it needs. I have several recordings of this work, but none come close to this one.

Several elements stand out here, beyond the quality of the playing and the excellent sound. First of all, Savall has taken a totally new approach to the work. The disc opens with a brief piece, which is not in Bach's score - a 29 second performance of the "royal theme" on solo flute. This is an excellent idea, for it gives the listener the essential musical element of the set on its own, allowing them to more easily follow the development of this theme through the various canons and fugues.

Savall has taken "liberties" with the order of the pieces - although, since no one is sure exactly what order should be respected, this is not unjustified. The work is made up of several small canons, a long (almost 9 minute) Ricerar a 6 for harpsichord, and other fugues. But its centrepiece is the sonata in four movements that gives the royal theme its most intricate exposition. Savall puts this sonata in the centre of the work, just after the Ricercar a 6, placing the two major pieces of this work in direct relationship.

Several of the canons are also played in an interesting manner. Savall plays four of these canons beginning with just one voice, then adding the next, then the next. This gives a much more interesting tone for these cryptic works (the canons were not fully written out in the work; they were enigmas that had to be figured out to be played). He also repeats one of the canons in a different manner, and repeats the Ricercar a 6 at the end of the disc, this time arranged for strings and harpsichord. (These additions explain why this disc is over 71 minutes long, compared to 45 to 50 minutes for most other recordings of the work.)

One of the high points of this disc, in my opinion, is Pierre Hantaï's solo performances of the Ricercar a 6, the Ricercar a 3, and two canons played by solo harpsichord. The instrument he is playing on is one of the best sounding harpsichords I have ever heard on a recording (the only one that I know that sounds better is the magnificent instrument used by Kenneth Gilbert on his Well-Tempered Clavier recording). No information is given in the notes as to the instrument, but it has a magical sound: it is dark and smooth, especially in the back register; its bass notes are rich and ample, its treble silky and bright, though not aggressive. The quills (the plectra used to strike the strings) are highly flexible, and sound as though they are real bird quills. In the Ricercar a 3, when he plays with the two registers coupled, the resonance of the instrument is even more astounding, and its sound when using the lute stop, in the Canon a 2 Quaerendo invenietis (A), is amazing. It sounds as if it is using brass strings, which could explain the magical sound, but, whatever the case, Hantaï's instrument and performance are extraordinary. His performance of the two Ricercars is so excellent that it is worth buying this disc for these two pieces alone.

The sound of the rest of the recording is excellent as well. There is an amazing separation of the instruments, even at high volumes, and they all fit together perfectly, with flawless¹ balance.

All these elements coalesce into what I feel is the finest recording of this work I have ever heard. The only thing I can say is that this is an essential disc for all lovers of Bach's music. And if you are not familiar with this work, this is even more essential - not only can you discover one of Bach's greatest works, but in perhaps its finest performance as well.

Feedback to the Review

Thomas Braatz wrote (September 9, 2001):
[To Kirk McElhearn] This certainly sounds like THE recording of the MO to get. So far I have not been disappointed with any other Savall recordings that I have.

One question on the harpsichord terminology that you use: I am not familiar with the English terms "back" and I assume also "front" register, although I can imagine that they mean the distance from the player. What if you have an instrument with two 8' registers? Do these terms indicate which register is nasal and which is not? What are all the available registers on Hantaï's harpsichord? Do the liner notes give this information? Is the name of the harpsichord maker given?

Thanks for the excellent review!

Kirk McElhearn wrote (September 9, 2001):
[To Thomas Braatz] The liner notes give no info at all regarding the instrument. Brad helped me better understand what made it sound good (I sent him an mp3).

The back and front are, indeed, the distance from the keyboard. The further from the jacks the plectra are, the richer the sound (correct me, Brad, if I'm wrong). You can here the same thing on a guitar - when you play close to the bridge, the sound is a bit higher.

Harry J. Steinman wrote (September 9, 2001):
[To Kirk McElhearn] Your review makes me want to run right out and get this recording. Alas, I can't find it! Tried Amazon.com; Tower.com; Berkshire; HB Direct, CD Universe. Where should I try??

Kirk McElhearn wrote (September 9, 2001):
[To Harry J. Steinman] It's not out yet. In France, it is shown as being available on Sept. 28. I guess it will take a while longer in the US.

Michael Grover wrote (September 9, 2001):
[To Kirk McElhearn] Thanks for the excellent review. Things I'm curious about, however, are the instrumentation used and the size of Le Concert des Nations. You mentioned flute and harpsichord but nothing else -- is it just strings for the rest or are other instruments present? Cello, gamba, or both? And does the CD booklet mention how many players there are?

Bradley Lehman wrote (September 9, 2001):
[To Kirk McElhearn] Kirk, a correction of your terminology here...the plectra are *in* the jacks. I think the phrase you were looking for was something like: "the further from the nut the plectra pluck,..."

The jack is the strip of wood (or other material) that carries the plectrum up and down past the string. The lower end of the jack rests on the back end of the key, which is a simple lever, so when the player presses the key down its back end and the jack go up. The plectrum (made of bird quill, stiff leather, or more recently plastic) plucks the string on the way past it, and a pivoting mechanism with a spring mounted in the jack allows the plectrum to pivot gracefully out of the way and not pluck the string a second time on the way back down. When the player has released the key, the string's vibration is stopped by a damper (or two): a small tag of felt attached to the jack. The dampers rest on top of the string to prevent that string from resonating sympathetically when it is not being played.

Each register in a harpsichord is a set of strings plus a set of jacks/plectra. When the register is turned off, all its jacks lean very slightly to the side so their plectra will miss the strings when the keys are played, but the dampers are still in contact with the strings to keep them from sounding sympathetically. When the register is turned on, those jacks are aligned to stand up straight so their plectra will play.

The sounding length of each string is stretched between a bridge (in a graceful curve, and mounted on the soundboard) and the nut (a second bridge, straight, directly in front of the player near the tuning pins).

If the harpsichord has two registers at 8-foot pitch (concert pitch), those will necessarily be at different distances from the nut, since two sets of jacks obviously cannot occupy the same space. The one whose jacks are closest to the player is called the "front 8." The one farthest away from the nut is the "back 8." (There might also be some middle 8's, rarely.) Those distances from the nut, both relatively and absolutely, are tremendously important to the tonal design of the instrument. A millimeter in either direction affects the tone quality, and the builder must be very careful to place these rows of jacks at exactly the right places for the tone s/he wants in the design. Then the shape, thickness, and material of the soundboard must also be exactly right to amplify the strings' vibrations appropriately, giving the instrument resonance and power with a pure tone. (This is all a very delicate science, just as in violin-making!) The materials used for the strings and plectra matter, too, but not as much as getting the soundboard and the physical placement of the registers exactly right.

In general, the farther the plucking point is away from the nut, the deeper the tone will be (not the pitch, but the tone). As Kirk mentioned, it is easy to illustrate this on a guitar, plucking the same string at different points and noticing the difference of tone quality.

Any single musical tone (except one from an electronic sine-wave generator) is a composite of different harmonics, different pitches really, in a balanced mix. If a string's basic vibration rate is 100cps, it is also vibrating at 200, 300, 400, 500, 600, 700... simultaneously, though of course with the 100 and 200 usually the loudest by far. 100 is approximately the G at the bottom of the bass staff; 200 is the next G up; 300 is middle D; 400 is the next G; 500 is B; 600 is D; 700 is F; 800 is G.... When that single low G is played, we are really hearing all those other notes simultaneously in a mix!

That brings us back to the placement of the jacks and plectra. If the plucking point is near the nut, i.e. near the end of the string, those higher harmonics become more prominent in the mix. If the plucking point is farther away, i.e. toward the middle of the string, we hear only the first few (lowest) harmonics in the mix. In that case we perceive the tone as darker and richer, but also less brilliant or penetrating. It's a trade-off.

The farther the "back 8" is from the nut, the more the harpsichord starts to sound like a virginal. A virginal plucks its strings at a point usually somewhere between 1/4 and 1/3 of the string's sounding length (like plucking over a guitar's fingerboard). A harpsichord plucks them much nearer the end, more like 1/32 to 1/8 of the length. Consequently a harpsichord has a brighter and more complex tone quality than a virginal.

Putting all this together: the "front 8" on a harpsichord gives the brighter, more brilliant, more nasal tone (weaker in the lowest harmonics), while the "back 8" supplies the deeper and more fundamental tone, a less complex sound. They can be played separately or together.

Hantai is (I believe) playing the Ricercar a6 on the "back 8" alone, and that register on that harpsichord seems particularly well designed tonally by the builder. Such a rich and mellow, yet still very interesting, sound!

Here is a fun exercise to hear the above harmonic mix for yourself, inside your own head. It might be shocking the first few times you try it, but I think it's nifty. Take a deep breath, and fairly loudly sing a low note. Holding that note steady, slowly and gradually change the shape of your mouth so it forms all the possible vowels and all the mixtures in between the vowels. U...Oh...Ah....Eh....I.... As your mouth's volume and shape exactly catch some of those higher harmonics (especially in the U to Oh range), you will suddenly hear these new high notes inside your head! Sometimes only a very slight movement of the lips or jaw will take you over to the next one.

In this way you can appear to be singing two or more notes at the same time. (Tell your neighbors you are demon-possessed....)

This is even better when done in a resonant room. Sing in the shower! Every room has some pitches that work better than others. For example, long ago I discovered a toilet stall at my college that went crazy with harmonics whenever I sang an A-flat, but did hardly anything when I sang a B. Rooms have good and bad pitches, and they also have "nodal" points where everything frankly sounds dead regardless of pitch. In a concert hall, of course, the designer hopes to minimize the number of seats that are at these nodes....

All these acoustical points get complex. How does an instrument make a single tone, and how well are that tone's harmonics balanced? Do the other tones from the same instrument have a similar tone color? How much range of variation does the player get to manipulate? How do multiple sounds from an instrument blend? How do several different instruments blend together? How does the room amplify those sounds additionally? How does each listener hear everything slightly differently, both because of placement in the room and because of head and ear shapes?

In a recording there is also the consideration of microphones: whatype, and exactly where should they be placed to give an accurate and pleasing sound? Any filtering? In which direction(s) should they be facing? (High pitches are much more directional than low pitches are...that's why your stereo system needs only one subwoofer; stereo doesn't matter much for low pitches.) The listener's playback room, speakers, headphones, etc all also add some tonal coloration to the mix....

Then it gets yet more complex with tuning and temperament. Those ratios of 100, 200, 300, etc. within a single tone make acoustically pure intervals. But if we tune the notes on the instrument exactly to all those pitches, it limits us to playing in only one key. So we start fudging the pitches to be almost but not quite in tune, giving us more usable keys but less resonance. (If two musical pitches have some component harmonics that coincide exactly, they reinforce each other and the combination seems considerably louder.) How much fudging is too much for the pieces to be played? How much is too much before the instruments start to sound dull, and/or "out of tune"? That is the art of temperament.

Kirk McElhean wrote (September 10, 2001):
[To Michael Grover] Well, the MO is always a small group. Altogether, there are 8 musicians, but with the exception of the second version of the Ricercar a 6, there are never more than 4 playing at a time, as per the score.

Jordi Savall: Short Biography | Hespèrion XX | La Capella Reial de Catalunya | Le Concert des Nations | Recordings of Instrumental Works | Recordings of Vocal Works
Reviews of Instrumental Recordings: Bach Sonatas for Gamba and Harpsichord | Review of Savall Musical Offering

Pierre Hantaï: Short Biography | Recordings of Instrumental Works
Reviews of Instrumental Recordings: Bach's Well Tempered Clavier Book 1 from Pierre Hantai | Bach's Toccatas for Harpsichord from Watchorn & Troeger (3 Parts) | Hantai’s Bach | Pierre Hantai's 2^nd Recording of the Goldberg Variations | Review of Savall Musical Offering
Discussions of Instrumental Recordings: Goldberg Variations BWV 988 - played by Pierre Hantaï