About singing, technology, and Oscar Serpa

Jointly authored by Daniela Pastina and El Victrolero

(Deutsche Version)

The 20^th century has been a century of unprecedented developments in technology, and it has disparately influenced our daily lives. Technological developments invariably bring about changes in the way things are done. Some changes are auspicious, others turn out to have detrimental consequences.

Without the invention of the phonograph, tango would not enjoy its current popularity. Of course, there were dance fads before tango became a world-wide phenomenon, but this was the first time that the originators of the music were recorded and preserved for posterity. It is the unsurpassed performance quality of the great tango orchestras of the “golden age” that is still attracting to tango music dancers and music lovers of all ages in all parts of the world.

1. Recording Music

The possibility to record music is a remarkable fact by itself. The technology at the time the recordings were made (1910's-1940's) was primitive, and so was the result: the sound quality produced by the old technology is meager measured by modern standards (let alone the degradation the records have suffered in the course of time). Nevertheless, it is the quality of the performances that shines and makes one soon hear past the surface scratchiness and appreciate the artistic value.

In the course of time, the recording technology improved, but one thing remained a constant until the 1950s: once made a recording could not be altered. Recordings were cut on wax or lacquer master discs from which the final record was pressed. There was no post-recording doctoring of the performance. If errors had been made and were unacceptable, the recording had to be repeated. In other words: in recordings made before the 1950s, we are essentially listening to live performances.

For a musician, playing for a recording is more taxing than playing for an audience. A mistake made during a public performance is, if noted at all, quickly forgotten. On a record, however, it is preserved permanently. Recording engineers of the wax matrix period have reported that recording sessions were particularly stressful experiences for classical musicians -- they played pieces that are considerably longer than a tango. Frequently the performers got so nervous that each take turned out worse than the one before. In the end, it was often decided to take the first recording since it had the fewest mistakes.

Compared to present-day standards, the recording techniques were primitive as well. At the time of acoustical recordings—when the earliest tango recordings were made—performers had to play or sing more or less directly into the “horn”, a large funnel that collected the sound. Electromagnetic microphones came into use in the late 1920s. They were much more sensitive and allowed musicians to perform at a greater distance. Orchestras were positioned in a circular sector in front of the microphone. The sound balance was arranged by distributing the musicians in the studio: soloists would be placed closer to the microphone to make them clearly audible whereas the instrument sections and louder instruments were placed in the back.

Ernesto Famá during a recording session with the orchestra of Francisco Canaro. Famá is closest to the microphone. In the first row behind him are seated the bandoneons, then the violins. Farthest from the microphone are the muted trumpet (left) and the piano (right).

The microphone, therefore, acted as an “imagined listener” to whom the orchestra played. The instruments were positioned in such a way that this listener received the sound in the desired balance. Today, the instruments are isolated during recording, either by playing into independently regulated microphones, by being placed in separate cubicles, or by being recorded successively on different tracks. The latter is a much more economical way of recording, since errors can be isolated and corrected one by one. It had, however, one great—and adverse, in our opinion—side effect: it shifted the control (and some responsibility) of the performance increasingly from the musician to the recording technician. In fact, it is not uncommon today in popular music that the musicians have no say in the final stage of a recording, the sound mixing. This phase is often left entirely to the sound engineers.

Before the late 1950s, the responsibility of playing well and flawlessly lay with the musicians. Playing well did not only mean to impress the audience, it also meant to get a job done quickly and get on to the next one. The tango recordings of the “golden age” are a testament to highly trained and experienced musicians that do not take a backseat to performers of more demanding musical genres.

2. Singing Without a Microphone

Microphones were available before the music record technology was developed. Without a microphone telephony, for example, would have been impossible. However, the quality of the sound registered by the early microphones was unsatisfactory for music. It was not until the late 1920s that microphones were developed that rendered music sufficiently well to be used for music recordings.

It may seem unthinkable today but it was a reality at least into the 1930s: singers of popular music performed without a microphone. Their voices had to be sufficiently strong to carry even with orchestra accompaniment and in large spaces.  

Libertad Lamarque during a concert performance in 1931, singing without microphone at the Teatro Colón, the opera house of Buenos Aires. She is accompanied by five guitarists. In the back, the Canaro orchestra can be seen, which also accompanied the singers during the concert.

Another striking aspect about tango singers of the “golden age” is that many of them sang with a very clear diction. That is to say, the text is readily understood: the words are so clearly articulated that they could be taken for dictation.

The volume of a voice and clear diction are among the indicators of trained singers. This should come as no surprise: many tango instrumentalists had attended conservatories or had studied with other known instrumentalists. The image of the singer of popular music, however, is not that of a trained vocalist. Unlike an opera singer, for example, who takes pride in belonging to a school of singing, the singer of popular music “just sings”. We have, therefore, practically no information on how tango singers learned to sing.

We have, however, little reason to doubt that singers were formally trained. Canaro, for instance, hired singers that could perform in the plays that he staged with Ivo Pelay. Singing in a theater without amplification, issuing a text with clear diction requires good vocal technique that must be learned.

We also know of one tango singer who had not received voice training: Tita Merello. Tita Merello was actually an actress, but at the time it was expected from actors to sing as well. Very early in her career Merello had tried her luck as a singer in varieté but it turned out to be such flop that she did not pursue a singing career any further. Yet, as she advanced in her acting career, singing became part of the job. In 1932, she was engaged to act in La muchachada del centro, a play by Ivo Pelay with music by Francisco Canaro. As part of the role, she had to sing two pieces with Canaro's orchestra. During the rehearsals she bemoaned that the orchestra was too loud, and that she was afraid not to be heard since she did not have a big voice. Canaro set her at ease and promised to keep the orchestra down. 
The play and Tita Merello were a huge success with the audience and was performed with the same cast nearly 600 times. Yet, two years later, the Argentinian magazine Caras y Caretas reported: “Tita Merello has finally decided to take voice lessons under the direction of Maestro Longomuto.”

3. The Singing Tradition

At the time when tango was developing as an indigenous urban music of Buenos Aires, Argentina witnessed the “great European immigration wave”. The largest groups of immigrants came from two countries: Italy and Spain. The influence of the immigrant communities on tango was decisive. Most tango musicians, if they were not born in Europe themselves, were children of “tanos” or “gallegos” who had immigrated.

The immigrants brought along their musical traditions, but not only that: there were musicians and music teachers among them to cultivate and propagate their aesthetics and styles of music making. Also, Buenos Aires was not a provincial backwater but a growing metropolis in which visiting or resident musicians from Europe where commonplace.

The common style of singing practiced and taught in southern European countries during the 19^th and early 20^th century originated from the Italian bel canto opera tradition. In this style, the quality of sound emitted by the singer is of paramount importance. The aim is to control the intensity and color of the tone in such a way that it is consistent throughout the singer's range, from high to low. Musical phrases and melodies must be articulated as larger, uninterrupted units with a smooth transition from note to note while being audible in a theater even when sung softly.

Singers sing, of course, not just musical notes but texts that are structured as syntactical phrases. In the bel canto style of singing, it is imperative that words and text lines are clearly articulated. The words are to be stressed, at the same time, so that they convey an interpretation of the text. In order not to disrupt the smooth transition from note to note, this is usually achieved through a mode of singing called rubato (“stolen”). Rubato singing means that the emphasized notes are sung longer than their value indicated in the score. This extra value is “stolen” from the surrounding notes, which must be sung shorter so that the performance still maintains the tempo and meter.

Tango was in its beginnings primarily instrumental music, but it turned into predominantly vocal music with the “invention” of the tango canción in the mid-1920s. Listening to the music that was recorded during the following 20 years, one cannot but be amazed about the abundance of excellent singers that performed and recorded tangos and its related song types at that time. Towering above all others was, of course, Carlos Gardel, an exceptional singer who also played the guitar, composed a number of enchanting songs, and was one of the Argentinian silver screen idols of the day. But there were others, no less worthy of exaltation. One of them, Oscar Serpa, though not completely forgotten but commonly only mentioned inter alia, deserves—in our opinion—far greater recognition.

4. Oscar Serpa (1919-1982)

Unfortunately, hardly anything is known about Serpa's musical training. His parents sent him to a conservatory where he learned to play the guitar and probably singing as well. At age 17, he was “discovered” by Abel Pelaia, a singer who had lost his partner and, looking for a replacement, had heard about the young singer. Serpa and Pelaia worked on a repertory and a few months later began to perform as a duo. Serpa's career as a tango singer began in 1942, when he joined the orchestra of Osvaldo Fresedo.

Alberto Podestá on Oscar Serpa (2012)

Serpa owes his singing style to a long-established tradition, which slowly but inexorably disappeared from popular music in the course of the 20^th century due to the introduction of the microphone and technological progress. His voice stands out for elegance of phrasing and homogeneity of timbre. The production of sound is clear and effortless, uninterrupted, and naturally vibrant. In Serpa's singing there is no search for dramatic effects, no signs of harshness, no mirky sounds. The impression is that he could sing beautifully despite being in front of a microphone, rather than thanks to it. Serpa learned to sing when microphones were not part of the singing technique and certainly, from his recordings, extending over a period of about 20 years (from Niebla del Riachuelo in 1937 to Verdemar in 1955), one can hear that he never changed his style.  

5. Niebla del Riachuelo

When he was only 18, a year after becoming a professional singer, Serpa recorded Niebla del Riachuelo (1937).  In this piece, one can appreciate the singer's clarity of diction which brings the text to life. It is an honest tenor voice, which does not try to sound like something it is not, does not use effects, and, perhaps against today's taste, never produces what to a classically-trained ear is an unpleasant sound. A comparison with a modern recording of Niebla del Riachuelo by the flamenco singer Diego “El Cigala” will clarify the differences.

One may object that Diego “El Cigala” sings a completely different style of music with its own aesthetics, one that values spontaneity over control. Certainly, Serpa and Cigala stand at opposite ends of a spectrum. Serpa's goals of producing clear and smooth sound, musical phrasing, and diction are not Cigala's. Unlike Serpa's, Cigala's voice is raspy, he swallows syllables, blurs vowel sounds, even changes words. Disruptive of the vocal line are also the strong guttural “k” sounds (see, for example the line “... barcos carboneros que ...”) which may be due to the fact that Cigala is singing directly into the microphone). Also, it is impossible to get an idea of the volume of Cigala's voice since the microphone equalizes the difference between loud and soft passages and an artificial reverberation is added.

Serpa's performance was a “live” performance, that is, a recording cut on a wax disc. All musicians (singer and three guitarists, we guess) were placed in front of the same microphone, the singer being closest yet still keeping some distance. As much as the technology of the day allows it, the recording offers a natural rendition of the voice, that is, without added technical effects.

In Serpa's version the text becomes clear, on the one hand, because Serpa produces clear and consistent vowels; and “o” sound is always an “o”. The sound may be shorter or longer, but it is always an “o”. On the other hand, Serpa enunciates the consonants even if they occur at the end of a word. One hears unmistakably “recalar”, “quedar”, and “dolor”, and not “recalá”, “quedá”, or “doló”. Like good classically trained singers, Serpa sings “on the vowels”, that is, he carries the vowel sound into the next pitch, sounding any consonant just a heartbeat before the next vowel. This is also a reason why Serpa's “k” sounds are not as conspicuous as Cigala's.

6. Uno

As a singer with tango dance orchestras, Serpa made his mark with two of the most important ensembles of the 1940s-50s: those of Osvaldo Fresedo and Carlos Di Sarli. Both Fresedo and Di Sarli cultivated a soft, violin-dominated orchestra sound that made a perfect match with Serpa's mellifluous and elegant singing. Perhaps it is due to different recording techniques, but Serpa blended particularly well with Fresedo's sound, making his voice sound as if it were another instrument of the orchestra. Fresedo's recording of Mariano Mores' tango Uno, recorded six years after Niebla del Riachuelo, presents Serpa as a matured singer with a remarkable singing technique and expressiveness.

Oscar Serpa (left) and Osvaldo Fresedo

Formally, Uno is a tango canción, that is, it consists of two sections, the first setting the verse of the poem to music, the second the refrain. The first part, in turn, divides into two phrases, antecedent and consequent, of 8 and 15 measures length, respectively. We shall examine here only the antecedent phrase and its transition to the consequent one as an example of Serpa's singing, but we invite the readers to listen to the full-length recording and compare it to versions by other singers (many of which are available through online resources).

The antecedent phrase is composed of two melodies with a length of four measures. They unfold an ascending chromatic (half-tone) scale in relatively fast note values. This ascending melodic line continues to be developed until it reaches its apex at the beginning of the consequent phrase. After the apex is reached, the melodic motion gradually descends again.

The melodic motion is emphasized through dynamics. The score indicates a piano at the beginning of the antecedent phrase, followed by a crescendo to a forte in the second melodic segment in measure 5. The crescendo continues to reach a fortissimo with the first note of the consequent phrase, the apex, and the dynamic strength decreases again with the descending melodic motion.

One may, thus, compare the musical setting of the first section to a wave that gradually surges to its highpoint and then ebbs away again. The Fresedo orchestra beautifully conveys this swelling and abatement in their recording. The violins articulate each note of the ascending melody with broad tenuto strokes (interrupted only by a short staccato passage added for variety), proceeding to the subsequent phrase by gradually increasing volume and intensity. With the subsequent phrase the surge of intensity reaches its climax in the fortissimo that is effectively underscored by the percussion instruments. Uno appears to be a composition tailor-made for the orchestration of Fresedo's orchestra.

Serpa's rendition of Uno with Fresedo is a textbook example for clarity of diction. It evinces another aspect of his excellent technique: his breath control, which makes the long ascending phrases of Uno sound effortless. He only takes breaths when the inevitable sound interruption does not interfere with the musical (or poetical) phrase. In this recording one can appreciate the perfect intonation and the free, constant vibrato in the voice. He shows great sensitivity in stressing the right syllables so that the phrase makes sense musically and poetically.

Serpa's voice blends perfectly with the orchestra, in fact, it seems more like another musical instrument rather than soloist singer. He makes a tasteful use of portamentos (that is, bringing a note up or down seamlessly to the next pitch) and of the rubato (see above). The second occurrence of the word “uno”, the fortissimo apex of the section, is breathtaking. Rather than bellowing out the word after a long crescendo in fortissimo, he beautifully serves the “u” on a tray and continues effortlessly to the end of the phrase (“...amor”) where he takes a quick breath before the next sentence—not out of necessity but for the sake of phrasing. If there is beauty in tango singing, it is exemplified in this passage.

We have chosen to write about Serpa because, to our opinion, he represents well the singing of his day. Serpa mastered the centuries-old art of singing in his production of a clear, effortless sound. His naturally vibrant voice was endowed with exceptionally clear diction where consonants were no obstacles but rather helped the vowels to shape words into meaningful text and thus song. This style of singing has all but disappeared from popular music. Modern technology—microphones, electronic sound enhancement, editing techniques, etc.—has made it easy for lesser voices to sound grand and it has affected popular taste. Yet, we must be grateful to technology for preserving so many great musical recordings for us to enjoy today.

© 2018 by Wolfgang Freis