This is the second article in an occasional series about tango songs that have special qualities but which are not heard as often as they deserve.

Uno, with music by Mariano Mores, setting a lyric by Enrique Discépolo, has several things in common with the song featured in the previous article, Discepolin. That song was a tribute to the lyricist of Uno, and two of the leading recordings were by the orchestras of Troilo and Fresedo. Now, with Uno, Troilo and Fresedo feature again, but I also consider several other recordings (and there are many more, besides).

Discépolo’s lyric is superb. It speaks of a man betrayed, fearful of loving again. Derrick Del Pillar’s translation to English is worth knowing.

Empty now from loving and from crying
over so much betrayal!

Most people are likely to choose their preferred version based on the singer, so here are the vocal entries of four competing versions – all recorded in 1943, when the song was newly-published. Each is very fine, and while I also consider six more recordings, I don’t think that any of them are serious rivals to the 1943 versions.

The Leading Contenders

The first recording was by the orchestra of Francisco Canaro, with singer, Carlos Roldán, recorded on 26 May 1943. The arrangement is straightforward, well-played and the recording is decent. However, the string writing sounds rather too much like the style of the De Angelis orchestra for my liking.

Uno: Canaro (Roldán), 1943 – Vocal entry.

Next in the studio was Aníbal Troilo with singer, Alberto Marino, and the recording was made on 30 June 1943. Marino is probably the best singer of the four and the orchestral playing is matchless.

Uno: Troilo (Marino), 1943 – Vocal entry.

Juan D’Arienzo recorded the song on 23 November 1943 with singer Héctor Mauré. The piano part, more than anything, tells us that the orchestra is D’Arienzo, but this is far from the rhythmic, driven, sound of the late 1930s.

Uno: D’Arienzo (Mauré), 1943 – Vocal entry.

A few days later, on 1 December 1943, it was the turn of Osvaldo Fresedo with singer, Oscar Serpa. It would be a mistake to think of this a lightweight performance. There is some beautifully shaded playing from members of the orchestra and Serpa’s vocal contribution is notable.

Uno: Fresedo (Serpa), 1943 – Vocal entry.

Other Versions

Hard on the heels of the four 1943 recordings is this one by the orchestra of Rodolfo Biagi with singer, Carlos Acuña, from April 1944. I can’t get on with it: arch-rival, D’Arienzo surely had the measure of the song. Carlos Acuña seems to be trying to get to the end as quickly as possible, while Biagi sounds completely outside his comfort zone. (D’Arienzo was out of his comfort zone with Héctor Mauré, but he managed better than this).

Uno: Biagi (Acuña), 1944 – Vocal entry.

Troilo returned to the song in 1952 with singer, Jorge Casal. Troilo had switched record company from Victor to TK, and the sound is pretty dreadful for 1952. The arrangement is substantially the same as the Marino recording of nine years before, and if a master like Troilo feels he has something new to say about a previously recorded song, then it must be worth hearing. In the end, Casal can’t match Marino, and there’s no escaping the poor sound. Even playing it with other Casal recordings of the 1950s would be problematic: Troilo was just adopting a significantly new style (and one which most people think puts his music just beyond the mainstream of tango dance music of the Golden Age), and Uno was just before the switch. Most of the other Casal recordings belong in the new phase, and they are not obvious partners.

Uno: Troilo (Casal), 1952 – Vocal entry.

Carlos Acuña recorded the song, again, in 1957, under the direction of Uno’s composer, Mariano Mores. This time we hear the short orchestral introduction as well as the singer’s entry. Often, a composer’s own recording has to be considered to be definitive, but this just makes me laugh: what was he thinking? Acuña doesn’t help by singing under the note.

Uno: Mores (Acuña ), 1957 – Intro & vocal entry.

Armando Pontier also recorded the song in 1957 with Julio Sosa, but this is not a dance arrangement – the Golden Age was over. On its own terms, it is very fine until the entry of Sosa. Sadly, his voice is processed with so much added reverberation that it just sounds silly. Here is the greater part of the introduction, leading into Sosa’s entry.

Uno: Pontier (Sosa), 1957 – Intro & vocal entry.

Pontier made another recording of Uno in 1968 with Roberto Goyeneche, generally considered to be the finest tango singer of his generation. It too is concert music, but on its own terms much more successful than the earlier recording with Sosa. It is also the longest of any of the recordings, coming in at 4:10.

Uno: Pontier (Goyeneche), 1968 – Vocal entry.

The last version I’ve considered is also the latest recording. Osvaldo Pugliese recorded the song with singer, Abel Córdoba, in 1976. Again, this is in concert style. The orchestra arrangement sounds more like a Pugliese ‘greatest hits’ compilation, while Córdoba sings along. It is full of Pugliese cliches and treats the song in a very free and rhapsodic way. You’ll either love it or hate it. I tend to the latter view, but it’s worth hearing, but perhaps not more than once …

Uno: Pugliese (Córdoba), 1976 – Intro & vocal entry.

Comparing the 1943 Arrangements

The song has a very unusual structure. Discepolo’s lyric has three verses, but all of the arrangements set just the first two. That is unusual, in itself, as typically in a 1940s dance arrangement, the thematic material is played right through orchestrally before the singer enters to deliver the first verse of the lyric. We usually don’t hear the second verse at all, because the orchestra once more takes up the themes. The singer returns with a final verse or chorus to wrap up the arrangement and that’s it.

These arrangements all begin with an orchestral introduction (although Canaro’s is abridged), but from the first entry of the singer, two complete verses of the lyric are sung and at the conclusion of the second verse, the song ends (although Canaro offers a brief repeat).

The musical phrasing is very unusual, too. Usually, we expect to hear two contrasting sections: you can refer to them as A or B, or perhaps as verse and chorus. They are normally each of sixteen bars’ duration (4 + 4 + 4 + 4), and the whole song can be written out as a thirty-two bar piano score (16 + 16), printed on two sides of paper. This song is much longer, and the sections are of irregular length. The first section (A) is of twenty-two bars, grouped 4 + 4 + 3 + 4 + 4 + 3, and the second section (B) is of twenty-eight bars, grouped 4 + 4 +4 +4 +4 +4 +4.

Canaro’s introduction seems simple. He gives us just the first eight bars of section A and then the whole of section B. One surprise, is that the introduction is played in a completely different key from any of the other versions. The song is notated in D (although the melody is very chromatic), but Canaro plays the whole introduction in G. Initially, I assumed that this was to suit the vocal range of Roldán (a baritone), as the other singers were tenors – but no, right at the end of the introduction, Canaro’s arrangement makes an audacious modulation (changing key), and Roldán enters on exactly the same A as all the others. Once you’ve noticed, it’s a striking effect, but you’d easily miss it. Here’s Canaro’s introduction, ending with the entry of the singer:

Uno: Canaro (Roldán), 1943 – Introduction.

The piano plays the eight bars of section A, accompanied by pizzicato (plucked) strings, and then a bandoneon solo introduces section B. At 1:09, this is the shortest of the introductions.

Troilo gives us the whole of sections A & B. The first section is 44s, so the whole introduction lasts 1:41: almost half the duration of the song. The arrangement (made by Astor Piazzolla) is much more sophisticated than Canaro’s. Troilo’s orchestra always created a unique sound: it sounds much bigger than it was. Its range of dynamics seems greater than anyone else (except perhaps Pugliese) could achieve – and he preferred the sound engineers to place their microphones slightly further from the instruments than his rivals. The recording technology of the day was stretched to the limits of what could be captured, but this is a magnificent sound – with Marino on top form, and a string section with a nearly full orchestral complement of violins I & II, cello & bass. It’s worth considering that this introduction is so complete and satisfying, that if there had been no singer, at all, no one would have complained. As it is, the delivery of Marino lifts the song to a special level, magnificently supported by the whole orchestra (and they are singing too: the Troilo trademark sound). Just wow.

Uno: Troilo (Marino), 1943 – Introduction.

D’Arienzo’s introduction plays exactly the same musical material as Troilo’s and is comparable in duration at 1:37. The arrangement lacks the subtlety of Troilo’s, and the piano (in particular) drives the performance in a way that doesn’t sit quite right with the material – but goodness, this is very unlike the D’Arienzo of the 1935-39 orchestra. There is light and shade in the playing, and section B begins with a lovely bandoneon solo. Mauré had a strong effect on the orchestra’s sound but Mauré only enters right at the very end of this introduction. The lyrical elements (and they are there in abundance) are produced by D’Arienzo and his regular players. Great stuff – but for me, anyway, while this is not in Troilo’s league (but then, I’d say that of D’Arienzo, full stop), it is perhaps the greatest recording of the D’Arienzo/Mauré partnership.

Uno: D’Arienzo (Mauré), 1943 – Introduction.

Once again, Fresedo presents the whole of sections A & B in his introduction, which also lasts 1:37. You might be forgiven for expecting something very typically sugary, even lightweight, from Fresedo – at least, in comparison with the others – but no, the arrangement is within the bounds of good taste (always a risk, with Fresedo from the 1940s onwards), and there is playing of great sensitivity and beauty, particularly from the piano.

Uno: Fresedo (Serpa), 1943 – Introduction.

So that just leaves section B. Each orchestra has already played through all of the thematic material, and it is the turn of the singer to present the second verse of the lyric, with the orchestra in a supporting role.

Roldán, for Canaro, continues to do a good job, but the arrangement sounds too cheerful, even jaunty – missing the mark. Roldán gives us the second verse, and then there is an orchestral interlude, taken from the middle of the verse, with Roldán returning with a repeat the end of the last lines – and that’s it. Ultimately, the performance doesn’t quite convince me.

Uno: Canaro (Roldán), 1943 – Section B.

Troilo’s arrangement has Marino sing straight through section B just once – there are no orchestral interludes or coda. We have been given all of the thematic material twice though: once orchestrally, and now, vocally. All is in perfect balance and it requires nothing more.

Uno: Troilo (Marino), 1943 – Section B.

Like Troilo, D’Arienzo takes section B straight through. The sound is out of balance, though, favouring Salamanca’s rather strident piano playing over Mauré’s rather understated delivery. I don’t feel able to set it aside, but this isn’t a version I turn to very often.

Uno: D’Arienzo (Mauré), 1943 – Section B.

Fresedo also plays section B straight through, without repeats. Serpa makes a beautiful sound, but I’m not sure he quite gets under the skin of the lyric. The orchestral arrangement, relying heavily on the piano’s role, complements the singer, but this doesn’t carry the depth of Troilo’s interpretation with Marino.

Uno: Fresedo (Serpa), 1943 – Section B.


Aníbal Troilo
Aníbal Troilo

Troilo’s version with Marino is my preferred choice: it has a depth and quality that the others can’t match. It’s a great song, but everyone still wants to hear the upbeat 1941 Troilo instead, either instrumentally, or with Fiorentino.

D’Arienzo’s version with Mauré would be my second choice. There are not many D’Arienzo/Mauré songs widely available in decent fidelity (and there are several transfers of Uno doing the rounds that are transferred far too fast, and which sound rather silly*). I have a soft spot for the Fresedo version. It doesn’t plumb the depths, but isn’t lightweight, either. I can take or leave the Canaro, but I’m not really a fan of Canaro from the 1940s onwards, generally.

Caveat emptor

*I did a quick check on downloadable versions of the D’Arienzo recording. Only a couple are at anything like the right speed/pitch (my library version comes in at 3:17 tuned to A = 440 Hz). This one managed to fit into 2:59 and needed slowing down by no less than 9.5% to get to concert pitch. It still sounded awful, too.

Uno: D’Arienzo (Mauré), 1939.


¡viejo Discepolín!

Tango 500 Book
Tango 500: the book.

This is the first in an occasional series about tango songs that I believe have special qualities but are only infrequently heard. I love the music of Aníbal Troilo and one of his little-heard masterpieces, Discepolín, is my first choice for this series.

Homero Manzi

The music was composed by Troilo, with a wonderful lyric by his friend and regular collaborator, Homero Manzi. It was written as a tribute to the great tango composer and lyricist, Enrique Discépolo, who had turned 50 in March 1951. Tragedy makes the song particularly poignant, as Manzi was to die of cancer on 3 May 1951, just days before the first recording of the song and Discépolo died on 23 December in that same year from a stroke.

Enrique Discépolo wrote the music and/or the lyrics of lots of very familiar tango compositions. His works include Yira, yira, Confesión, Esta noche me emborracho, Sueño de juventud, Cafetín de Buenos Aires, Carrillón de la Merced, Cambalache, Secreto, Alma del bandoneón, Soy un arlequín, Canción desesperada, Condena, Tormenta, Mensaje and dozens of others.

Enrique Discépolo

The translation of the lyric for Discepolín into English (below) is a machine translation made by Google Translate, with just a few minor amendments. You can find others, online, but I’m regularly surprised at what a good job Google Translate does, even with poetry.


There are three recordings by dance orchestras and they all date from 1951. The earliest was not Troilo’s own but was made by the orchestra of Enrique Francini & Armando Pontier together with a little-known singer, Héctor Montes. The recording was made on 11 May 1951 (for Victor) and it has the best sound of the three. Troilo’s own recording, with Raúl Berón on vocals, was made on 29 May 1951 (for TK), and while the recording must be considered definitive, the sound quality is terrible. Osvaldo Fresedo made the last recording, with Héctor Pacheco on 13 June 1951 (for Columbia), and the sound is decent. There is also a much later recording by Orquesta Típica Porteña and Roberto Goyeneche in 1976, but I’ll not consider it further, here.

For the following extracts, I have taken versions of each song from my own library and rendered them at the same pitch, processed them to reduce clicks & crackle and matched the gain levels as far as possible. Finally, because the sound of the TK recording is so poor and constricted, I have applied pseudo-stereo processing to it, which opens up the texture a little (which helps on headphones, in particular), and for consistency, have done the same for the others. To keep file sizes small (particularly as they are in ‘stereo’), I have rendered them as 128 kbps MP3 files.

The duration of the recordings varies considerably. The Francini-Pontier version lasts 2:58, Fresedo’s is 3:12 while Troilo’s is significantly longer at 3:39. This is partly a matter of the chosen pace, but mainly because of the form or structure of the arrangements.

In the early 1940s any Troilo vocal arrangement would have the orchestra play through the thematic material (verse & chorus, usually 8 bars + 8 bars) before the singer delivered the first verse (and chorus) of the lyric. The orchestra would return for 16 bars (meaning that the second verse of the lyric was never sung), and then the singer would be heard again, with the arrangement wrapping up soon after. By Discepolín, things were less rigidly fixed. Troilo wrote three 8-bar thematic sections: let’s call them A, B & C. The lyric has three verses (each set to 16 bars of music: sections A + B). Section C is only heard orchestrally, but all three verses of the lyric are sung in each of the arrangements. There are interesting variations in the structure of all three arrangements, as you will hear.


The most straightforward arrangement is that by Francini-Pontier. It opens with a short 2-bar introduction and then the three sections A, B & C are played through by the orchestra:

The first two sections are an exchange between the strings and the piano, but I can’t claim to like the way either are arranged much. There’s a moment from 0:29 when the violins briefly adopt a Pugliese-like Yumba figure (and more of that, later) and the introduction ends with very short solos for bandoneón and then violin (presumably, Pontier & Francini playing). At 53s, that’s it: and the singer enters with the first verse.

Troilo’s introduction is straight away much darker in colour and more dramatic:

Section A opens with strings and bandoneóns playing together, creating the rich and complex timbre that is unmistakeably Troilo; it ends with a sweet-toned violin solo. The piano initially carries the melody for section B, supported by the strings, and the section ends with full orchestra and a rising figure for cello right at the end.

That’s not the end of the introduction, though, because he now repeats both sections A & B and adds the additional section C, although it is not heard, again:

Section A builds to a climax, which melts away into section B and a beautiful but short bandoneon solo (Troilo, himself?) The full orchestra begins section C, which ends with another violin solo which balances the earlier one. The whole introduction has the structure ABABC and lasts 1:29.

Fresedo adopts the same structure for his orchestral introduction, beginning with sections A & B:

The strings are in the foreground, throughout, with a prominent role for piano.

He repeats sections A & B with little change in orchestration, but we don’t get section C (so it is missing from the arrangement, altogether), but instead, a 4-bar linking passage leading to the first sung verse quoting from Discépolo’s own composition, Cambalache:

First verse

On the frozen marble, croissant crumbs

and an absurd woman who eats in a corner ...

Your muse is bleeding and she has breakfast ...

the dawn does not forgive nor does it have a heart.

In the end, who is guilty of the grotesque life

and of the soul stained with carmine blood?

It is better that we leave before dawn,

before we cry, old Discepolin! ...

After a lightweight introduction, Francini-Pontier surprises us with the rich and slightly plummy baritone voice of Héctor Montes:

Troilo and Berón give the verse more colour, beginning quietly and building to a climax before falling back with wonderful word-setting:

It is sometimes said that Raúl Berón was at his best during his years with Troilo, and certainly he brings a depth to this interpretation that was not always apparent in earlier years. I don’t always enjoy Berón’s rather nasal crooning, but he musters some power, here, even if the sotto voce opening sounds as though he is struggling, a bit, at the bottom of his vocal range.

Fresedo accompanies Pacheco mainly with the strings, and the arrangement has real depth and subtlety:

Pacheco’s vocal delivery is a surprise; he has a fairly light, tenor voice, and yet he is more comfortable with the opening phrase and builds just as powerful a climax as either of the others. Methinks he is underrated, and I have been enjoying many of his other recordings of this period with Fresedo, too.

Second Verse

I know of your long boredom

and I understand what it costs to be happy,

and with the sound of each tango I feel your presence

with your enormous talent, and your nose;

with your bitter and hidden tears,

with your pale clown mask,

and with that sad smile

that flourishes in verse and song.

Continuing with Fresedo and Pacheco:

There is some lovely word-painting, here, and the orchestral writing balances the singer with interesting rhythms and colour that points up the lyric in a most affecting way.

Final verse

People come to you with their pile of sorrows

and you caress them with almost a tremor ...

It hurts as your own, the scar of others:

he had no luck and she didn't find love.

The ronda has been packed by the sound of the orchestra

They embrace under the spotlight like sawdust dolls ...

Can’t you see they're dancing?

Can’t you see they're partying?

Come on, everything hurts, old Discepolín ...

For the final verse, Francini-Pontier’s arrangement is nearly but not quite up to the job of supporting the singer:

Montes delivers the verse, and the climax of the song, with conviction, but overall he doesn’t quite have the depth of interpretation to pull it off, and I’m left thinking that everything here adds up to slightly less than the sum of its parts. There’s a nice touch, though, right at the end, with a short orchestral coda, with fleeting references to two other Discépolo compositions, Soy un arlequín and Yira yira.

Fresedo builds a satisfying climax, but with string writing that nearly gets carried away:

Pacheco’s voice is well-suited to the shape of the melody, rising in pitch and then falling back. He delivers the last line of the lyric with almost no accompaniment, and then orchestra ends the arrangement with some very bluesy chord progressions.

Troilo delays the final verse by adding a short linking section quoting from Uno, a song with a Discépolo lyric (but, ironically, the music being quoted was by Mariano Mores):

Berón builds the tension throughout section A, but the climax comes in the final section supported by driving chords played with strong arrastre (that Pugliese-like Yumba effect, again). I wonder whether the arranger of the Francini-Pontier version heard Troilo’s interpretation, played live, before the recording and ‘borrowed’ the idea. The climax melts away, and the strings, dying away, support Berón’s voice first with tremolo and then pizzicato chords. The song ends simply, with a perfect cadence.


Montes’ singing is the relative strength of the Francini-Pontier version. He only recorded one other song (a vals, Una triste verdad) and then got married and quit. The style of the orchestra is not really to my taste and the singer is very forward in the sound balance. I can’t see me ever choosing to play it for dancing (to say nothing of the challenge of finding a home for it in a good tanda).

I’m torn between the other two versions: Troilo ought to be the clear winner: it is the composer’s own interpretation, with a brilliant arrangement – a big, dramatic sound. I have reservations about Berón’s crooning and perhaps wish that Troilo had recorded it with Edmundo Rivero, instead.

It’s easy to dismiss Fresedo in this period as a purveyor of lightweight lyrical repertoire, but there’s more to him than that. This arrangement is inventive and well-played. Yes, it has many of the Fresedo trademark sounds (vibraphone and string glissandi) that make you either smile or wince (at your choice). Pacheco has a light tenor voice and yet he carries the lyric with sensitivity and conviction. I like it a lot.

Still torn, I can’t name a favourite: Troilo is poorly served by the recording, but the arrangement and performance have real gravitas and the music-making is of the highest quality. Fresedo has a different feel, but no less valid. I must have them both.

Here are the complete songs, including some unwelcome pitch variations:

A week of preparations for a DJ Set

DJ Clive Harrison
DJ Clive Harrison

Every DJ has their own way of putting together a DJ set. Some plan everything in advance (or do so some of the time); others might just choose a few opening tandas (and they might not do even that in advance) and then select music as an event unfolds – aiming to choose music that seems to best reflect or serve the dancers’ preferences/needs in the moment.

Quite a lot of completely bogus claims are made for the superiority of the latter approach, and there are not many DJs around that can make consistently better choices on the fly. My advice to inexperienced DJs has always been to compile sets in advance until you are sure that you can produce better ones, live. And, of course, there are lots of points between the two extremes, and no intrinsically right solution.

I tend to compile a complete set in advance, so that I can have it duplicated to my phone (as a backup device) in the event of equipment failure. I have had to use my backup, once; and as we were only twenty minutes in, the evening would have come to a rather premature end had I not had it.

As often as not, I end up using the set I had compiled, largely without change – but I’m usually DJing in a familiar venue, and I know what to expect. I can, and do, change sets on the fly, but often the changes are minor ones – substituting a tanda here or there – rather than going off on a tangent that has no way back, and demanding continued fresh choices for the rest of an event. One of the fallacies of ‘live’ DJing is that a DJ can magically ‘read’ the energy of the dancefloor, and ‘know’ what the dancers’ current response means in terms of their next preference. Individual dancer’s responses are rarely homogeneous, in that way, anyway, and ‘reading’ the response to ‘A’, rarely says anything useful about the likely response to ‘B’.

If you play a tanda that clearly has not hit the spot, the signs may be all too obvious: there are fewer couples dancing than you had expected, and the noise of conversation may have risen above the norm. Xyz, over in the corner, has shot you a dirty ‘WTF’ look – and you know you have erred. The obvious solution, is not to immediately play a similar tanda by the same orchestra – but surely no one, anywhere, ever, was going to do just that. You have to pick yourself up, and perhaps check whether that next tanda, already planned, really is a good idea (and change it, if you think not) – and that’s about the best you can do.

So in the period leading up to an event, most DJs will at least be thinking about what they might play. Really nervous ones, new to the game, may have planned everything in detail, weeks before. I used to be like that – but it gets easier – and I reckon that I take about half an hour, these days, to compile a four-hour set. If I am doing it ‘live’, under the pressure of knowing that the tanda now playing will end in six minutes, oops, no, now four, and that I have to select something – I find that the task will take all my attention, pretty well, all night. And I don’t believe that I make better choices, most of the time, and feel quite unembarrassed to say so.

Saturday: Una emoción

There’s a week to go before the next Milonga at Redmarley and I’m putting together an outline of my DJ set – and the challenge is always to whittle down all the possibilities to just four hours of great music. My sets are created more-or-less to a formula: a certain proportion of different musical styles, periods, orchestras and singers. The thing with a formula is that it can easily become formulaic, but I try and play up the creativity of the process. There are obvious and easy choices, but you can’t play the same handful of ‘greatest hits’ all the time, but no one will thank you for playing a whole load of unknown repertoire that you have ‘discovered’ either. It’s harder than you might guess.

Anyway, no milonga could omit a tanda from Tanturi. The choices are several: he worked with two really great singers, Castillo and then Campos. Tango enthusiasts usually have a very firm and settled view over their preference. Mine is for Campos (but I play and enjoy both). After Campos, came other, lesser, singers, and generally I don’t play those. If there’s only time for one Tanturi tanda, why play a 2nd rate one? The few instrumentals are good too, but there aren’t many, so you can only play them now and again.

The tanda I have chosen opens with Una emoción, recorded in 1943 – a fantastic song and great for dancing. The other songs all come from a tight time period: August to November 1943. I have always believed that the most powerful building-block of a good tanda is the relationship between the songs, and choosing repertoire recorded in a short time period gives you a coherent style and feel that is hard to beat. Other DJs have a different approach – and in matters of taste, there can be no right or wrong – but it’s a good thing that individual DJs develop a strong musical character and that dancers get to know what to expect (at least in general terms).

Sunday: La capilla blanca

Nearly every DJ set I play has a tanda featuring the golden voice of Alberto Podestá. He recorded with several leading orchestras, most notably Di Sarli, Caló and Laurenz, so it’s no hardship. La capilla blanca, recorded with Di Sarli in 1944 is one of the highlights of the repertoire of the mid-40s – sophisticated and urbane music that just makes your heart melt.

Di Sarli recorded the song, again, in 1952 with Mario Pomar, and that is a very fine version too. Choices, choices …

I give the songs in my tango library a star rating, and unlike TripAdvisor reviews, the five-star ratings in my collection are very few and far between. A five-star song is very special, usually in more than one way. I just checked: I have just forty five of them (out of several thousand). All four of the songs in the Di Sarli/Podestá tanda in my forthcoming set at Redmarley on Saturday have five, precious, stars. In fact, it would be worth going just to dance them – but just make sure that when the preceding cortina fades, you’re not one of the lost souls who are checking your phone, chatting to your friends and oblivious to that mirada from your favourite partner, or otherwise not focussed on the reason to you went: to dance La capilla blanca.

Monday: Pobre novia

Nearly everyone loves valses. Pobre novia was recorded in 1955 by the orchestra of Domingo Federico with singer, Armando Moreno (who is more usually associated with the orchestra of Enrique Rodríguez). Federico’s orchestra is not very well known, but after leaving the orchestra of Miguel Caló in 1944, he produced a steady stream of fine recordings for just over a decade (and intermittently, for years after that) – and they deserve to be heard more often than they are.

There’ll be time for four vals tandas in my DJ set at Redmarley on Saturday, and I’m planning a nice spread of styles and periods, with vals recordings spanning 1932-1955. One of those tandas will be by Caló, providing a point of stylistic similarity with Federico, while the others are well-contrasted. Juxtaposing the familiar with the less so, making connections through linked musicians – putting together a set that is both varied and coherent – is part of the challenge and fun of constructing a DJ set.

Of course, you can (and should) just take the music at face value and enjoy the dancing, but many sets are built around one or more themes, and when you are DJing regularly at one venue (as I am currently doing at Redmarley), there is the opportunity for a theme or programming idea to run through several sets – although I’m probably the only person aware of the underlying structure. I keep notes, and know exactly what I have played, where and when. I guess that makes me a nerd: Oh well…

Tuesday: Maragata

For a tango artist of the first rank like Aníbal Troilo – who produced recordings of the highest quality from 1938 to 1971 – it seems a bit odd that dancers and DJs seem to favour the recordings from just one year: 1941, mostly vocals, featuring the voice of Francisco Fiorentino. I checked my own notes. I have played forty Troilo songs over the last twelve months, recorded between 1938 and 1949 and yet no fewer than fourteen of them were recorded in 1941 (and I play a wider range than many DJs). This thing is that they are so good! He had one of the greatest vocal partnerships in tango: Troilo and Fiorentino virtually invented the cantor de orquesta role (certainly no one did it better) and the music perfectly captures the upbeat and exciting pace of tango music at that time.

Tango music slowed down, considerably, over the following couple of years and strangely enough, I have only played a handful of songs from 1942 or 1943 in the last year. It seems that I steer away from his transitional phase – although I hadn’t really been conscious of doing so. By 1944, of course, Fiorentino had left Troilo for a solo career – the vocalists over the next few years were Alberto Marino and Floreal Ruíz – but the flame never burned as brightly, again. Other notable singers followed, but by the late-1940s Troilo’s orchestra was developing in directions that didn’t much suit social dancers, and my interest wanes to almost nothing. So 1941 it is for my set at Redmarley on Saturday.

The ‘Troilo tanda’ (if there is to be only one) is often the highlight of a DJ’s set (at least for me – perhaps its just me?). Rather like Pugliese (and for many of the same reasons), Troilo has to be sequenced very carefully as the music has such a strong character. Maragata is a wonderful song, strongly rhythmical and with lots of syncopated accents. Every section of the orchestra is on top form and Fiorentino makes it all sound so easy. Great stuff!

Wednesday: Ventanita de arrabal

In about 1926 the ‘new’ electrical recording technology reached Argentina and suddenly the quality of sound recordings was transformed into something that could sound really very good, compared with being invariably awful. A song that received multiple recordings in 1927 was Ventanita de arrabal: two by singers Corsini and Gardel (with guitar accompaniment), and three (purely instrumental) by the orchestras of Canaro, Lomuto and Maglio. It’s a pleasant but unremarkable song; the studios produced hundreds of such recordings, every year, and there was nothing especially distinctive about more than a handful of them – they are largely unplayed today.

Ventanita de arrabal seemed to disappear into obscurity, but then in 1950 it resurfaced in several new vocal recordings, most notably with Pugliese (Vidal), but also by Del Piano (Vargas) and Pedevilla (Serpa). Troilo recorded it in 1952 (Casal) and again in 1965 (Reyes). There were others, too, so the song obviously had qualities that were recognised by at least two generations of tango musicians, including several of the first rank.

Some DJs play quite a lot of ‘early tango’. They have their fans, but they are in a minority. Some will play nothing recorded before 1935 (tango ‘begins’ with D’Arienzo, you see, and doesn’t seem to last very long, as they won’t play anything much after 1945, either). Thankfully, they are in a minority too. Even within the usually accepted limits of ‘traditional’ tango music there is a wide variety of music styles and periods, spanning the very earliest electrical recordings, right through to the beginning of the digital recording era. Some tango musicians, like Pugliese and Fresedo were around for the whole of that time. Of the music I select to play in my DJ sets, a high proportion (around two thirds) is drawn from the decade 1935-44, but I always play a tanda or two of earlier music and a slightly higher ‘quota’ from the later period (which is much longer, in duration).

At the end of the evening at Redmarley on Saturday, I’m planning to play a really lovely instrumental tanda from the sextet of Juan Maglio (Pacho) – one of tango’s real old-timers – and the tanda opens with Ventanita de arrabal. For the recording date, the sound is more than acceptable, but I have done quite a lot of work on making it sound better. For all sorts of reasons, the pitch at which many vintage recordings were made was not always faithfully reproduced on shellacs played at 78rpm; and as the music captured on shellacs was later transferred to vinyl, and then CD, further pitch errors crept in. Probably more than 80% of commercial tango recordings available today are audibly out of tune (really, and sometimes laughably so). Everything that I play has been pitch-checked (and corrected), and I also use specialist software to help minimise the clicks and crackle that are the inevitable product of the shellac medium, as the clicks are nothing to do with what could be heard in the recording studio.

I’m pretty happy with the sound of this Maglio tanda, but as an experiment, I have gone further and processed the sound in pseudo-stereo. It opens up sometimes-congested musical textures and creates a feeling of space around the instruments. I want to see how an effect that can be heard in a domestic setting transfers to a larger venue – and if you’re there, and notice, I’d be interested to know what you think of the finished result. Of course, if you just don’t like early music, you can take the announcement of the last tanda as your signal to change shoes and leave, and then you won’t have to face the dilemma of whether also to stay an extra couple of minutes to help stack a few chairs at the end of the evening.

Thursday: El porteñito

Lots of tango dancers are a bit wary of milonga tandas – and then along comes D’Agostino.

El porteñito is perfect in every way: a 5-star jewel of a milonga. It takes a very steady pace and its gentle rhythms are just irresistible. If you don’t want to dance it, there’s something very wrong with you, or you’ve just done one milonga workshop too many, lately.

Almost everyone can identify the unique sound of the orchestra of Ángel D’Agostino (even if they couldn’t put a name to it) in about two seconds flat – it’s utterly distinctive. The partnership of the two Angels: Ángel D’Agostino and his best singer, Ángel Vargas, was almost remarkable – there are no duds – and the orchestra is rightly a firm favourite of many discerning dancers for its subtlety and finesse. These are not qualities normally associated with the milonga genre, but D’Agostino’s touch was sure and the magic just works.

Whatever you do, don’t sit out the tanda at Redmarley on Saturday.

The Redmarley milonga is usually held in the evening of the 2nd Saturday of the month at The Village Hall in Redmarley D’Abitot in rural Gloucestershire (GL19 3HS). It’s easy to reach from a wide area, being just five minutes from the junction of the M50 (J2) with the Ledbury Road (A417).

Sonata and Temo

Tango 500 Book
Tango 500: the book.

One of the vals tandas included in my book, Tango 500, opens with a song called Sonata sung by its composer Augstín Magaldi. The tanda continues with Sin rumbo fijo (with Ángel Vargas) and ends with the wonderful Temo (with Mario Corrales). The orchestra is Orquesta Típica Victor under the direction of Federico Scorticati – or is it?

Well, as far as Sonata is concerned, no, it probably isn’t. attributes it to Orquesta Típica Victor, giving the recording date as 3 June 1937, but the song is missing from other OTV discographies. The original shellac disc label says solo con orquesta, but doesn’t name the ensemble. The A-side of the same disc was also a Magaldi recording, but that’s with guitar accompaniment.

OTV was a recording orchestra without fixed personnel. It had a studio session on 26 May and another on 8 July, each producing two recordings, but nothing else was recorded on 3 June. The orchestra usually recorded with ten to twelve players, but the ensemble from 3 June sounds smaller. It seems likely that a handful of players were hired for the session, drawn from the pool of OTV regulars. If anyone has further information, I’d be interested, but it’s a nice song, anyway. Here’s an excerpt:


Many tango recordings are published at the wrong pitch/speed. All three of the songs in this tanda require retuning to return them to concert pitch. I analysed the tuning of my transfer of Sonata using the nnls-chroma tuning plugin with the Audacity sound editor and found A = 432.8 Hz (and I am assuming that A = 435 would still have been in use at the time of recording). With some rounding of the values, the published pitch is 9¢ (0.5%) flat, a small (but audible) variance. Correcting it shaves one second off the song duration, and most people wouldn’t notice it – the variance is only just above the threshold for hearing a pitch difference.

I also checked Sin rumbo fijo and found A = 446.9. That is a much more significant 47¢ (2.6%) sharp. The variance is nearly half a semitone, so my transfer (Euro Records Coleccion 78 rpm) is almost exactly in the middle of any key you would find on a piano tuned to A = 435. Correcting the error adds four seconds to the transfer but, more significantly, has a very clear effect on the pace of the song and of the timbre of the singer’s voice. It sounds much better.

Here are two 15s excerpts, before and after:

Lastly, I checked Temo (also Euro Records Coleccion 78 rpm), although I already knew that something was very wrong. In the opening phrase of the introduction, the violin vibrato is obviously too fast, and when the singer enters we hear a rather feeble tenor with an over-fast vibrato, rather than the baritone, Mario Corrales (better known as Mario Pomar from his 1950s recordings with Di Sarli). The measured pitch was A = 445.6 Hz (42¢ fast), almost the same as for Sin rumbo fijo, but correcting it left the transfer still sounding odd and Corrales’ voice barely recognisable. There’s no harm in trying alternatives, so rather than retuning down to the nearest key, I corrected by 142¢ or 7.8% (another complete semitone), and listened again.

Straight away, the violin tone sounds right, and at the singer’s entry, I recognise Corrales’ voice. The pace of the song is very different, too, of course. Here are three 45s excerpts, the original pitch, corrected by 42¢ and then by 142¢.

One final issue with Temo is that by 1940 (the recording date) most orchestras in BsAs had retuned to modern concert pitch, A = 440 Hz. I’ve never seen any source commenting on the changeover date, orchestra by orchestra, so there remains a question over whether OTV were still recording at A = 435, or had already changed. The pitch difference is 20¢ (a fifth of a semitone) – quite audible, but not dramatic. If I’m playing Temo in a tanda with earlier valses, tuning to A = 435 means no jarring clash of relative pitch, but the required speed variance is reduced slightly at the higher pitch.

I don’t know whether I’m right. It seems extraordinary that such a massive pitch error could creep into a published transfer (and as far as I can see, all the alternatives fall into line*), and yet the timbre of Corrales’ voice sounds right at the slower speed. At the slower pace, lots of little details in the violin parts like vibrato and bowing articulations sound so much more natural. The tempo is still 204 bpm, so by no means a ‘slow’ vals. The original rattles along at 220 bpm, which is quick.

[Edit: *The database lists one version (sourced from a private collector) timed at 3:07, compared with the 2:54 of the Euro Records transfer or 2:55 for CdT, and while I haven’t heard it, I have to assume that it, too, is tuned down a whole semitone.]

Sangre de mi sangre

Tango 500 Book
My new book, Tango 500, is available now from Amazon.

One of the best of the recordings made by Rodolfo Biagi with Hugo Duval in the mid-1950s was Sangre de mi sangre, recorded for Odeon on 29 June 1954. There seem to be lots of versions available on the usual music listening sites and for download, but most of them turn out to be the much later version Duval recorded with Trío Yumba, and that’s not really my cup of tea.

I have only found the Biagi original in two distinct transfers (but they are widely duplicated); one is from the album Tango Classics 095: La copa del olvido and it has rather a lot of noise but sounds dull, while the other is from the album Sangre de mi sangre (1953 – 1959), which sounds rather better. Here are the two 30s clips available to preview the tracks:

The clips end just as Duval begins to sing (which is a shame) but the pitch/speed is very obviously too high/fast in both transfers. I measured the pitch in Audacity, using the nnls-chroma tuning plugin. The second clip reported tuning of A=431.7 Hz, suggesting that the tuning was 33¢ flat. However, as far as I can determine, the plugin is not reporting absolute pitch, but the variance from the closest standard key (ie the maximum it will report is 50¢ up or down). Where the tuning error is more than half a semitone, it may report 33¢ flat, when it should report 67¢ sharp, so judgement is required in interpreting the measured result. The correction required, here, is to stretch the sample by 3.795%. Not surprisingly, it now sounds rather better:

There is quite a lot of mains hum, so removing that (I use Brian Davies’ DeNoiseLF), improves the sound again. Here’s another 30s sample, but this time, picking up where the others finished – to hear the effect of the pitch correction on the timbre of Duval’s voice:

Of course, it shouldn’t be necessary to check the music files we buy for anything as basic as correct pitch, but it is comparatively rare to find something so far out of tune: a correction of nearly 4% completely alters the pace and feel of the song.


I do most of my listening tests on headphones, as it’s easier to hear fine detail, but I don’t much like the ‘inside your head’ feeling I get when listening to mono recordings on headphones.

Just for fun, I sometimes render a song in pseudo-stereo, and I’ve been amazed at how it can bring a ‘difficult’ transfer to life. The effect is completely phoney, of course, and I’d never use it for my DJ library.

I combine two different effects: Audacity’s mda-stereo plugin and also one of the reverb presets (small room bright). I won’t bore you with the details of exactly how I do it, but here’s a 60s excerpt (combining both of the above samples) in ‘stereo’. You need headphones, really, to ‘enjoy’ the effect – but I think it actually sounds quite good, and it’s certainly good to hear the song in the proper key.

DJ Clive Harrison
DJ Clive Harrison

Barrio de tango is the tango blog and online home of tango DJ, Clive Harrison, based in the English Midlands. Now retired from teaching and hosting dance events, Clive remains available to DJ, playing exclusively traditional tango music from the great tango orchestras.

RCA 100 Años – Carlos Di Sarli (Revised)

Many people will have this album, either as a CD or download. I bought mine to fill some gaps in my collection of the late Di Sarli instrumentals, and so I was only ever interested in certain tracks. Recently, I noticed an apparent pitch discrepancy in one track, and that caused me to take out the original CD again and check the rest: I was quite surprised at what I found.


Four songs: A la gran muñeca, El amanecerEl ingeniero and Organito de la tarde seem to have unpleasant added reverberation. I already had these songs (dry) on the ‘Solo Tango: Carlos Di Sarli – Instrumental (v1)’ CD, and here are short excerpts from each song to compare, taken from the two releases.

A la gran muñeca, 1954

El amanecer, 1954

El ingeniero, 1955

Organito de la tarde, 1954

Surely, these are the same recordings? If the Solo Tango transfers are ‘dry’, the RCA 100 Años must be at least one generation (further) removed from the master tapes. Incidentally, the same reverberation can be heard on these songs on the ‘Involvidables – Carlos Di Sarli’ CD.

As a secondary matter, all four songs appear to have been transferred at the wrong speed & pitch.

Pitch Variations

It is obvious, listening to the opening seconds of a handful of tracks on the CD that there are irregular pitch variations. With standard pitch (A = 440 Hz), you can play along (on a keyboard, perhaps) and soon find the key of any piece; except that the majority of the tracks have pitches that fall between the available range of notes – they’re out of tune. It’s possible, but unlikely, that the files have been processed to alter the pitch independently of the speed. It is more likely that the mastering has been done with insufficient care to preserve the pitch properly (ie that at the point the tape was digitised, the tape machine was running at the wrong speed).

I originally didn’t spend a huge amount of time on the exercise, but the pitch/speed variations seemed to lie in the range of 2-3% – more than enough to make a significant and audible difference to the pace of the performances, regardless of whether the actual pitch difference was troubling. My attention was then directed to some software called beaTunes which (among other functions) can analyse a file and determine the key in which the music is arranged and the degree (if any) to which it is out of tune (measured in cents: 100th of a semi-tone).

On checking the whole CD, I found pitch variations in the range of just a few cents (but very few people can hear a pitch discrepancy of 5 cents or less), right up to 67 cents (-3.79%). Here are a few examples:

A la gran muñeca, 1954 (Solo Tango) -1.94% (34¢)

Don Juan, 1955 (RCA 100 Años) -1.83% (32¢)

El amanecer, 1954 (Solo Tango) -1.55% (27¢)

El ingeniero, 1955 (Solo Tango) -3.79% (67¢)

El once, 1954 (RCA 100 Años) -1.89% (33¢)

La morocha, 1954 (RCA 100 Años) -2% (35¢)

Milonguero viejo, 1955  (RCA 100 Años) -1.94% (34¢)

Organito de la tarde, 1954 (Solo Tango) -2.06% (36¢)

If these transfers were derived directly from the original master tapes (as I have heard claimed) then I don’t think that very much care was taken in the process. Thirteen out of twenty-two tracks were sufficiently out of tune for the pitch difference to be audible.

I have rendered ‘new’ versions for my own library (which was tedious and time-consuming) but I’m always looking for the very best sound and will try and convince myself that it was worth the effort.

DJ Clive Harrison
DJ Clive Harrison

Barrio de tango is the tango blog and online home of tango DJ, Clive Harrison, based in the English Midlands. Now retired from teaching and hosting dance events, Clive remains available to DJ, playing exclusively traditional tango music from the great tango orchestras.

Carlos Di Sarli – El señor del tango


Carlos Di Sarli was born in 1903 in Bahía Blanca. He was a pianist and played in several orchestras (including that of Osvaldo Fresedo) before forming his own orchestra in 1928. He was active in the recording studio from the early days of electrical recording in the late 20s, throughout the Golden Age and into the late 50s.

The standard of music-making was uniformly high, and while his distinctive style developed slowly over many years, he produced great music for dancing throughout his career.

Early Period

Carlos Di Sarli

Di Sarli was just twenty-five when he entered the recording studio for the first time for Víctor, cutting two sides in November 1928. Twenty-two songs followed in 1929, sixteen in 1930, eight more in 1931 – and then there was nothing until 1939.

The early recordings were all made with just six players and they sound very small-scale. The sound balance favours the violins. The bass is surprisingly clear, but the piano (Di Sarli, himself) sounds as though it was in the next room. Recording technology was very primitive and we have to live with restricted sound, but if you can find transfers that have been carefully made and turn up the volume a little, you may be surprised at how much musical vitality is coming out of your speakers.

Di Sarli (Instrumental) 1929-30
  • Pobre yo > Belén > No te aguanto más > No cantes victoria

Ernesto Famá

In September 1930, Di Sarli started recording with tenor, Ernesto Famá, five years Di Sarli’s junior, but already an old hand in the recording studio, having previously recorded dozens of songs with Osvaldo Fresedo and with Orquesta Típica Victor. At this time the singer’s rôle was limited usually to delivering just one chorus towards the end of the song. Famá’s time with the sextet was short-lived, but he was simultaneously making a name for himself singing with the much larger orchestra of Francisco Canaro.

Di Sarli (Famá) 1930-31
  • Maldita > Chau Pinela > Flora > La estancia 

The Forties

Carlos Di Sarli
Carlos Di Sarli

In the last days of 1939, Di Sarli was back with a new orchestra (initially of only eight musicians, but soon to grow to twelve) and a new recording contract with Víctor. With the impetus of the success of Juan D’Arienzo (from 1935) whose orchestra had been built on the foundation of rhythm, the late 30s had been a time of strong growth in the popularity of tango and the number of orchestras. Di Sarli’s answer was not to copy any other orchestra’s style but to introduce his own strong blend of underlying rhythmicity (led from the piano and underpinned by the bandoneons and bass) together with lyricism (led by the violins). The mix favoured rhythm for the first year or so, but as the pace of tango started to slow, generally, Di Sarli responded by moving more and more towards lyricism – stretching out the phrases. The transition was slow and steady and led, inevitably, to the glorious late instrumentals that are so well-loved.

The very first recording session, held on 11 December 1939, was his manifesto: with a marvellous instrumental, El retirao, which bubbles along with the strings almost laughing their way through the initial melody, played staccato. With the improved sound quality now available we also hear clearly the thing that sets Di Sarli’s orchestra apart from any other: the piano playing of Di Sarli, himself. It is usual to say that Di Sarli’s most distinctive feature is the way he uses the violins – the sound of the strings. It’s obviously true that they frequently take the lead in carrying the melody, but music is about much more than melody. Listen to El retirao a couple of times: once to get the general feel of the piece – its mood and vitality – and then again, for Di Sarli’s piano playing. So much of the colour and rhythm is led from the piano. No one else played this way and in my opinion, it is Di Sarli’s piano playing, not his fondness for the sound of violins, that makes Di Sarli unique.

Roberto Rufino
Roberto Rufino

The other song recorded at that first session was Corazón, with the young baritone, Roberto Rufino (not yet eighteen) on vocals. The limited rôle of the estribilista (chorus singer) was giving way to the singer taking a much more prominent position, delivering the lyric but with the voice fully integrated into the sound of the orchestra as though an additional instrument. Rufino’s voice was perfect for this rôle and Di Sarli’s early success was secured by the popularity of this orchestral/vocal partnership.

Di Sarli (Rufino) 1939-40
  • Corazón > Lo pasao pasó > Cosas olvidadas > En un beso la vida

Throughout his working life, Di Sarli recorded a significant number of instrumentals and they are wonderful for dancing. In this relatively early stage, the pace is still quite brisk. But by 1941 all the elements of Di Sarli’s mature style were present and over the following fifteen years he would steadily slow the pace and draw more and more from a growing body of string players. His own piano playing has a new confidence: just listen carefully to La cachila – it is almost a piano solo with orchestral accompaniment.

Di Sarli (Instrumental) 1940-41
  • El incendio > La trilla > La cachila > La morocha

Alberto Podestá, another young baritone, joined the Di Sarli orchestra in 1942 as 2nd singer. There was a considerable rivalry between the two singers, and Podestá left at the end of the year, before returning in 1944 following the departure of Rufino. He recorded twelve songs in 1942 and a further ten in 1944. There isn’t a dud among them and the output of the Di Sarli orchestra of the early 40s sits firmly at the centre of the lyrical tango repertoire.

Di Sarli (Podestá) 1944
  • Tú el cielo y tú > Llueve otra vez > Lloran las campanas > Vamos 

Following the departure of Podestá, Di Sarli engaged another baritone, Jorge Durán, from the beginning of 1945. He stayed with the orchestra for two years (and it was his departure that allowed Podestá to return for his third and final period with the orchestra). Durán had a darker voice than his predecessors and while limitations in the sound quality of the mid-40s technology partly mask its tone, he was to return to the orchestra in 1956 and those recordings are among the highlights of Di Sarli’s late period (and of the 50s in general).

Di Sarli (Durán) 1945
  • Un tango y nada más > Hoy al recordarla > Que no sepan las estrellas > Yo


The Fifties

Transition – Music Hall

After working steadily since 1939, Di Sarli recorded just two sides in 1948 and then there was nothing until 1951 when he returned to the studio with a new record company, Music Hall. He worked with two singers: baritone, Mario Pomar and tenor, Oscar Serpa, although all but one of the 1951 sides are instrumentals.

The Music Hall recordings were some of the first to be released on vinyl (rather than shellac), but the sound quality is variable. Di Sarli was to return to Víctor in 1954. The tape-mastered recordings of his late recordings are of the highest quality but the early 50s material is valued for its musical qualities rather than the sound.

In these recordings we start to hear the characteristics of the late Di Sarli orchestra (post-54) – a greater feeling of space and a balance tipping in favour of the melody and a big string sound (but always with a steady and notably slower beat). The Music Hall recordings are the bridge from the sound of the 40s to the late 50s – although the transition was slow and steady – there is a coherence of fundamental style and many elements of the later style can be heard from the earliest days.

Di Sarli (Pomar) 1952
  • Tangueando te quiero > No me pregunten por qué > Se muere de amor > Domani 

Oscar Serpa had first sung with Di Sarli’s in 1948 and featured regularly in the orchestra’s recording schedule between 1952 and 1955. He had sung for several years with the orchestra of Osvaldo Fresedo, and his voice suited both orchestras very well. This tanda ends with the 1953 recording of Verdemar. Serpa recorded it again in September 1955 – his last recording with the orchestra. The comparison is interesting. The sound quality of the later recording is better, but some of the tenderness of the earlier interpretation is missing. Interestingly, the pace seems slower in the 1952 version too. Di Sarli re-recorded much of his popular repertoire with Víctor, but apart from the better sound, I rarely prefer the later versions.

Di Sarli (Serpa) 1953-54
  • Al compás del corazón > La canción más triste > Buenos Aires, yo te canto > Verdemar 

Late Period

Carlos Di Sarli
Carlos Di Sarli

It is the late instrumentals that many tango dancers particularly associate with Di Sarli’s music. They have a clear pulse, limpid texture and Di Sarli’s piano-playing sparkles with a bell-like clarity while an enlarged violin section carries the melody (and frequently points up the rhythm too). The following version of Milonguero viejo is mislabeled as his 3rd – there were four: 1940, 1944, 1951 and this, from 1955.

Di Sarli (Instrumental) 1955-56
  • Milonguero viejo > Germaine > Los treinta y tres orientales > El jagüel

BMG seems to have a mastering pitch/speed error in Milonguero viejo. The key is  E, and if you bring down the speed by 3% to arrive at the correct pitch the speed difference makes a surprising difference to the pace of the interpretation, as you can hear in this short clip:

Di Sarli recorded for Víctor between 1954 and 1958. The singers were tenors Argentino Ledesma and Roberto Florio and baritones, Rodolfo Galé and Jorge Durán. Here’s a tanda from Florio, stretching the compás to the limit. The violin section has grown to eight players.

Di Sarli recorded a final LP with Polygram in late 1958. He continued to perform, live, during 1959, but then illness forced his retirement. He died in January 1960 a few days after his 57th birthday.

Di Sarli (Florio) 1956-58
  • Derrotado > Soñemos > Nuestra noche > Adiós, corazón


Di Sarli’s recordings fall into four distinct periods and each produced work of the highest quality.

    • The early sextet recordings (1928-31).
    • The middle period (1939-48), with wonderful vocals, particularly from Rufino, but with fine contributions from Podestá and Durán. The early years featured a smaller orchestra and a more rhythmic style of playing. The later years laid the foundations of the slower, more lyrical style. The instrumentals are good too.
    • The transitional period (1951-54) with a fine blend of instrumentals (often in versions sounding fresher than the late period, but with inferior sound) and lovely vocals from Pomar and Serpa.
  • The late period (1954-58), fine instrumentals featuring an enlarged orchestra – eight violins – and a slow pace. Vocals not always as successful. Sound excellent.

More …

In this blog the music of Di Sarli featured regularly in The Tango Years series and also in the Tango 250 series.

DJ Clive Harrison
DJ Clive Harrison

Barrio de tango is the tango blog and online home of tango DJ, Clive Harrison, based in the English Midlands. Now retired from teaching and hosting dance events, Clive remains available to DJ, playing exclusively traditional tango music from the great tango orchestras.

Juan D’Arienzo – El Rey del compás


Juan D'Arienzo
Juan D’Arienzo

Juan D’Arienzo was born in 1900 in Buenos Aires. He was a violinist and a notable composer but he is mostly remembered as an orchestral director. In a long career, he recorded for the Disco Electra label between 1928-29 and for Victor between 1935-75. He died in 1976 aged 75.

He led the rhythmic school of tango music and was very important in popularising social dancing through a reinvigoration of an earlier style of playing with a clear beat and strong rhythms.

The Thirties

There are very few duds in the 1930s D’Arienzo discography. He continued to record until a year before his death, but it is the recordings of 1935-39 that are most frequently heard at milongas, today, and quite a lot of the later repertoire is never played.

Pianist: Lidio Fasoli (1935)

In a 1975 interview, just a month before he died, D’Arienzo said,

The foundation of my orchestra is the piano. I regard it as irreplaceable.

Three different pianists played in the D’Arienzo orchestra between 1935 and 1939. The first, Lidio Fasoli, is the least notable, but the ten recordings made between July and December that year are still very fine. There are five tangos, four valses and a milonga, all instrumentals.

D’Arienzo (Instrumental) 1935

Pianist: Rodolfo Biagi (1935-38)

Rodolfo Biagi replaced Fasoli at the piano after a few months and the first recording session in which he plays was on 31 December 1935. He stayed until he was sacked in mid-1938 (effectively for being too popular), making sixty-four recordings: forty tangos, eighteen valses and six milongas. These are D’Arienzo’s greatest recordings: they have an irresistible drive and are loved by dancers.

D’Arienzo (Instrumental) 1937

D’Arienzo (Instrumental) 1937-38 (Vals)

It’s a shame that all the versions of Valsecito de antes available on listening service Spotify have the same added reverberation, robbing the music of much of its energy. Typically, reverberation was added to LP compilations to disguise surface noise and nearly all CD releases by the big record labels are derived from LPs. Here’s an excerpt from a recent transfer made directly from a shellac disc (by TangoTunes) for comparison. I have processed the file to further reduce running noise:

D’Arienzo (Instrumental) 1937-38 (Milonga)

Pianist: Juan Polito (1938-39)

Juan Polito succeeded Biagi at the piano and while he couldn’t quite replicate Biagi’s flair, the orchestra maintained its high musical standards and commercial success. Singers were growing in importance too, and just before Biagi’s departure, D’Arienzo had recruited his most notable singer: Alberto Echagüe. At the end of 1939 Polito left to form his own orchestra, but he also persuaded Echagüe and nearly every one of D’Arienzo’s musicians to join him. D’Arienzo, who had to reform his orchestra with new players, was absent from the recording studio for several months, but he returned as popular as ever.

There are forty recordings: twenty-nine tangos (twelve with Echagüe), five valses and six milongas (all with Echagüe).

D’Arienzo (Echagüe) 1938-39

D’Arienzo (Echagüe) 1938-39 (Milonga)

The Forties

… tango is, above all, rhythm, nerve, strength and character. Early tango, that of the old stream (guardia vieja), had all that, and we must try never to lose it. (D’Arienzo, 1949)

Pianist: Fulvio Salamanca (1940-57)

By April 1940, D’Arienzo was back in the studio, with 18-year-old pianist, Fulvio Salamanca and with Carlos Casares and Alberto Reynal on vocals. The first recording, Entre dos fuegos, presented Salamanca’s credentials, leaving us in no doubt that the role of pianist remained central to the orchestra’s sound:

The instrumental recordings of the early 1940s are straightforward vintage D’Arienzo. Compared with the repertoire of the later 1930s they have perhaps traded some finesse for vigour. But by 1942 and Tierra negra the orchestra has adopted a more sophisticated sound – a response to changing times?

D’Arienzo (Instrumental) 1940-42

The early 1940s were a time of considerable change in tango music. The high energy rhythmic music-making that D’Arienzo had done so much to promote was slowing down, mainly to accommodate the rising popularity of the orchestral singer. The change was not lost on D’Arienzo, who replaced Casares with the voice of  Héctor Mauré at the end of 1940. This called for a change of gear for D’Arienzo but also presented a paradox: his orchestra was built on a foundation of driving rhythm, while Mauré took the orchestra further and further into the realms of lyricism.

Mauré’s recording debut came in April 1940 with Ya lo ves, which has the usual D’Arienzo rhythmic stamp, and yet you can hear Mauré beginning to stretch the compás to suit the delivery of the lyric, but not yet sufficient to force the orchestra to depart from the pulse to follow him:

By November 1943, Mauré has taken the orchestra to a new world of sound with a masterpiece, Uno. I can’t improve on Michael Lavocah’s description of this version as being …

… perhaps the greatest ever recorded. The music swells and tumbles in great tides in a most un-D’Arienzo like manner.

D’Arienzo (Mauré) 1942-44

The tension of the clash between Mauré’s lyricism and D’Arienzo’s hard-driven rhythms led to Mauré’s inevitable departure in 1944, prompted by the return of Echagüe. Amarras was his last recording, made in July 1944.

It is a personal view, but I don’t much enjoy the repertoire from the remainder of the 1940s. Perhaps D’Arienzo over-compensated for his brush with lyricism, for the music-making becomes increasingly strident and starts to sound, well, just loud. It would be silly to write off the recordings completely, and yet, I don’t play them. There are always limited opportunities to play the music of any single orchestra, even in a milonga of four or five hours duration, and I find that this period always gets squeezed out in favour of better D’Arienzo.

The Fifties

Juan D'Arienzo
Juan D’Arienzo

As though someone had thrown a switch, there is a change in the feel of the orchestra’s sound from 1950. You could say that the orchestra had rather lost its way for most of the 1940s, but had now recovered its energy and sense of purpose. However, before the year was out, principal bandoneonist and arranger Héctor Varela had left to form his own orchestra. D’Arienzo took a short recording break to regroup and appointed pianist Salamanca to the role of principal arranger.

D’Arienzo (Instrumental) 1950-51

Tape mastering was introduced by Victor in 1954 and D’Arienzo became particularly active in the studio, re-recording his then-current repertoire for the sake of the improved sound.

D’Arienzo (Instrumental) 1954-55

Armando Laborde had sung with the orchestra since 1946 and overall I think his vocal contribution to the D’Arienzo orchestra was second only to Echagüe’s in the late 1930s (pushing Héctor Mauré into a close third place).

D’Arienzo (Laborde) 1954-56 (Vals)

Piano: Juan Polito (1958-75)

In 1957, Salamanca left (after seventeen years at the keyboard) to form his own orchestra. Interestingly, he seemed to take nothing of the D’Arienzo style with him, but Juan Polito (who had been his predecessor) returned, and thus the orchestra entered its final period. Singers Echagüe & Laborde left too, and from that point, I restrict my own interest in the orchestra’s recordings entirely to the instrumentals.

These milongas, while not particularly well-known, are a good example of the style of the orchestra in the late 1950s after Polito had returned at the keyboard. There is a better balance between the instruments: Polito doesn’t seem to hit the piano keys quite as hard as Salamanca!

D’Arienzo (Instrumental) 1957-59 (Milonga)


By the early 1960s, many tango orchestras had disbanded as social dancing was much diminished – the Golden Age was well and truly over – and yet several orchestras remained popular and continued recording. D’Arienzo was active until the year before his death in 1976. From 1963 Victor, seeking a new international market for its tango recordings, introduced stereo sound and made a series of albums under the title Tango for Export most notably with D’Arienzo and Troilo.

The music is very danceable and the sound is excellent.

D’Arienzo (Instrumental) 1963-67


The best period is 1935-39. Nearly everything is good. The earliest sides with Fasoli (before the orchestra really got into its stride) are good, the later sides with Polito are better and the middle period with Biagi is the best (by a long way).

The Salamanca years (1940-57) were dominated by a hard-driven sound that I can only take in small doses. The vocal contribution of Mauré (1940-44) is fine, but not typical D’Arienzo (but none the worse for that). The later vocals are largely undistinguished, but there are a lot of recordings with Laborde and some of them are good. It is necessary to be very selective.

The final period with Polito (1958-75) is only really of interest because of the good stereo sound from 1963. Purists will be horrified to think that a respectable DJ would play these recordings – and they can please themselves – I quite like them.

Postscript: The Twenties

Juan D’Arienzo

D’Arienzo was already in his mid-thirties when he secured the recording contract with Victor that made his reputation. However, he had already been actively involved in the tango scene for over a decade, making around forty tango recordings between 1928-29 on the Electra label. He also played on the radio (then, as important as recording) and also performed regularly at the prestigious Chantecler Cabaret.

I have never heard one of his early recordings played for social dancing and the sound quality of the transfers I have been able to find is pretty dreadful. These recordings are usually written off as being of no real interest but I don’t think that’s fair. I’ve listened to most of them, and if the limitations of the sound quality are put to one side, the arrangements and performing style are very typical of their date. Here’s an excerpt from Esta noche me emborracho, recorded in 1928, with Carlos Dante (who later sang with Alfredo De Angelis):

For comparison, here are excerpts from two other recordings of the same song, also made in 1928. The first is by Juan Maglio with singer Carlos Viván and the second by Franciso Canaro with singer Charlo. The arrangements are all slightly different, of course, but the D’Arienzo recording holds its own.

The other vocalist to feature on the early D’Arienzo recordings was Francisco Fiorentino (who was later to make his name with the orchestra of Aníbal Troilo). Here’s an excerpt from Noche silenciosa, recorded in 1929. It has quite an adventurous arrangement for its date:

If an enterprising record company were able to make new transfers from the original 78s of this period it’s likely that at least some of these early recordings would be heard more regularly.

More …

In this blog the music of D’Arienzo featured regularly in The Tango Years series and also in the Tango 250 series:

DJ Clive Harrison
DJ Clive Harrison

Barrio de tango is the tango blog and online home of tango DJ, Clive Harrison, based in the English Midlands. Now retired from teaching and hosting dance events, Clive remains available to DJ, playing exclusively traditional tango music from the great tango orchestras.

Tango in 1958-59

There is no one definition that satisfactorily defines tango’s época d’oro – the Golden Age. The formation of D’Arienzo’s orchestra in 1935 is frequently cited as marking its beginning, with the political crisis following the coup of 1955 marking its end. In the same breath, the Golden Age is often described as lasting 30 years, while others name the single decade of the 40s. However you reckon it, the Golden Age was over by the end of the 50s.

Even at the end of the 50s, there was still great music to come – some of it, good for dancing. Tape mastering had transformed the quality of recorded sound by the mid-50s and by the early 60s, several orchestras were recording in stereo. Incredibly, Fresedo (who had formed his first orchestra in 1918) carried on recording until 1980.

Di Sarli (Instrumental) 1958

Carlos Di Sarli
Carlos Di Sarli

Di Sarli ceased recording for Victor after August 1958 but he returned to the studio one last time (for Polygram) in late 1958 to record a final LP. The sound quality doesn’t match Victor’s, but the LP included two instrumental masterpieces not previously recorded: Una fija and Indio manso. Although very ill, Di Sarli continued to make some live appearances, but the last of those were during the carnival season of 1959. He died, just days after his 57th birthday on 12 January 1960.

  • Champagne tango (13-Nov-1958)
  • Una fija (12-Nov-1958)
  • El abrojo (13-Nov-1958)
  • Indio manso (13-Nov-1958)

Pugliese (Instrumental) 1955-59

Osvaldo Pugliese
Osvaldo Pugliese

These instrumentals take the emphatic pulse of La yumba and seem to distil its essence. For me, there was only one remaining masterpiece still to come, A Evaristo Carriego in 1969. Pugliese’s influence on tango dance music had been profound, but there was, perhaps, little left to say.

  • Emancipación (2-Sep-1955)
  • Nochero soy (28-Nov-1956)
  • La bordona (6-Aug-1958)
  • Gallo ciego (23-Jul-1959)

Roggero (Mixed) 1958-59

Aquilies Roggero
Aquilies Roggero

Aquiles Roggero had played violin in the orchestra of Osmar Maderna, who had been tragically killed in a flying accident in 1951. Roggero formed the orchestra in 1952, naming it Orquesta Símbolo Osmar Maderna – effectively a tribute band. The music is both romantic and dramatic and very much of its time. Tenor, Adolfo Rivas features in each of these songs (except instrumental, La huella) and he is joined in the famous (or is that infamous) Merceditas by baritone, Carlos Aldao.

  • Lucecitas de mi pueblo (30-Jun-1959)
  • La huella (29-Sep-1959)
  • Mejor así (29-Sep-1959)
  • Merceditas (21-Sep-1958)

Salamanca (Guerrico) 1957-59

Fulvio Salamanca
Fulvio Salamanca

Fulvio Salamanca had been D’Arienzo’s pianist for seventeen years when he left to form his own orchestra in 1957. No one could have guessed what would emerge – for the music is utterly unrelated to D’Arienzo’s style. He recruited legendary tango violinist, Elvino Vardaro, and never looked back …

This is music about which it is difficult to feel ambivalent: you’ll either love it or hate it.

  • Bomboncito (6-Jun-1958)
  • Todo es amor (18-Sep-1958)
  • Hasta siempre, amor (17-Jul-1959)
  • Adiós, corazón (29-Nov-1957)

Click here for ‘Tango in 1956-57’.

Click here for an index to all the articles and music included in this series.

DJ Clive Harrison
DJ Clive Harrison

Barrio de tango is the tango blog and online home of tango DJ, Clive Harrison, based in the English Midlands. Now retired from teaching and hosting dance events, Clive remains available to DJ, playing exclusively traditional tango music from the great tango orchestras.

Tango in 1956-57

The political climate that followed the coup of 1955 had a profound effect on tango, beginning a steady decline that was only reversed after the return to democracy in 1983. The top orchestras remained active, at least in the recording studio, but many others folded, and what remained was increasingly for listening rather than dancing. The Golden Age was over.

D’Arienzo (Laborde) 1954-56 (Vals)

Juan D'Arienzo
Juan D’Arienzo

Armando Laborde recorded with D’Arienzo from the end of 1944 until 1957 (and for an extended second spell from the mid-60s). Much of this repertoire is undistinguished, but these valses are well-worth hearing and typical of the mid-50s D’Arienzo style.

    • La sonrisa de mamá (1-Sep-1954)
    • Lloré por los dos (20-Sep-1956)
    • Me quieres y te quiero (17-May-1956)

Di Sarli (Instrumental) 1956-57

Carlos Di Sarli
Carlos Di Sarli

These instrumentals were among Di Sarli’s last recordings. The pace is slow, the style expansive, with the orchestra producing a sound dominated by violins and the piano. Many dancers love these late recordings for their emotional intensity and the way in which Di Sarli manages to combine lyrical and rhythmic elements in the music – the culmination of three decades of steady musical development.

  • Bahía Blanca (21-Nov-1957)
  • Nueve puntos (7-Mar-1956)
  • Cara sucia (12-Jul-1957)
  • Viviani (19-Dec-1956)

Di Sarli (Florio) 1956-58

Tenor, Roberto Florio, was the last of the great singers to work regularly with Di Sarli, singing alongside Jorge Durán, who had previously worked with the orchestra in the mid-40s. The ‘late’ recordings (from 1956 onwards) were mostly vocals, but in truth, I find the instrumentals (although few in number) more satisfying. Adiós, corazón was the last of Florio’s recordings with Di Sarli.

  • Derrotado (27-Sep-1956)
  • Soñemos (25-Apr-1957)
  • Nuestra noche (24-May-1957)
  • Adiós, corazón (16-Jan-1958)

Pugliese (Maciel) 1954-56

Osvaldo Pugliese
Osvaldo Pugliese

Jorge Maciel joined Pugliese’s orchestra in 1954, replacing Alberto Morán. He had been poached from the orchestra of Alfredo Gobbi, with whom he had already recorded Remembranza and Canzoneta. It was a successful partnership, but I’m not sure that the music is very good for social dancing (although this repertoire is regularly chosen for performances). If you like it, you probably love it, but I can’t hear the last bars of Canzoneta and not snigger a bit.

  • Remembranza (4-Jul-1956)
  • No juegues a la guerra (31-Jan-1956)
  • Canzoneta (29-Oct-1954)
  • Cascabelito (22-Sep-1955)

Quinteto Pirincho (Instrumental) 1957-59

Francisco Canaro

Francisco Canaro was never one to miss an opportunity to make money. At the height of the popularity of D’Arienzo’s orchestra in the late 30s, Canaro formed a quintet (initially called Quinteto Don Pancho) to play music in the guardia vieja style that D’Arienzo had reinvigorated. It recorded, on and off, throughout the rest of the 30s and 40s (having changed name to Quinteto Pirincho early on) but it recorded more regularly throughout the 50s. The recordings feature a fairly conservative repertoire of tango standards in straightforward arrangements and with good sound. I don’t enjoy Canaro’s orchestral recordings after the end of the 30s much, but the Quintet recordings have a different feel, even if a little goes a long way.

  • Mala junta  (2-Sep-1957)
  • Zorro gris (31-Jan-1957)
  • El pollito (22-Apr-1959)
  • Derecho viejo (24-Apr-1956)

Click here for ‘Tango in 1954-55’.

Click here for ‘Tango in 1958-59’.

DJ Clive Harrison
DJ Clive Harrison

Barrio de tango is the tango blog and online home of tango DJ, Clive Harrison, based in the English Midlands. Now retired from teaching and hosting dance events, Clive remains available to DJ, playing exclusively traditional tango music from the great tango orchestras.