Occasionally, when I listen to an orchestra playing a Strauss waltz or something similar, I hear something interesting in the background rhythm of 'boom chik chik' in that the first 'chik' happens somewhere before beat 2.

Take, for example, this recording of The Blue Danube Waltz and compare to any score on IMSLP. (The excerpt in question is marked Waltz I, and occurs right after the first fermata.) The notated rhythm is different from what the orchestra plays!

What are the conventions for performing waltzes in this style? Is it is specific to a composer, time period, or school of interpretation for all waltzes?

Furthermore, where exactly should the second note be played? Is the note shifted forward by a full eighth note, a sixteenth note, or maybe a portion of a triplet?

  • Dance a Viennese waltz, then you will understand. Or play for dancers and watch them move.
    – user24860
    Nov 25, 2015 at 21:24

7 Answers 7


Even if I would not take the late Karajan as a reference for this, you are right, the first beat of each bar of a Wiener Waltzer is almost always played short; usually the leading voice is even slightly in advance of the bass when the melody is written in 3 quarter notes. This can be approximated as removing 1/6th of the first beat and adding it in advance of the second one.

It is also a very important clue to use when dancing the waltz (we are talking of an original quick Austrian waltz, not what is often called waltz in the US but is usually danced two times slower). When turning (by opposition to Pendel Schrift) during the time between the first two beats the man builds momentum and release it a little more slowly thereafter. Most people trying to dance the waltz with strictly even beats fail to be gracious or enjoy the dance; that is why a teacher I know advises to begin as early as possible to count bars and not inner steps.

This is specific to Viennese waltzes composed for dancing in the European tradition (Strauss family, Lanner, van Suppe, Waldteufel, ...), in opposition to the American slow waltz, Boston, etc. It is difficult to say what to do about the waltz form as a piano form, especially Chopin. They are very specific and often played at tempo incompatible with dancing (either too slow or too quick), and Chopin reportedly said he was not able to write waltzes like Mr Strauss -- this could have been a very strong critic.

In an orchestra, you would depend on the conductor's cue to choose the placement of the second beat. As a solo player this is a difficult choice to make. It requires more dexterity to play the second beat closer to the first but it could add some welcome lightness to the music.

  • Interesting... I have a book of a German drummer, Werner Schmitt (called Tanzmusik für Drummer, "Dance music for drummers"). He played a lot of dance music in big bands for professional dance championships. He claims that the drummer should - in contrast to the way classical orchestras play this music - enforce an exactly equal distribution of the beats in the Viennese Waltz to simplify dancing. I'm not a good dancer, so I don't know if he's right. May 31, 2011 at 18:22
  • @groovingandi : interesting as well. It might come down to a common convention between dancers and drummers in that kind of competitions. Dancing is always about negotiating what can be and using certain aspects as if they were rock and steel. Changing apparently little in your technique and timing would change a lot in the experience and the relation to music. I am not a competition dancer, I have been trained for social dancing and this limits what you can expect and what you are told, in that case: to expect Viennese beat and pulse accordingly.
    – ogerard
    May 31, 2011 at 21:18

Since I have been studying and living in Austria as a South African music student I have had several opportunities to play waltzes in orchestras. Especially New Years concerts consist of a couple of famous Viennese Waltzes and since the first time I had to play one of these, I have also been fascinated by this interesting rhythmical appearance. Being both violinist and violist, I get the chance to sometimes play the flashy violin melodies and other times I play the less interesting "boom-cha-cha" voice as part of the lower strings (yet now a very interesting topic) and because of this I think that I have a quite valuable opinion.

I find all of the above opinions and observations very true, but it is important to take note that the early placement of the second beat in the accompanying rhythm (I will refer to this onwards as the EPSB) does not occur straight through the piece in every bar. If one would listen to the above mentioned recording of the Blue Danube conducted by Herbert von Karajan, one would hear that in some sections of the piece the second beat comes early and in other sections the beats are very equally spaced. I find that the EPSB is used in sections which has a very clear swinging one-bar feeling with a very clear emphasis on the every first beat, where the melody consists of longer lines. This way the music still has a very clear swinging dance feeling in spite of the long and somewhat heavy melody. Then, in other sections, where the melodic character is lighter and has a clear feeling of three regular beats, the accompanying rhythm is less apparent and the three beats are equally spaced.

In both variations it is interesting to note the relation between the melodical and rhythmical factors of the composition and how the character of the melody is balanced out by the accompanying rhythm to ensure the constant flow of the dance. In the sections where the melody consists of longer, carried tones (and when played on its own doesn't have any clear rhythmical indication), more importance is given to the accompanying rhythm by the strong emphasis of the first beat and the attractive early placement of the EPSB which encourages a swinging dance feeling. The third beat is also played with some importance in order to lead more to the important first beat in the next bar.

Otherwise in the sections where the melody itself is lighter and has more rhythmical characteristics, the accompanying rhythm is of less apparent and the beats are spaced equally and played with more equal importance with a slight emphasis on the first.

To comment on the opinion of the German Drummer mentioned above: Yes, it would probably make life easier for dancers if the three beats were spaced equally, but that would lead to a very boring and "never ending" performance of a very unique tradition of good Austrian Waltzes. Therefore why not rather improve your dancing skills to fit to the music. Also think of the poor violist which only gets to play the "cha-cha" and not even the "boom" in these waltzes - to vary the rhythm slightly now and then could only give some spice to his tedious job. Besides, the drummer actually doesn't have much say in the placement of the second or third beat, because 90% of the time he only gets to play the first beat of each bar.

  • An excellent first answer! Would you be able to comment on whether EPSBs are planned out in advance, and whether they are communicated to the ensemble by the concertmaster or conductor in performance?
    – NReilingh
    Jan 4, 2013 at 5:03
  • Is it possible that those irregular beats are for the dancers' sudden left-right foot switching something like boom-chick-chick...boom-da-chick-chick..boom-chick-chick. I can imagine if you change the rotation you might need change the leading foot.
    – user1306
    Jan 4, 2013 at 14:51

Speaking as a recovering percussionist and as a conductor, it is important to note that the methods used to talk descriptively about beat placement have two contradictory qualities:

  1. They come from a very modern and commercial worldview. I use the term commercial to refer specifically to the advent of electronic recording techniques and more specifically to quantization.
  2. They are (almost) the best we have.

It may seem strange, but modern electronic metronomes with the ability to create a variety of quantized subdivisions of the beat and commercial pop recordings have slowly eroded our historically more fluid notion of musical time. (Think of other musical styles in which notes of equally notated value receive non-uniform durational values in performance, such as notes inegales, ragtime, or swing. The truth is that it is problematic to say simply that the second beat often falls "early" in many Viennese waltzes. It is accurate, only if ones accepts a modern notion of time that recognizes a universal sense of "on time" anchored on sequential quantized pulses. In truth, the word "early" is an implicitly relativistic term and is problematic in the case of making an analysis of a musical culture.

As an example: what does it mean to say that I woke up early this Sunday morning? It implies a cultural norm regarding when people wake up on a Sunday, and that I woke up earlier than that. So "early" actually means "earlier than the norm". It only gives you a sense of a time range in which I probably woke up because you have a sense of when people normally wake up on Sundays. If that cultural norm shifted, as many such norms often do, someone could read this very differently in the future.

Now, unfortunately for us, commercial culture has become so widespread, that it becomes increasingly difficult within the developed world to grow up with any other possible understanding about time. This makes it even harder to know for certain how even the relatively homogenous, almost all-male Vienna Philharmonic might conceive of these ideas. But it is worth noting that time is actually significantly more fluid, and can be thought of in a variety of more fluid terms.

Just one very simple example within the present discussion: you could just as easily and (arguably) just as rightly say that the first beat is clipped short. We imply that the second beat to be the pattern-offending beat when we say it is early. We know that musical note values are only articulate-able in terms of their relationships to other note values. We know that we experience note values in sequence, i.e. one after the other, and so we cannot appreciate a note value in musical terms until (at least) the following note value is performed.(In other words, when we listen, we don't know how long a note is until the next note or rest is performed. This is opposed to measurement in scientific terms of duration in milliseconds, which is possible but not germane.) Also, we cannot deduce patterns until a sequence has repeated itself at least once (for a total of two full iterations). So we can't make hierarchical relationships (such as first beat, second beat, third beat) valid without having heard that pattern iterated at least twice. (And if my 8th grade algebra teacher is correct, then we really need at least three points [iterations] on that line to confirm it.)

All of this said, we return to my example of an alternative description of the three interior pulses of many Viennese waltz performances. As I said, beat one is clipped, and we add to that description that: beat two is pregnant, and beat three is neutral. These are still relational terms mind you, but they don't single out beat two as being an offending beat. The only thing it offends is a quantized notion of where the beat "should" land. It is also important because these terms relate the three pulses to each other and not to an exterior musical culture. Think of the three bears. One's too short, one's too long, and one is just right.

This might seem like an interesting (and admittedly incomplete) discussion about broader concepts suitable for a theory classroom rather than a specific answer to the original question. But it can lead to a change in the framework over which broader interpretive decisions are made. For instance, I have heardmany performances by orchestras and conductors less skilled, perhaps less mindful, and certainly less nuanced than those mentioned above. I often find that they exaggerate by over-emphasizing the second beat, causing it to stick out. It can almost begin to sound like a form of duple-based syncopation: eighth-quarter-eighth. This, I believe, is not the desired effect. It is always hard to say for certain what conception an interpreter (be it a conductor or otherwise) was using as the basis for a performance without having heard them describe it in rehearsals, but it is safe to say that small details in descriptive language can often have far-reaching implications in performance.

The best remedy for these problems of language has been, in my experience, something that is (thankfully) more immediately gratifying than worrying about the semantics of the word "early" as I have just done and (fortunately) much easier to do: LISTEN! Listen to as much music as you possibly can. A great way to make a useful comparative study would be to select 5 of the past Vienna New Year's Eve concert recordings. Same orchestra, same hall, mostly same players, different conductor, and choose just one waltz common to all. The easiest of course is the most well-known, Blue Danube, since it is performed virtually every year. Choose only one or two strains to listen to at a time, and then go through each recording, noticing as many fine distinctions as possible. One useful tool to also consider is to find cases where the same conductor has returned and performed at least twice, such as the legendary Carlos Kleiber in '89 and '92. Still, we find small differences that can be enlightening. Absorbing the style as best one can through listening is by far the best way to understand it. It is really the difference between learning German in a classroom and understanding the mechanics of verb forms and cases, and the level at which you can go to Germany and listen to the language to understand how it is actually spoken by ordinary people.

  • 1
    +1, interesting, although it seems to somewhat miss the question (it was asked what the conventions are whereas you seem to argue why they are justified...) Anyway, there's one point I'd like to make. You seem to say that the sense of being on time is somehow a modern notion. However if I rember correctly, for example in baroque music it was just as important that every measure is on time as that the individual beats are not on time. What I think has changed is that now we interpret the notes literally. Before a dot meant the note is longer, now it means it's exactly 3/2 longer.
    – nonpop
    Jan 6, 2013 at 17:57
  • 1
    You make a good point. I think my hope, though, is to show that it is dangerous to try to get answers to this kind of stylistic question from sources other than listening to more performances. I showed a case where uncareful verbal answers that fail to convincingly capture the essential practice can easily lead to bad performances that multiply the problem. On the other hand, finding the answer in the originating medium (musical performance, or more generally, sound) is more effective and (conveniently) more enjoyable. Jan 6, 2013 at 19:09
  • "I woke up early this Sunday morning". 'Early' might be absolute. No-one disputes that 5.00 a.m is 'early. Or it might be relative - "I normally rise at 8.00 but this Sunday I woke at 7.00. Player of notes inegales KNEW they were 'inegale'. They had a perfectly good concept of where the beat 'should' be.
    – Laurence
    Oct 11, 2022 at 10:05

Attempting to quantise a viennese waltz has much in common with trying to quantise swing. When everyone in the ensemble feels it the same way, magic happens.

There's a degree of truth in both of these notations. But it's horrific to imagine them being taken as mathematical instruction.

enter image description here

  • 1
    What, Laurence, mathematics is horrific? But you're right: math has its proper domain. And the Viennese waltz is not in it. Cheers from stormy Vienna, Scott Oct 23, 2018 at 15:07

This is specific to a city, Vienna. I have heard it called "Wiener Blut", i.e. "Viennese Blood". "Put more Wiener Blut into it," a teacher once said to me.


In the Viennese Waltz, the second beat is slightly anticipated. This is a particular characteristic of the Viennese Waltz. When conducting such waltzes, conductors generally only indicate each bar rather than each beat.

The timing is roughly as if the beats are triplets with the second beat syncopated.


I've just been listening to traditional Blaskapelle music from 2018 Oktoberfest in Munich, and I noticed that when they played waltzes they also used the Viennese style of an anticipated 2nd beat:

so this seems to be more universal (but of course Munich is quite close to Austria)!

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.