Organ Recitals: Audience Preferences
by Colin Pykett
Posted: 8 August 2012
Last revised: 31 August 2012
Copyright © C E Pykett 2012
Abstract. This short article takes a look at the download statistics relating to the music files on this website in an attempt to gauge today's audience preferences for organ music. The composers represented span five centuries and their works are played on eleven simulated styles of organ, from Schnitger to WurliTzer. Total playing time exceeds three hours, equivalent to the contents of about three CD's.
To assess whether the downloads of these files reflect useful information about audience preferences for organ music, data for the six months from July to December 2011 inclusive were analysed. A 'Top 20' chart is presented for this period, reflecting 7530 downloads of the 20 most popular numbers. The absence of Bach in the Top 20 was surprising, given that he is by far the best-represented composer on the site. On the contrary, the taste of this cyber-audience for more recent and 'tuneful' works was marked, as was their apparent preference for organs with a Romantic rather than a Baroque flavour.
The data might be useful to those compiling recital or concert programmes, as it is difficult to see how such a comprehensive survey could be conducted in any other way. It is worth pointing out that the results are based on what people actually listened to rather than what they might say they listen to, and in this sense the outcome probably reflects reality more closely than trying to assess audience preferences using alternative means.
(click on the headings below to access the desired section)
does one compile a recital programme, or merely decide what to play before and
after a service? All organists will
have their own views, but one ingredient in the mix must surely be what their
audiences want to hear, rather than what we think they ought to hear.
It amused me to read of a player who was surprised by the fact that only half a
dozen people turned up to hear her playing a long programme entirely of works by
Soler! At the other end of the spectrum we have those players, sometimes
of 'celebrity' status, whose recitals feature the same dreary core items
wherever they go. Whether we like it or not, people do have preferences
which might deserve more respect than some performers seem to give them.
So, taking this
as a given, how can we find out what they like or dislike - what these
This is the subject of this article.
are at least three major issues. The
first is obvious - that different people like different things.
The second is that we can sometimes get right into the hearts of at least
some in the audience if we play the right stuff.
The third follows from this - although
it is nice that we can sometimes please the odd one or two people, most of us
probably want to please the majority rather than a minority.
All three are important, but we probably cannot do much about the first.
As the saying goes, you can't please all of the people all of the time.
But the second and third issues are within our control.
They remind us that people can be moved by our choice of music, and
presumably we would like to maximise the impact of the music we choose.
So how do we achieve this?
recently I have never found a satisfactory answer to this question, one which is
asked in circles far wider than that populated by us organists.
Recording companies, for instance, devote considerable resources to it so
they can optimise their sales of 'classical' music, and in the popular music
field they go a step further to create new markets for new releases by inventing
new boy/girl bands for example
However it is not surprising that the commercial music business does not disseminate
its market intelligence widely. But coming back to organs, it is only from having run this website for
many years that I have become more aware of what the global cyber-audience
for organ music seems to like and dislike on the basis of the download statistics relating to the music
files on the site. So in this article I have used them in an attempt to gauge today's audience preferences for
The results are interesting, some are surprising, and I thought you might
like to share them.
Boring as it might be, I first need to say something about how much music is on this website, the composers represented and which organs it is played on, though if you prefer you can skip directly to the results where the 'Top 20' favourite hits are presented.
At the time of writing (2012) there are 68
music tracks (mp3 files) available with
a total playing time exceeding three hours.
This equates to about three CD's worth of music in round figures.
The number of files downloaded by visitors to the site continues to
increase steadily month by month, and well over 2000 tracks were downloaded in
December 2011 for example. In terms of listening time this
corresponds to at least ten typical organ
recitals per month at this download level, so the volume of material
involved is probably representative enough of listeners' tastes to warrant
analysing it in more detail.
are these listeners? One cannot
know of course, though the vast majority of files are downloaded by humans as
an act of deliberate choice. Unlike
ordinary html web pages which contain text, music files are not widely sought by
the web robots,
spiders, crawlers and the like which search engines such as Google use to index
the contents of web sites (though some of my music files have been pirated by sites which
then offer them as downloads for iPods, ringtones for phones, etc.
Neglecting the legal issues, I am not sure whether to regard this as a compliment). The statistics programs provided by my web hosting company
confirm this picture by showing details such as the country of residence of each
downloader and even the mp3 player s/he used (Windows Media Player, NSPlayer,
etc). This information would not be
available were the downloads initiated automatically by robots, thus it confirms that
analysing the statistics to assess the preferences of a human audience is valid.
a third of the 68 numbers are played on a three manual electronic organ with an
English romantic specification. The
remaining majority is rendered on a virtual pipe organ (Prog
simulates ten different instruments, ranging from one by Arp Schnitger to
various twentieth century organs.
are 34 composers represented over a 500-year span from the fifteenth century to
the twentieth, and the way their music (the 68 files mentioned above) is
distributed across this timespan appears in Figure 1.
1. Distribution of mp3 music files
versus year on this website
one big peak shows that 20 compositions are featured from the first half of the
18th century - these are all by J S Bach. Otherwise
the distribution is relatively flat apart from a smaller peak featuring romantic
works from around 1850.
the foregoing has shown that there is a reasonably wide range of organ music
available on the site, and that it attracts enough global interest to make
further analysis worthwhile. Although
I could bore you to death by continuing to massage the raw data for ever, it is
probably best to keep this article short by drawing it to a close with the
'top twenty' titles. These vary
somewhat from month to month, but by grossing up the results for the six-month period from July to
December 2011 most of these variations will be smoothed out.
This 'top of the organ pops' chart, representing 7530 separate downloads
of these twenty titles over this period, is in Table 1 below.
Table 1. The 'Top 20 Organ Pops' downloaded from this website between July and December 2011 inclusive
most surprising feature to my mind was the absence of works in the Top 20 by Bach, even though
Figure 1 shows that he is by far the best-represented composer on the site.
(Note this does not mean he was not downloaded during this period, only that he
did not feature in the Top 20). On the contrary, the overwhelming preference of those who downloaded this
large number of files was clearly for romantic works.
These downloaders also seemed to favour romantically-voiced instruments.
Another aspect is that I believe some of the works are long out of print. This might suggest a willingness on the part of the audience to try numbers they had not come across before, provided they are 'tuneful' in the romantic sense. Indeed, this is confirmed by the number of requests I receive for the sheet music - which of course I cannot supply. But this, in turn, shows that at least some of the audience are probably organists themselves. There is also some slight evidence that lesser-known composers, such as Rheinberger and Karg-Elert, are attractive.
there you have it. When you next
give an organ recital, make sure you choose numbers with memorable tunes which your audience can
hum or sing on the way home if you want to gain a reputation as a 'popular'
performer! Your church will
probably welcome this as well, because larger audiences mean heavier collection
Sorry, JSB. You just do not seem to be pressing the right buttons at the moment. Nor, perhaps, are those organ builders who are making 'baroque' style instruments rather than more eclectic ones. Obviously these remarks are made with a dash of tongue in cheek, but against the background of such a large body of evidence it seems difficult to ignore the outcomes completely.
finally, I should like to thank that army of anonymous downloaders out there in
cyberspace who have accorded me the privilege of visiting my website, and who
have thereby enabled this article to be written.