Digitising an old lady
Home Up Other Publications Contact Me About the Author About this Website Prog Organ - a VPO Sitemap

 

  Digitising an old lady - simulating the 1858 Walker organ at St Mary Ponsbourne, Hertfordshire  

 

Colin Pykett

 

"A good simulation ... gives us a sense of mastery over experience" 
"The Dreams of Reason", Heinz Pagels

 

Posted: 27 March 2013

Last revised: 13 June 2013

Copyright © C E Pykett 2013

 

 

Abstract.  The church of St Mary Ponsbourne in Newgate Street village near St Albans in Hertfordshire contains an authentic and historic English organ built in the mid-Victorian era by the well known firm of J W Walker in 1858.  This article describes how its sound was captured digitally  for posterity on a virtual pipe organ (Prog Organ), just prior to a major overhaul in 2013.  Some audio examples of the result are included.

 

 

Contents

(click on the headings below to access the desired section)

 

Introduction

 

The Organ at St Mary Ponsbourne

 

Digital simulation of the organ

 

What does it sound like?

 

Acknowledgement

 

References

 

Introduction

 

The church of St Mary Ponsbourne is one of the Hartford Hundred West Group of five churches in the Diocese of St Albans, all of which nestle in rural Hertfordshire, England.  The area lies just north of the M25 motorway and about equidistant between London and Stevenage, and one of its more memorable aspects is that it is about the first sustained glimpse of the countryside one gets as one travels north out of the capital on the A1.  The church contains an interesting and historic English organ built in 1858 by J W Walker which, thanks to a generous grant from the Heritage Lottery Fund, is about to be overhauled (early 2013).  This article is about a digital simulation of the instrument which was developed just before the work commenced.  To me, this is a point of particular interest as I regard the aural record which resulted as having some intrinsic value for posterity, because all major interventions in the affairs of any organ will always result in changes to the way it sounds.  This is in no way a criticism of the way organ builders carry out their business but simply an acceptance of reality.  In fact a digital simulation, when carefully done, is far more than the passive sort of aural record one would get merely by listening to a CD.  It is a living, musical record because it enables one to recapture the greater aesthetic of an organ by actually playing it.  Consequently I was grateful to the organist at St Mary's, Paul Minchinton, for having provided me with a complete set of raw recordings of every pipe in the instrument as it sounded in January 2013. 

 

Those who wish to know more about the church before reading further might like to visit its website

 

 

The Organ at St Mary Ponsbourne

 

A relatively untouched organ still in regular use and dating from 1858 (155 years ago at the time of writing) is something of a rarity.  So I found myself reflecting on what life in Britain was like when this venerable old lady was built all those years ago, as it puts the instrument into a context so utterly different from that we are familiar with today.  Thus we find the steam railway augmenting transport which, until about twenty years earlier, did not exist beyond one's own legs, waterways or the horse.  Mains electricity was unknown, though gas was lighting towns.  Elsewhere people still relied on oil or candles, as did organ builders much of the time (which must have made working in a country church difficult during the short dark days of winter).  Communications were benefiting from the landline telegraphs which were being installed in post offices, though the Atlantic was not to be bridged successfully for some years and the telephone was yet further away.  Painless surgery had been available to those with the means for a mere twelve years, though post-operative mortality would remain a near-certainty until Lister's introduction of antiseptics a decade later.  The British penchant for war had been temporarily sated by that in the Crimea which had ended two years previously.

 

Musically, Beethoven would have been known to many then still living.  Mendelssohn had died only eleven years before the organ was built, and Chopin but nine.  Saint-SaŽns had taken up his post at La Madeleine the year before.  Robert Hope-Jones would be born the following year.  And against this backdrop, the firm of Joseph William Walker was doing very nicely in 1858 thank you, with a burgeoning order book and a substantial workforce.  As well as making worthy small instruments like that at St Mary's, he also built the large and famous one at Romsey Abbey that same year.  His work set that standard of mechanical and tonal excellence, including the sound of those beautiful rolling diapasons in larger buildings, for which the company name became a byword.

 

Organs like this one are a national treasure.  Therefore it is gratifying that it has earned an Historic Organs certificate from BIOS (the British Institute of Organ Studies), and it is shortly to be overhauled by Mander Organs.  It has been described in an engaging article [1] by the present organist, so only a brief summary of its history is included here.  The current stop list is in Table 1.

 

Great (56 notes) Swell (56 notes) Pedal (29 notes)
Open Diapason 8 Double Diapason 16  (TC)   Open Wood 16
Dulciana 8 (TC) Open Diapason 8    (TC)
Stop Diapason Bass 8 (12 pipes)  Stop Diapason 8    (TC) Great to Pedal
Stop Diapason Treble 8 (TC) Principal 4    (TC)
Gamba 8 (TC) Fifteenth 2    (TC)
Principal 4 Oboe 8    (TC)
Flute
Twelfth 2 2/3
Fifteenth 2
  
Swell to Great

 

Table 1.  St Mary Ponsbourne (J W Walker 1858)

 

Notes:

 

  1. No Swell to Pedal coupler.

  2. All swell stops revert to the great Stop Diapason Bass in the lowest octave.

  3. The swell Double Diapason is of stopped pipes.

  4. Mechanical action throughout.

  5. 3 combination pedals to great.

  6. Trigger swell pedal.

 

Summary of history:

 

  1. 1858 - built with console in north transept.

  2. 1860 - pedal compass increased from 20 to 29 notes.

  3. 1867 - tuning changed to equal temperament.  Previous temperament unknown.

  4. 1884 - cleaned and overhauled.

  5. 1887 - organ turned through 90 degrees - console moved to north side of new chancel.

  6. 1909 - cleaned and overhauled.  Great Gamba replaced a three-rank Mixture (composition unknown)

 

In recent years a new blower has been installed, but beyond that the organ is a good example of the longevity which can be expected of a pipe organ by a reputable builder when it is maintained periodically.  Devotees of electronics, eat your hearts out!

 

Some pictures of the church and organ are appended below.  

 

 

 

St Mary Ponsbourne: the church and some great organ pipework

(© St Mary Ponsbourne & P R Minchinton)

 

 

 

St Mary Ponsbourne: console detail - builder's nameplate, candle holder and blower tell-tale, all in brass.  The drawstops are beautifully hand-engraved.

(© St Mary Ponsbourne & P R Minchinton)

 

 

The organ still sounds well in the building.  It does have one or two singular features though, not all of which can be considered advantageous.  In particular, no pipe longer than 4 feet lives inside the swell box (hence the reliance on the Stop Diapason Bass for the lowest octave), so it is perhaps odd that Joseph put in a full length pedal Open Wood at 16 foot pitch.  These aspects must have had something to do with the the chamber he had to contend with, as one pedal pipe was awkwardly mitred to fit.  However few would wish to dispense with this splendid stop in favour of the usual indistinct Bourdon at half the length, no doubt afflicted with a severe cough.

 

 

Digital simulation of the organ

 

At this point you might be wondering why this article appears in the 'electronic organs' category on this website rather than that dealing with pipe organs.  As mentioned in the Introduction, the answer is that the organist at St Mary's, Paul Minchinton, sampled (made separate recordings of each pipe on) the entire organ and kindly provided me with the results.  Realising that the samples represented a complete aural record of an historic organ as it sounded in 2013 just prior to a major rebuild, I decided to simulate it digitally.  Of the 717 samples, one per pipe, virtually all were useable directly in a sampled sound synthesiser.  Only a handful (fewer than ten) had to be discarded.  Each sample was several seconds in duration, recorded in stereo and each included the release transient as the sound died away.  These raw recordings were denoised and then chopped up into separate files, one for each sampled pipe.  They were then individually looped [2] and tuned before importing the complete set into my Prog Organ virtual pipe organ simulation system.  At that point the organ became playable and some tonal finishing such as regulation (adjusting the volume within and between stops) was carried out.

 

Forgive me saying so, but the simulation makes for a highly satisfactory little instrument.  Plaudits for this belong to J W Walker and Sons, not me, as they made these ancient pipes in the first place.  However I also need to seek forgiveness in other ways because I surrendered to the temptation to indulge in - ahem - a little 'creative augmentation' of the stop list as follows:

 

  1. The swell stops were all carried down to bottom C so they no longer need to rely on the Stop Diapason Bass on the great organ.

  2. A Tremulant was added to the swell.

  3. Those stops on the great organ which were of incomplete compass (the Gamba and Dulciana) were also carried downwards.

  4. The separate Stop Diapason Bass and Treble stops on the great were combined into a single Stop Diapason.

  5. A three-rank Mixture was added to the swell (see below for further details).

  6. A 16 foot reed, a Contra Oboe, was added to the swell as a downward extension of the existing Oboe.

  7. A Quint 10 2/3 ft stop was added to the pedal organ, derived from the Open Wood. 

  8. The new swell 16 foot reed, together with three other swell stops, were duplexed on the pedal organ.

  9. The pedal compass was increased from 29 to 30 notes.

  10. A separate Swell to Pedal coupler was added.

 

Other than the Mixture, all of the tonal additions used samples from the original set.

 

With these changes, the revised stop list is in Table 2 below:

 

 

Great (56 notes) Swell (56 notes) Pedal (30 notes)
Open Diapason 8 Double Diapason 16   Open Wood 16
Dulciana Open Diapason 8   Stop Diapason 16       (swell Double Diapason 16)
Stop Diapason  Stop Diapason 8   Quint 10 2/3   (from Open Wood 16) 
Gamba Principal 4   Principal 8         (swell Open Diapason 8)
Principal 4 Fifteenth 2   Fifteenth 4         (swell Principal 4)
Flute 4 Mixture III  (additional samples)   Oboe Bass 16       (swell Contra Oboe 16)
Twelfth 2 2/  Contra Oboe 16  (from Oboe 8) 
Fifteenth 2 Oboe 8   Swell to Pedal                (new)
Great to Pedal
Swell to Great (couples pedals through)  Tremulant          (new)

 

Table 2.  St Mary Ponsbourne (J W Walker 1858) - augmented stop list of the simulated organ

 

 

Regarding the new Mixture, nothing is known of the one originally present on the great when the organ was built (it was replaced by the Gamba in 1909).  Since there was already a Twelfth and a Fifteenth on the great, a logical scheme was to complete the diapason chorus when the manuals are coupled by adding a three-rank Mixture with a composition of 19.22.26 at bottom C, breaking successively at each octave to become 8.12.15 at the top of the keyboard.  This is a 'full' mixture rather than a 'sharp' one, and Joseph Walker included both types in his larger instruments.  Although based on real pipe sounds from another organ, I accept that this is an imported stop which might not have a close affiliation to the one Walker made.  One cannot know.

 

The four duplexed stops on the pedals are all from the swell and currently they are controlled in volume by the simulated swell box, just as they would be if duplexed on a pipe organ.  They are also affected by the Tremulant.  However the Prog Organ system allows their volume to be set independently of those on the swell if desired, though I have not done this because it smacks of cheating in the sense of using digital technology to achieve something that would be impossible with pipes.  All duplexing and extension is also exact in the sense that each sample ('pipe') can only sound once regardless of how many times it might be keyed, the situation which would obtain in a pipe organ. 

 

 

What does it sound like? 

 

The following audio clips might give some idea of how the simulated organ sounds with various stop combinations ranging from soft to loud.  In those below only those stops present on the pipe organ were used (see Table 1), not the new ones which have been added (Table 2).  

 

  Extracts from violin sonata in E (M C Festing) - 4.32 MB/4m 43 s

  Andante Religioso (Henry Smart) - 1.5 MB/1m 38s

 

The augmented stop list (Table 2) was used for the examples below:

 

  Gelobet seist du, Jesu Christ - BWV 604 (J S Bach) - 1.21 MB/1m 19s

 

The instrument has turned out unusually lifelike compared to many digital organs and VPO simulations of my acquaintance.  It is certainly one of the more attractive of the many I have implemented, and in quite a different league to those commercial bland and dreary digitals which are so common.  The Ponsbourne samples are undoubtedly 'live' in the sense of having pronounced, though not unpleasant, variations from note to note.  This feature can take one by surprise while playing it, as you might detect from time to time on my imperfect recordings, almost as though one is seated at the actual tracker action instrument!  In the interests of preserving the sound of the organ as it existed just before the 2013 rebuild I have not modified these attributes, thus the occasional off-speech pipe can still be heard.  One of these has lost its fundamental completely and whistles charismatically away at the twelfth (third harmonic), whereas another struggles to reach stable speech over several seconds!  Some others are over-reticent, and a few tend to scream somewhat.  However some samples included noticeable action clatter at the onset of the note, and this was attenuated deliberately (though not removed completely) when processing the raw samples.  Because the same noises occurred for each sampled pipe supplied by the same pallet, they would otherwise have built up in an unrealistic and unpleasant manner as more stops were added when playing the simulation.

 

The outcome of the exercise confirms my view, arrived at through hands-on experience over many years, that organ simulation using sampled sounds is currently the best way to capture the subtlety of real pipes.  Other methods cannot come as close to rendering the understated delicacies of every pipe on a particular instrument which was achieved in this case.  How can the sounds of the actual pipes be considered inferior to modelling them mathematically, as some claim?  It defeats me how anyone can possibly imagine they can improve them!  Considering its age and the limited work which has been done on it, the pipework on this venerable organ seems in remarkably fine fettle, presumably leaving relatively little for the organ builders to do in this regard when they overhaul the instrument in the near future.  

 

 

Acknowledgement

 

Many thanks to Mr Minchinton for providing the samples of this historic organ, recorded over several sessions in freezing temperatures during the winter of 2012/13, and also for permission to use his photographs in this article.

 

 

References

 

1. "Gleanings from the Cash Book, St Mary's Hatfield: Church Expenses 1874-1923", Paul Minchinton, Organists' Review, May 1999.

 

An updated and expanded version of the text of this article is also available on this website as a PDF download (148 kB) by clicking here.

 

2. 'Looping' is the process whereby a waveform sample of finite duration, no more than a few seconds in length, can be used in a sampled-sound synthesiser to enable notes of any duration to be played.  The sample set producer (me in this case) defines loop-start and loop-end markers on the waveform which are recognised by the synthesiser, which then loops repetitively between them for as long as desired.  The loop points are chosen so that audible discontinuities in the looped sound (clicks, etc) are unnoticeable.  Looping is a standard technique used in digital music.