The Sound of Bells

Home Page Back to list

Painswick, Gloucestershire, St Mary the Virgin, continued

The previous page described the original Rudhall ten of 1731/2. This ten was augmented to a 12 in 1819, but the two trebles were replaced almost immediately by Rudhalls in 1821, again by Mears in 1887, and most recently in 1993 by Whitechapel. Why should there have been such disatisfaction with the previous three sets of trebles? The Whitechapel bells are analysed here alongside the original Rudhall ten. It will become clear that fitting new bells to old was not straightforward. The 1993 bells, including an extra treble to give a light ten, and the flat 6th provided in 1986, were cast by Whitechapel to Rudhall profiles, and the tuning of them is of some interest.

Recordings of all the bells are available via the tuning table on the previous page.

As noted on the previous page, my thanks are due to Alan Hodges who made me so welcome when I came to record his bells.


First, we should examine the nominals of these bells. The orginal ten are stretched, about 23 cents in the octave. The stretch is not uniform across the peal, it kicks in rather dramatically between 7th and 6th. The closest fit for the tuning of the bells is equal temperament, surprisingly, rather than any version of meantone. The diagram below shows the tuning of all fourteen bells, with the new Whitechapel ones marked with a red 'w':

Painswick nominals - all fourteen

The first point to note is that the flat sixth is neatly fitted mid-way between 6th and 7th, as you would expect. The two trebles to the twelve continue the stretch in the next three or four bells pretty exactly, though this does put the treble nearly 30 cents sharp of the tenor. There was probably no other option for these bells, though they have octave primes, not the flat primes which are probably the explanation for the stretch in the original bells. The extra treble is a little flatter than the trend might predict. However, this bell (and the flat 6) are not meant to be rung with the twelve, and to fully understand the tuning of these bells it is easier to look at the front ten alone:

Painswick nominals - front ten

Here, the tuning of the flat sixth is obviously correct, and the logic behind the tuning of the front three bells becomes clear - the stretch is allowed to run out rather than extend all the way, which would have made the treble a full 35 cents sharp of the tenor, and this in just ten bells. The bell which is furthest from equal is the 8th (the 7th of the twelve), one of the original Rudhall ones. However, this bell is only about 10 cents flat. Given it is the third of the scale, this deviation is towards meantone or just tuning, and quite in order. The flatness of the extra treble also tends towards this tuning. The plot shows that, through careful choice of the nominals of the top bells, a peal has been produced, albeit stretched, with quite acceptable nominal tuning.

Other partials in the new bells

Tuning figures are given on the previous page. All the new bells have octave primes. To flatten these partials after the manner of old-style trebles would have produced a 'screaming' tone. The hums are all major sevenths, very comparable with the original bells and in particular the Rudhall trebles. There is some scatter in the tierces but this is probably inevitable given the old-style profiles.

What is really striking about the front three bells is the superquints and octave nominals. These partials are not tuned, but dictated by the profile of the bell. The front bells show a steady progression in these partials, but they get very flat in the small bells. The octave nominal of the extra treble is quite the flattest I have ever seen in any bell. I did not take any dimensions of these bells, but did note that the extra treble, which hangs in a frame above the others, is a very thick bell indeed. It also has a very large clapper, no doubt to make the bell loud enough, which seems to take up a significant proportion of the mouth of the bell! This effect of thick trebles leading to flat higher partials is present at Llandaff but is even more extreme here. The extra treble has been reclappered - the mark of the previous clapper is visible on the soundbow. The clapper currently installed strikes slightly lower. I was told this was done to improve the sound of the bell.

Intensity profiles of the new bells

Here are the spectra of the sixth (Rudhall), the flat sixth (Whitechapel) and the seventh (Rudhall):

Painswick 6, flat 6 and 7

and the extra treble, the treble and the second:

Painswick sharp treble, 1 and 2

The spectrums of the sixth, flat sixth and seventh show that from the perspective of tuning of all partials, and intensity of the higher partials, the Whitechapel bell is pretty similar to the bells either side. The most noticeable difference is that the newest bell has the flattest hum. The intensities of the lower partials vary quite considerably, but then the sixth is rather different from the seventh in any case.

The spectrums of the front three are remarkably similar, showing the close control Whitechapel have over their profiles. Comparisons with the front three at Llandaff are quite striking. These two sets of bells from the same foundry are about the same weight and date but have quite different spectrums. This is substantiation of the claim, if it were needed, that the Painswick bells were cast to old-style profiles. The Llandaff bells - at least the front two - have strong hums and weak nominals, with not too much tierce either. The Painswick bells have weaker hums and very strong tierces. It is the weaker hums of these bells which saves them. Had they been cast with Llandaff-style strong hums, tuned to the major seventh, they might have sounded quite poor. As it is, the strong octave primes give them a clean strike note, with the strong tierces adding interest. However, it is interesting to compare the sound of the second against the third - the flatter hum in the third clearly has a beneficial effect.


Adding trebles to the orginal ten which complement them as a peal has not been easy, with three failures along the way. I rang on these bells with the previous trebles, but as a child, and do not remember them. However, one can assume that Rudhalls and Mears must have had difficulties both with the considerable stretch in the original ten, and in producing really musical small old-style bells with flat primes and sharp hums. Whitechapel in the latest trebles have reduced the former problem by 'shading out' the stretch at the top end - though adding an extra treble was quite daring! The latter problem they solved by putting in trebles with octave primes - not historically correct but necessary to avoid poor tone. The new trebles sound bright, but not harsh, and have the authentic old-style sound about them. But many of the original ten have near true-harmonic primes, hums, or both; and the tenors, though they have sharp hums, have them muted so as not to spoil the tone. I wonder what true-harmonic trebles would have sounded like?

Click here to return to the previous page covering the original ten.


Last updated October 18, 2001. Site created by Bill Hibbert, Great Bookham, Surrey