Tower bell ringing is not possible at the moment in the UK and US because of coronavirus. People still seem to be ringing in Australia and New Zealand. Even handbell ringing depends on having a whole band living in the same house. People have been speculating about using videoconferencing systems for remote handbell ringing. Is this possible?

Back in 2013 there was a handbell peal on a video link between Derby and Edinburgh. Tina and I were invited to ring in it, but it took place while we were away on our summer holiday, so we had to decline. That peal used a special high-performance video link between two universities, and I think it was facilitated by a research project on high-speed networking. But is it possible with the kind of standard videoconferencing that many of us are using increasingly at work, especially with the current travel restrictions?

Dorothy and I have just done a little experiment between two rooms in the house, using Zoom. The results were not good. One problem, which is really a show-stopper in itself, is that there was a significant delay over the network, making it impossible to maintain a rhythm. Each of us was ringing our bells as soon as we saw/heard the other person's bells ring, but each of us experienced a long delay (maybe up to a second) between ringing our bells and hearing the other person's bells. The other problem is that after the first strike of each bell, there was a large decrease in the volume, perhaps because the software was trying to reduce feedback. Even when we both used headsets, the same thing happened.

So, we have recorded our Plain Hunt Minimus on BellBoard for posterity, but videoconferencing isn't going to be a substitute for getting together.

Even if there were no delays over the network, another potential problem would be having bells at each end with compatible tuning. In the past, when trying to make rings of 14 or 16 by combining more than one set of bells, I have found that it's possible for two bells from different sets to be marked with the same note but be a semitone apart in their actual pitch. I expect this wouldn't be a problem with two sets from the same founder.

We did well with our quarter of CYNR this week, so I would like to go for a peal of it during the handbell weekend that I'm planning in April. Here's a survey of compositions to choose from. First, some essential criteria. I'm looking at tenors-together all-the-work compositions. It's possible to get all the work in a five part composition (either an exact five part, or sometimes with a variation in one part in order to adjust the length), so I will consider five-parts to reduce the learning required. CompLib has 12 compositions satisfying these conditions (searching in CompLib actually gives 15 results, but I have combined some of them which are variations). I will classify them first according to the part ends, then by other factors. I have removed the musical analysis as I am mainly interested in how easy the compositions would be to call and ring, but it can be found on CompLib.

1. Part ends 12345678, 13526478, 15634278, 16452378, 14263578

This part end group consists of the lead ends of Plain Bob Minor with 78 added at the end. I find them easy to check because of familiarity with Bob Minor, and their coursing orders are familiar because we ring a lot of compositions with blocks of 5 befores, which produce the same coursing orders. There are 8 compositions with this part end group. I will subdivide them according to the calling.

1.1. The calling of Middleton's (2M 2W 3H)

There are 5 compositions with this calling. It potentially gives 5600 if every course is 7 leads. Some of the compositions are indeed for 5600, shortened to 5056 in the same way that Middleton's composition of Cambridge is usually shortened: by replacing 2M 2W with B in one part. One is shortened to 5024 by shortening some courses. Others use short courses and then lengthen some of them to get a length of 5024.

1.1.1. Compositions reduced from 5600 to 5056 by replacing 2M 2W with B

5056 by Robert D S Brown

M  W  H             23456
-------------------------
-        NNN.NYNY   43652
-  -     CCC.N.YCY  56234
   -  -  YYYY.NYN.  23564
      -  CNCCCYC.   52364
      -  RRRRRRR.   35264
-------------------------
5 part.
In one part, replace 2M 2W by B and omit the associated leads.
1344 Y, 1312 C, 1280 N, 1120 R
80 changes of method; all the work.

The next one has rather more Lincolnshire. We rang the full length of 5600 for my 600th peal.

5056 by Robert D S Brown

M  W  H             23456
-------------------------
-        CNC.YNNN   43652
-  -     NCN.C.YCC  56234
   -  -  YYYY.CYY.  23564
      -  CNNNNYC.   52364
      -  RRRRRRR.   35264
-------------------------
5 part.
In one part, replace 2M 2W with B and omit the associated leads.
1440 N, 1248 C, 1248 Y, 1120 R
79 changes of method; all the work.

1.1.2. Compositions reduced from 5600 to 5024 by shortening courses

This is the first composition, shortened in a different way which results in much more Rutland.

5056 by Robert D S Brown

M  W  H               23456
---------------------------
-        NNN.N[YNY]   43652
-  -     CCC.N.[YCY]  56234
   -  -  YYYY.[NYN].  23564
      -  CNCCCYC.     52364
      -  RRRRRRR.     35264
---------------------------
5 part.
In parts 3, 4 and 5, replace the bracketed methods with R.
1408 R, 1344 C, 1152 N, 1120 Y
74 changes of method; all the work.

1.1.3. Compositions extended from 4860 to 5024

The first of these two compositions is rung quite frequently on handbells. Peter Randall seems to like it. It has the fewest changes of method of any of these compositions, which can help with stability. Also it looks easy to learn. It can be made into a palindrome by either changing the fourth course to NNNCCCC or changing the fifth course to CCCNNNN.

5024 by Robert W Lee

M  W  H               23456
---------------------------
-        YYY.YR       43652
-  -     R.RRRRR.[R]  56234
   -  -  RY.YYY.      23564
      -  NNNNCCC.     52364
      -  CCCCNNN.     35264
---------------------------
5 part.
In part 2, replace [R] with NNN.
1408 R, 1280 Y, 1216 N, 1120 C
31 changes of method; all the work.

The next one has more changes of method, but a similar method balance.

5024 by Emma J Southerington

M  W  H             23456
-------------------------
-        [R].CCNY   43652
-  -     CCC.YR.    56234
   -  -  RRRRRR.R.  23564
      -  NNNNNYC.   52364
      -  YYYYYCN.   35264
-------------------------
5 part.
In part 1, replace [R] with NYN.
1408 R, 1312 Y, 1184 N, 1120 C
66 changes of method; all the work.

1.2. The calling B, extended with blocks of 3M and 3H

This is an attractive idea because of the simplicity of a B as a 5-part transposition, and then the composition is extended with two simple blocks of three bobs. If you think 5120 is too long, it's easy to shorten it by two leads. We rang the full length for the first peal of spliced by the Albany Quadrant band, in 2011, and again with Julia in 2012.

5120 (5056) by John S Warboys

M  B  H               23456
---------------------------
-        R.Y[NNN]     43652
-        YYY.CCCC     63254
-  -  -  NNC.RR.RRR.  23564
      -  CYNNYYY.     52364
      -  CCRNN.       35264
---------------------------
5 part.
1440 N, 1280 C, 1280 Y, 1120 R
74 changes of method; all the work.
For 5056 replace [NNN] with R in part 3.
1344 N, 1280 C, 1280 Y, 1152 R

1.3. The calling W B 2H, extended with blocks of 3H and/or 3M

These compositions again use B to give the 5-part structure. The W and 2H are a block of three bobs on the same three bells, but the calling positions are different because the B shifts all the bells. The first composition uses one more block of three to increase the length. CompLib also has a trivial variation in which the third lead of one part is replaced by Y. Similarly to the previous composition, this one can be shortened by substituting a lead of R in one part.

5120 (5056) by Philip G K Davies

B  W  H              23456
--------------------------
   -  -  CCCY.CCC.   45236
      -  YCYCNNN.    24536
      -  YYNN[YYY].  52436
-     -  NN.NY.      52364
      -  RRRRRRR.    35264
--------------------------
5 part.
1344 Y, 1280 C, 1280 N, 1152 R
69 changes of method; all the work.
For 5056 replace [YYY] with R in part 1. 
1344 N, 1280 C, 1280 Y, 1152 R

The next composition uses more short courses and therefore needs an extra 3M to get the length. It has the most changes of method of any of these compositions.

5120 by Donald F Morrison

M  B  W  H                     23456
------------------------------------
      -  -  CYYN.CYC.          45236
         -  YRR.               24536
         -  NCRNN.             52436
   -     -  CN.YN.             52364
3        -  R.CYCRC.YNRYN.RY.  35264
------------------------------------
5 part.
1440 Y, 1280 C, 1280 N, 1120 R
144 changes of method; all the work.

2. Part ends 12345678, 13456278, 14562378, 15623478, 16234578

This part end group consists of cycles on 23456 with 8 added at the end. The part ends should be easy to check, but I am not familiar with the coursing orders.

5024 by Robert D S Brown

M  W  H                 23456
-----------------------------
   2     RY.YNRCY.[R]   35426
   -  -  RC.R.          42356
   -     CR.NCY         54326
2     -  NYN.NNRNY.RN.  56423
      -  YCRNC.         45623
-----------------------------
5 part.
Replace [R] with CYC in part 3.
1440 N, 1408 R, 1152 Y, 1024 C
137 changes of method; all the work.

3. Part ends 12345678, 14653278, 15236478, 13462578, 16524378

I don't see anything special about this part end group, either as changes or as coursing orders, and I don't find it appealing. Maybe I am missing something.

5056 by Robert D S Brown

M  W  H             23456
-------------------------
-     -  NCN.NNNC.  64352
   -  -  RY.YYY.    35642
      -  [RRRRRRR].   63542
   -  -  CYCN.CCN.  54632
      -  YNYYR.     65432
      -  RCR.       46532
-------------------------
5 part.
Replace bracketed leads with RCR in parts 3, 4 and 5.
1280 N, 1280 R, 1280 Y, 1216 C
100 changes of method; all the work.

The final two compositions have a simpler calling and are similar to each other, with significantly different numbers of changes of method.

5056 by Robert D S Brown

M  W  H             23456
-------------------------
   -  -  CCCN.YNY.  45236
      -  NYNCNYN.   24536
      -  YNYCR.     52436
-  -  -  R.Y.R.     63425
      -  RYCCC.     46325
      -  [RNR].     34625
-------------------------
5 part.
Replace bracketed leads with RRRRRRR in parts 4 and 5.
1280 C, 1280 R, 1280 Y, 1216 N
115 changes of method; all the work.

5056 by Robert D S Brown

M  W  H             23456
-------------------------
   -  -  CCCC.NNN.  45236
      -  YYYNYYY.   24536
      -  NNNNR.     52436
-  -  -  R.Y.R.     63425
      -  RCCCC.     46325
      -  [RYR].     34625
-------------------------
5 part.
Replace bracketed leads with RRRRRRR in parts 4 and 5.
1280 C, 1280 N, 1280 R, 1216 Y
60 changes of method; all the work.

Conclusion

I like the Robert Lee composition for its simplicity, and I think I will call it next time we ring these methods. I'm not sure why I didn't choose it for our first peal of spliced. Maybe I didn't like the idea of the method substitutions in one part. When searching for quarter peal compositions, two callings come up that aren't used as the basis for any of these peals: M H W and W M. It could be interesting to explore them.

Yesterday was the annual SACR training day, at Tulloch. It's primarily a tower bell event, but we also offered some handbell ringing. In most cases this was plain hunting on six, including some with absolute beginners, as well as some Bob Minor. Good progress was made.

We also rang some Bristol Major with me, Jonathan, Angela and Nick, and half a course of Cambridge Royal with the addition of Jenny. The Cambridge wasn't perfect, but we rang with a good rhythm and recovered quickly from mistakes. We've really come a long way with our 10-bell ringing.

Inevitably there was a lot of tower bell ringing in which the rhythm was far from perfect, and the contrast with the handbell ringing reminded me of how much easier it is to get an excellent rhythm on handbells. We still find 12-bell ringing difficult, but we've improved enormously on 10, and on 8 we can ring very well indeed, even if there are a few trips.

Tomorrow we're ringing on 8, as Angela isn't free. This is another good thing about having 5 people now - there should be fewer weeks when we can't ring at all. The plan for tomorrow is a quarter of Cambridge, Yorkshire, Lincolnshire and Rutland, which will be an advance for Peter. The first suggestion was Cambridge, Yorkshire and Lincolnshire, but this is an interesting situation in which adding a method makes it easier to find a simple composition. Jonathan's going to call it, and I don't know which composition he will choose, but with CYNR it's possible to get a simple 5-part composition with the calling M H W or W M, using a 3-lead course (2 leads of Rutland and 1 lead of another method) and a 5-lead course (1 lead of Rutland and 4 leads of other methods).