All the previously mentioned makers produced fine music but the
actual tuning scale varied considerably.  The one thing that is
consistent is that most were not consistent in the way that actual
intervals, note-to-note, were created!  What John Powell did in his
limited comb tuning analysis was quite unique; he created a common
base line of 440 cycles per second [Hz].  He then plotted the results
of frequency of individual teeth against the base line.

The result shows clearly that there were a variety of ways in which
combs were tuned.  Where one would expect some sort of linear or
curvilinear result from base to treble, few did but they all employed
stretch.  John concluded that they did not use a master comb but tuned
by ear.  He plotted 8 charts for disc and cylinder musical boxes and
makers such as (possibly) Bremond, Rivenc, Nicole (i.e., Pierre-Moise
and David-Elie, not Francois) a Reymond-Nicole (son-in-law of
Françoise), Lecoultre Frères (Francois-Louis, I think).

I believe that, in the early days of comb tuning there were no master
combs.  If one studies the late H.A.V. Bulleid's date charts, which are
quite accurate and well tested, the Geneva makers accounted for some
fairly high and consistent rates of production during the early cartel
years.

Bulleid estimated Nicole, Ducommun and F. Lecoultre Frères at about
at about 900 per year with Henri-Joseph Lecoultre at 200 per year and
brother David Lecoultre at 300 per year.  That was quite a lot of combs
being tuned -- say, at least 3000 each year in a very small area of
Geneva.  If so, that meant about 600 a week and about 100 a day.
Nobody seems to know how long it took to tune a comb, but those with
100 to 200 teeth must surely have taken long hours and maybe several
days.

Thus it is uncertain exactly when master combs were introduced.
There is little doubt that they were in common use later but to what
extent they became an absolute basis for tuning or just a reference for
further tuning by ear is not known.  What we do know is that disc box
combs have a pre-determined gamme to suit all possible interchangeable
discs, but the tuning within the gamme is not always consistent; it's
close enough to sound okay but actually is slightly different when each
tooth pitch is analysed -- clearly indicative of final tuning by hand.

I also believe there were 'master tuners' with probably a few
apprentices.

Thus we arrive back to the ear again.  The comb arrives at the tuner's
workshop, possibly with or without leads but probably with the brass
comb base and comb base scratched to indicate the pitch of each tooth,
the octaves, the incidentals and the teeth to be tuned to the same
pitch; perhaps also with the dampers fitted.

There would have been no cylinder and no musical score -- just the
comb!  But, the tuner would have had feedback sufficient to know what
was required from the comb he was about to tune.  How many times had he
tuned for, say, 'The Last Rose Of Summer,' or any other popular air of
the time?  How many times would an order come in for the 'tune of the
day' with the expectation of other orders?  (We know batch production
was quite common.)  So the tuner does his work; he hears the result and
maybe makes a few changes, then he proceeds to tune combs for the
remainder of the batch.

These were skilled people.  Is it any different in principle to the
way top-class violin makers carved, shaped, and glued together the
pieces of wood, tapping with a finger or scraper as they went, using
their ears to detect the slight changes in resonance, knowing that
the final case with its tensioned strings would resonate to every note
the skilled player would make?

I believe the tuners developed their skills through some sort of
apprenticeship with a master tuner by hearing the result of their work,
learning about the nuances demanded by the arranger's score and being
'told-off' by the maker if he did not reach the standard of tuning
required!

Paul Bellamy
Kent, UK