Homophone
Date Codes, and the ‘Invicta
Problem’.
The Homophone codes were first written up by Peter Hulpusch. (“Dating Homophone Records”: Record Collector
Vol. 24 Nos.3/4, June 1979, pp80-1). In addition, a Homophone dating article by
Peter Copeland appeared in ‘The Historic Record’ – I have a photocopy, but the
number of the magazine is not known. Thanks to the late Arthur Badrock for supplying the reference & photocopy. What
follows is merely an informal ‘gossip’ on the topic – with particular reference
to the early and short-lived period (1912 – 1914) during which Invicta records bore a smiliar
code. As on all pages of this website, accuracy is not guaranteed – alas, I
make frequent errors… 8^)
Unfortunately for us, very few 78 records carry dates to indicate
when they were recorded, plated or pressed. But the German Homophon
Company is a notable exception. Indeed, between about 1905 and 1926, their
discs actually carry two dates: as it
were, a gift to future Discographers! But what the dates actually mean is, as
far as we know, still open to question. It would be very nice indeed if one of
them was a recording date. But in that case, what would the other date signify?
Here is a typical early Homophone as sold in
Well, let us look at the other marking to the right of the label,
which is G10V. Thanks to the pioneering work by Peter Hulpusch,
we know that is a date in the form month-day-year. The number in the middle is
clearly the day of the month, confirmed by the fact that no number larger that
31 is ever found. The first letter had to be the month, because only the twelve
letters A to M (omitting ‘I’) are ever found in this position. (It must be
assumed that A is January, B is February and so on). The final letter,
inevitably, was the year. To cut a long story short, it is known that these
letters run backwards, and seem to be based on Z representing 1901. It is true
that no Homophone record dates back to 1901, but it must be admitted that 1901
is a logical starting point.
Here is a table to help you – and
me! – to decode these dates:
|
Letter |
Month |
Day |
Letter |
Year |
|
A |
January |
1 |
Z |
1901 |
|
B |
February |
2 |
Y |
1902 |
|
C |
March |
3 |
X |
1903 |
|
D |
April |
4 |
W |
1904 |
|
E |
May |
5 |
V |
1905 |
|
F |
June |
6 |
U |
1906 |
|
G |
July |
7 |
T |
1907 |
|
H |
August |
8 |
S |
1908 |
|
J |
September |
9 |
R |
1909 |
|
K |
October |
10 |
Q |
1910 |
|
L |
November |
11 |
P |
1911 |
|
M |
December |
12 |
O |
1912 |
|
|
|
13 |
N |
1913 |
|
|
|
14 |
M |
1914 |
|
|
|
15 |
L |
1915 |
|
|
|
16 |
K |
1916 |
|
|
|
17 |
J |
1917 |
|
|
|
18 |
H |
1918 |
|
|
|
19 |
G |
1919 |
|
|
|
20 |
F |
1920 |
|
|
|
21 |
E |
1921 |
|
|
|
22 |
D |
1922 |
|
|
|
23 |
C |
1923 |
|
|
|
24 |
B |
1924 |
|
|
|
25 |
A |
1925 |
|
|
|
&c. |
|
|
So G10V gives us 10th July 1905. The other date, 25th July 1907,
is long afterwards. A side cannot be recorded after it has undergone another
process later on! Therefore the later date must refer to something that
happened after the recording date.
This leaves a depressingly large number of options open. The
sequence of events in the simple ‘five-step’ process of making gramophone
records is as follows:
1. The original and delicate wax master, which is positive.
2. Electroplating onto this to form a relatively robust matrix, which is negative.
3. Electroplating from the matrix to produce a still more robust mother, which is positive. A number of
mothers may be so derived.
4. Electroplating from the mother to produce a very robust stamper. A number of stampers
may be so derived from each mother.
5. Pressing a (hopefully quite large) number of discs from the stamper, until it wears out & another is needed.
So: which of these two dates refers to which process? What
happened on G10V = 10th July 1905?
It might be: (a) the plating of the matrix from the original wax
master; (a) the plating of a mother from the matrix; (c) the plating of a stamper from the mother; or even (d) a pressing date?
But before delving deeper into these esoteric matters, let us
study a few more Homophones – or Homophons, or Homochords, Homokord, Homocord; the name evolved over the years – in order to
look for general trends in these dates.
We’ll begin with the other side of the above record….
… we see that the date at the bottom is the same as the other
side: 25th July 1907. While the date at the right, E21U, decodes to 21st May
1906. Now here are the 2 sides of a later disc:
The only possible date to be derived from 22713 is 22nd July
1913. On the other side, 1813 might, I suppose, be
construed as 18th January 1903; but as we are pretty certain that Homophone did
not exist at that time, let us choose something more plausible: say 1st August
1913? As to F20O and F22O, they yield 20th June 1912 and 22nd June 1912
respectively.
Now we may go on and examine as many of these Homophone discs as
we can. In fact, we have been doing this, and intend to carry on; and one of
the purposes of this web-page is to ask you to do likewise on Invicta records! More on this later…
When the Great War of 1914-1918 broke out, these German-pressed
records naturally disappeared from the British market. But of course they
continued to be made in
Apart from the fact that the ‘A’ has shifted to the front of the
‘plain date’, and the digits slightly spread to make it even more obviously a
date, nothing has visually changed on these pressings since 1913 – and twelve
years was a long time in the Record Industry in those days! Compare these to
the right-hand black label above. The same double run-off groove still
originates at the top of the label, and sweeps swiftly, within half a
revolution, to the same double locked groove. As to the dates, the plain date
on 8092 reads 26th July 1922; the code G17D gives us 17th July 1922. On 8626,
we have 28th April 1925 and D21A = 21st April 1925.
It’s time to sum up. We can read these dates, for sure. But what
events do they refer to? From looking at many examples of this code, it is
found that the ‘letter code’ date is
always before the ‘plain date’. This has led to the not unreasonable
conclusion – indeed, according to Occam’s Razor, the inevitable
conclusion – that the earlier date may possibly be the recording date.
And that the later one is probably the date when the stamper was
‘grown’, or electroplated, from the mother. (The possibility that the later
date is the date of pressing can, I
think, be firmly ruled out. A big selling record would probably require a press
or presses to work for several days to satisfy the demand for it; and to change
the date code on the metalwork for each new day would be frankly, ludicrous.)
So is the G11P-type date the recording date? Possibly, yes. But
alas, it has been reported that two copies of the same side may have different
codes of this type – see below. There are two possibilities which might account
for this. Firstly, if our hypothetical disc has G11P, but we find another copy
of the same disc bearing, say, K30P, what are we to think? One explanation may
be that the side was originally recorded on G11P = 11th July 1911. And that,
for some reason, a ‘re-make’ was necessary; and this re-make was recorded on
K30P = 30th October 1911. Or secondly,
G11P may refer not to the recording date, but to the date on which the mother
was plated from the matrix. And after having several stampers
produced from it, that first mother was worn out. Accordingly, a second mother
would need to be plated from the matrix. So the later date might indicate when
that second mother was plated. And of course, if a third or fourth &c.
mother was required, each would have a different and ever later date.
Where does this leave our ‘plain date’, which is always later
than the ‘letter-code’ date? There aren’t many options left! As we have ruled
it out as a pressing date, it is most probable that this is the date the stamper was derived from the mother, just as has always
been surmised.
Lastly, can these numbers tell us anything from the way they
appear on the discs: e.g. are they impressed, or embossed? Well, both the letter-code
date and the plain date are in relief (embossed) on the pressings. This means
they have been struck into either the matrix or the stamper.
Anything struck into the matrix would remain in relief on the final pressing.
Likewise, anything struck into the stamper would also
appear in relief on the final disc. For all we know, the letter-code date could
have been struck on the matrix, and the plating code struck into the stamper – they certainly don’t have to be struck at the
same stage of manufacture. Indeed, if the plain code really is the stamper date, then it must of necessity have been struck into the stamper;
there is simply no option there. There remains only the letter-code date to
consider. Surely, it would be ridiculous to keep punching this ‘standard
letter-code’ into all the stampers, in addition to
their date of manufacture? I suggest – though I may be wrong! – that the
letter-code date was impressed onto the matrix itself: and thus appeared,
embossed, on all pressings that were derived from that matrix.
In short, I am suggesting that G11P most likely does not
represent a recording date, but probably the plating date of the matrix. The
immediate question you ask is: ‘How long after the wax was recorded was the
matrix produced?’ And the answer is: ‘As soon as possible!’ Once the original wax master was cut, it was
always a high priority to turn it into a matrix; that is, to convert the
delicate & fragile wax into a more durable metallic form. We can be fairly
sure that in well-regulated recording companies, this process was carried out
almost immediately. It would not do to leave vulnerable waxes lying around for
several days, in case they ‘crept’ out of shape, or even melted in extreme
heat; or simply broke as they were moved about. We can be fairly confident that
a matrix was produced from the wax master if not on the same day, then within a
day or so of the recording.
It follows then, that the same final pressing with a later code,
would be a ‘re-make’.
Mind you: there still remains the possibility that our new matrix
could hang around for some time before its date was stamped into it. Weeks;
months even? It would have been stored in a nice clean & strong envelope,
with all its details written on it, until it was required for use. Perhaps only
then was G11P stamped on it? But still, surely the operative would have stamped
on it the date it was plated, as
written on the envelope? Or maybe – heaven forbid! – the date stamped on it was
the date the factory requested that a mother be made from it? We simply do not
know – or at least, I do not know! Do you? If so, please read on…
Most likely, is that G11P is quite
soon after the recording date? How soon, you may guess as well as I. After
all, recording companies did not usually engage in the practice of making
matrices which they carefully hoarded on their shelves with little intention of
making pressings from them… especially in the ‘Boom Years’ of 1908 – 1914!
We now move on to the second part of this inordinately long &
boring web-page, which concerns the label Invicta.
This was initially a product of the German company Berolina
G. m. b. H, and the sole agency for these discs in this country was taken up by
the new & suitably-named Invicta Record Co., of
which William Barraud was a principal. (He was the
brother of Francis Barraud, the artist who created
the ‘His Master’s Voice’ painting: but that’s not important right now.)
These Invicta records bore a letter
code date. But although appearing exactly the same, i.e.
‘letter-number-letter’, it has proved the bane of many – including the present
writer! – who have tried to interpret it.
Our first fundamental mistake, was to combine long lists of
Homophone letter codes with Invicta letter codes. It
seemed the obvious thing to do: they looked as though they were produced by the
same company: and the more statistics, the more accuracy of analysis &
prediction, surely?
But no. Only confusion resulted from studying these mixed codes.
Eventually it emerged that a variant letter code was in use on Invicta. And to cut a long story short, it looks like a
simple inversion of the Homophone code. (a) Instead of being in the form
‘month-day-year’, they were ‘year-day-month’. (b) Instead of the months being A
to M (omitting I), they were O to Z. (c) Instead of the years starting at Z in
1901 and going to A in 1925( omitting I), they were A for 1901 and Z for 1925
(omitting I). Now why two different companies (Homophone and Berolina) should have date codes so closely related, and
even punched with the same sort of dies, is completely beyond us: so we leave
it to others more skilled. Occam’s Razor would suggest perhaps that Homophone & Berolina were ‘less unrelated’ than we might think, but let
us not speculate.
Anyhow, once this enchantingly logical process of code inversion
was grasped, the rest was easy. It was of course the years that ‘gave the game away’. For unlike Homophone itself, which
made records from 1905 until after its letter code had expired in 1925, Invicta records were produced for only two years or so.
This accounts for only the letters M and N appearing on the letter codes for Invicta discs. Let us look at an Invicta…
The letter code appears below the label – M24R. Thanks to the
generous help of many discographers, dozens of Invicta
letter codes have been studied – far
more than for Homophone – and without further discussion, here is a proposed
table of Invicta Codes.
|
Letter |
Year |
Day |
Letter |
Month |
|
A |
1901 |
1 |
O |
January |
|
B |
1902 |
2 |
P |
February |
|
C |
1903 |
3 |
Q |
March |
|
D |
1904 |
4 |
R |
April |
|
E |
1905 |
5 |
S |
May |
|
F |
1906 |
6 |
T |
June |
|
G |
1907 |
7 |
U |
July |
|
H |
1908 |
8 |
V |
August |
|
J |
1909 |
9 |
W |
September |
|
K |
1910 |
10 |
X |
October |
|
L |
1911 |
11 |
Y |
November |
|
M |
1912 |
12 |
Z |
December |
|
N |
1913 |
13 |
|
|
|
O |
1914 |
14 |
|
|
|
P |
1915 |
15 |
|
|
|
Q |
1916 |
16 |
|
|
|
R |
1917 |
17 |
|
|
|
S |
1918 |
18 |
|
|
|
T |
1919 |
19 |
|
|
|
U |
1920 |
20 |
|
|
|
V |
1921 |
21 |
|
|
|
W |
1922 |
22 |
|
|
|
X |
1923 |
23 |
|
|
|
Y |
1924 |
24 |
|
|
|
Z |
1925 |
25 |
|
|
|
|
|
&c. |
|
|
The only two year letters that are found in Invicta
date codes are highlighted above. It is admitted that the distribution of month
letters is patchy. Only O, P, Q, R, V, W and Z have been found. But still, none
of these stray outside our suggested parameters, and the absence of activity in
the months S, T and U (May, June and July) is tentatively accounted for as
being the Summer months, when the gramophone trade was traditionally quiet, as
opposed to activity in V and W (August and September), preparing for the Autumn
‘Season’ when gramophones & records began to sell again. Followed by a
burst of activity in December, Z, to provide discs for the New Year’s trade?
In the earlier part of 1914, the Invicta
Record Company began its metamorphosis into ‘Guardsman Records’ – which label
continued under the aegis of various
producers until 1928. The separation of Invicta
records from their original supplier, Berolina, is
therefore not to be attributed to the outbreak of the Great War in August1914.
At this catastrophic time, there was naturally a very sharp ‘cut-off’ of cheap
imported records from
What we are looking for is the same disc with different codes. We already know of two Invictas
like this. If you have any Homophones, Invictas, Homochords &c. in your collection which bear letter
codes (‘plain dates’ don’t appear on Invictas as far
as we know), it would be very good indeed if you would please send details –
catalogue number, artist, title, and please list every number that appears ‘in
the wax’, on the label, under the label, &c., and these will added to our
info. in the hope of sorting out exactly what these date codes mean. You will
of course have our grateful thanks, and will receive written credit hereunder.
Besides, updated information, if it becomes available, will be posted on this
page.
Please
send any information, queries, observations &c., to: invicta@normanfield.com Thank
you!
Thanks are already due to: the late Arthur Badrock,
Steven Walker, William Dean-Myatt, Dr. Rainer Lotz, Mike Thomas, Robert Girling,
Joe Moore & others, for supplying much data, and also the inspiration to
attempt to solve this mystery once and for all…
Page
written 31st December 2008. Happy New Year!
Revised
and corrected 11th November 2009.