On 9/30/19 7:05 PM, Bob kb8tq wrote:
Hi
Based on only dimly remembered conversations long long ago:
Getting all the “message fragments” so they sound natural and not choppy is
not quite as easy as it seems at first. It’s by not quite rocket science, but
there
is more fiddling involved than one might think.
One “solution” is to use fewer fragments and record larger portions of the
message.
Back in the day, storage limited your ability to record every message “full up”.
Typically one records 60 phrases - that is, rather than recording
twenty, thirty, forty, and one, two, three and try to assemble them -
you record all the numbers 00,01,02,03...
After all, your time is *free* to do the recording, and it makes the
software to play it back easier (you don't have to figure out "is this
less than 20 in which case play phrase[0] through phrase[19], and then
play phrasetens[t/10] + phrase[t mod 10]
Assuming you record the “at the stroke the time will be” only once, the rest is
under 3 seconds of audio. At maybe 16 bits / 32K sps. (yes that’s overkill).
this comes
up just under 200 K bytes. Recording the full time message for every minute of
the
day would be less than 270 megabytes.
That would be a bit tedious..
There are, of course, myriad text to speech programs of varying quality
available for just about every OS imaginable. Although, in 10 minutes
of casual browsing, I've not yet found one for Latin - a suitable voice
intoning the time in Latin periodically would seem to be a good addition
to any time cave.
Google translate does speak Latin, although I've not had my family
classics scholar evaluate it. To my ear, it seems to have a distinctly
Italianate accent (a terminal vowel added to words), although it does
say "Timeo Danaos et dona ferentes" and "Carthago delenda est" correctly.
That’s a pretty small flash drive ….
Bob
On Sep 30, 2019, at 4:00 PM, Neville Michie <namic...@gmail.com> wrote:
Here in Australia we are suffering the loss
of one of the significant developments in accurate time keeping and
dissemination.
The talking clock, built in England, with sound tracks on rotating glass disks,
has been on the Australian telephone system for more than half a century.
The system was timed by quartz oscillators, synchronised to the local
observatory time.
Now in spite of the trivial cost of maintaining the system it has been removed
by
the money-hungry telco which took over the government run telephone system.
Now it occurs to me that the sound tracks occupy a very small digital space, and
with modern flash drives and a little logic the talking clock could be driven by
any time nut's disciplined time source.
So is there a time nut who could design a voice output that we could all use?
“At the third stroke the time will be…”
cheers,
Neville Michie
_______________________________________________
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe, go to
http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe, go to
http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
_______________________________________________
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe, go to
http://lists.febo.com/mailman/listinfo/time-nuts_lists.febo.com
and follow the instructions there.
