[GreenKeys] [die at dieconsulting.com: Re: BAUD -vs- WPM]

[David I. Emery]

Date: Fri, 9 May 2014 18:12:28 -0400
From: "David I. Emery" <die at dieconsulting.com>
To: "tony.podrasky" <tony.podrasky at gmail.com>
Subject: Re: [GreenKeys] BAUD -vs- WPM

On Fri, May 09, 2014 at 01:49:54PM -0700, tony.podrasky wrote:
> GA OMs;
> 
> OK - "Riddle Me This, Batman":
> 
>  45.45 BAUD = 60  WPM
> 110.00 BAUD = 100 WPM
> 
> How come, when I print the same (text/file/data), the
> 35-KSR (8-level ASCII) finishes WAY BEFORE the 28-KSR
> (5-level Baudot) printing the same file?
> 
> True: I haven't timed the 2 beasts - but I know that
> the 110 Baud machine finishes more than 2 times before
> the 45.45 Baud machine does.
> 
> "Inquiring Minds Want To know"....


	Very easy explanation .

	A 35/33 KSR is typically running 11 bits per character (two stop
bits) because of timing issues with some character sequences and the
mechanics.   110/11 = 10 characters a second.   Not many mechanical
ASCII ttys were run at 10 bits per character ( 1 stop bit - more common
for electronic serial port ASCII today) though apparently some could.

	A 15/19/28/32 nominally runs at 7.42 bits per character (the
stop pulse is longer 1.42 units).   45.45/7.42 = 6.125 characters a
second.

	One might naively think this would imply a "60 wpm" Baudot tty
should print something in about 163% of the time it takes on a 10
character per second machine.

	BUT !!! there is another difference.   ASCII contains numbers,
letters, punctuation all in ONE character with 128 different
possibilities.   So text with numbers, punctuation and letters mixed
(typical of random English tty text documents) prints all of those and
spaces (blanks) in the same 1 character time per space on the page.

	Baudot has less than 32 printing symbols per character and a
shift unshift mechanism to expand that to the 50 some odd symbols a
Baudot machine can print and a blank or space.   But those shifts and
unshifts are controlled by a character that takes up one character time
- but does not space the carriage forward or print.  So printing a
number or the comma in a sentence involves AT least a FIGS shift, the
number or punctuation symbol and a LETTERS shift (or three characters
instead of one.   And some teletypists and software include more than
one LETTERs shift in case one got missed.   What this implies is that
Baudot has an overhead of the FIGS/LETTERs shifts that ASCII does not.
And this can easily be enough to run the time difference between the two
up beyond 200% (two to one).

	Also it was common to do CR CR LF LETTERs to allow Baudot
machines time to return the carriage and this may not be true of ASCII
text - some ASCII uses a single New Line character to end a line
and space to the next - though mechanical TTYs generally didn't

	And finally depending on what UART is generating the 45.45
Baudot you may or may not be outputting 7.50 bits per character or 8.0
bits per character.   Making digital circuitry do the 7.42 took more
gates in the chip... and was often not provided as a UART capability.  
Mechanical TD's and keyboards of course generally did do 7.42... Granted
this difference would be small (165% versus 163%) but it is there and
measurable.

	Hope this is clear... it is all in the less efficient coding
(more characters on the line required) of Baudot versus ASCII.   Of
course the smaller number of bits in Baudot characters makes it more
competitive on a bit for bit basis... or text/bit sent.

-- 
  Dave Emery N1PRE/AE, die at dieconsulting.com  DIE Consulting, Weston, Mass 02493
"An empty zombie mind with a forlorn barely readable weatherbeaten
'For Rent' sign still vainly flapping outside on the weed encrusted pole - in 
celebration of what could have been, but wasn't and is not to be now either."