24 May
2012
24 May
'12
5:52 p.m.
cctalk-request at classiccmp.org wrote:
Message: 1
Date: Thu, 24 May 2012 11:20:55 +0200 (CEST)
From: Christian Corti <cc at informatik.uni-stuttgart.de>
To: "General Discussion: On-Topic and Off-Topic Posts"
<cctalk at classiccmp.org>
Subject: Re: Software for OCR'ing paper tape?
Message-ID: <alpine.DEB.2.02.1205241119530.20068 at linuxserv.home>
Content-Type: TEXT/PLAIN; charset=US-ASCII; format=flowed
On Wed, 23 May 2012, Jonas Otter wrote:
This was done YEARS ago, in the very earliest microcomputer days.
Jon
Am I completely wrong if I think I remember that
the sprocket holes have a
fixed relationship to the data holes? If so, it ought to be possible to build
an optical decoder, illuminate it with a simple light bulb, connect it to an
input-capable parallel port on a computer with suitable software, and simply
pull the tape through by hand? The sprocket holes would function as a clock
for the data.
Now guess how a classic photoreader (optical paper tape reader) works...
25 May
25 May
11:52 a.m.
Now guess how
a classic photoreader (optical paper tape reader) works...
This was done YEARS ago, in the very earliest microcomputer days. 
1:49 p.m.
On 5/25/2012 12:52 PM, Tony Duell wrote:
_Way_ before that!. There were Elliot optical tape
readers that used this
principle in the 1950s.
Come to think of it, didn;t the Colossus tape reader work this way?
That is a Pre-555 timer era, you don't expect to have paper tape back then.
The Colossus has the fastest paper taper reader out there I expect.
-tony
Ben.
BTW... I love the paper tape quote but I can't find it at the moment
about any faster reading and you got sticky bits of paper tape all
over the place.
26 May
26 May
11:30 a.m.
On 5/25/2012 12:52 PM, Tony Duell wrote:
I expect paper tape goign right back to a 2 track (+sprockes) used for
morse code about 100 years ago :-)
_Way_ before that!. There were Elliot optical
tape readers that used this
principle in the 1950s.
Come to think of it, didn;t the Colossus tape reader work this way?
That is a Pre-555 timer era, you don't expect to have paper tape back then. The Colossus has the fastest paper taper reader out
there I expect.
I think so. The point wwas it didn't haev to start and stop the tape.
That's what complicates high speed readers. If you just letthe tape run
and grab the data as it occures, you can get high speeds very easily.
-tony
noon
On 26 May 2012 at 19:30, Tony Duell wrote:
I expect paper tape goign right back to a 2 track
(+sprockes) used for
morse code about 100 years ago :-)
I remember using a military code practice machine that used a paper
tape, not perforated but with a line printed on it (think of a square
wave trace).
Were there any computer/telecomm devices that used a similar
technology (i.e. ink on paper tape) for I/O?
--Chuck
1:28 p.m.
I remember using a military code practice machine that
used a paper
tape, not perforated but with a line printed on it (think of a square
wave trace).
Pretty standard Boehme inker stuff - in very common use in the 1930s.
Mostly speeds were limited by the photo cells of the day.
Were there any computer/telecomm devices that used a
similar
technology (i.e. ink on paper tape) for I/O?
There were ink to TTY (and TTY to ink) machines back in the 1950s.
--
Will
3:50 p.m.
On 26/05/12 3:00 PM, Chuck Guzis wrote:
On 26 May 2012 at 19:30, Tony Duell wrote:
Not so long ago, film soundtracks worked this way...
--Toby
I expect paper tape goign right back to a 2 track
(+sprockes) used for
morse code about 100 years ago :-)
I remember using a military code practice machine that used a paper
tape, not perforated but with a line printed on it (think of a square
wave trace).
Were there any computer/telecomm devices that used a similar
technology (i.e. ink on paper tape) for I/O?
--Chuck
4:48 p.m.
On 26 May 2012 at 18:50, Toby Thain wrote:
Not so long ago, film soundtracks worked this way...
Excuse me?
I'm familiar with both common sound-on-film systems (Fox andWestern
Electric Variable density and RCA Photophone Variable Area).
Neither uses ink. The RCA system used a galvanometer-and-mirror
setup that recorded a variable-width-spot on the film sound track.
The WE system used a variable-width-slit type of galvanometer light
valve that varied the amount of light passing; the Fox system used a
gas-discharge tube to provide varying light intensity. The RCA
system was not affected to the extent of the VD systems were by the
developing and copying process, but RCA was not quite as sensitive.
Both date from the late 20s/early 30s and can be played back using
the same equipment.
But no ink-on-paper.
--Chuck
6:08 p.m.
On 26/05/12 7:48 PM, Chuck Guzis wrote:
On 26 May 2012 at 18:50, Toby Thain wrote:
Thanks for the extra detail.
Not so long ago, film soundtracks worked this
way...
Excuse me?
I'm familiar with both common sound-on-film systems (Fox andWestern
Electric Variable density and RCA Photophone Variable Area).
Neither uses ink. The RCA system used a galvanometer-and-mirror
setup that recorded a variable-width-spot on the film sound track.
The WE system used a variable-width-slit type of galvanometer light
valve that varied the amount of light passing; the Fox system used a
gas-discharge tube to provide varying light intensity. The RCA
system was not affected to the extent of the VD systems were by the
developing and copying process, but RCA was not quite as sensitive.
Both date from the late 20s/early 30s and can be played back using
the same equipment.
But no ink-on-paper.
Right. I just meant a visual encoding. But it was analogue, and not
computer or telecoms, so my answer was not responsive anyway. :)
--Toby
--Chuck
6:23 p.m.
On Sat, 26 May 2012, Toby Thain wrote:
Right. I just meant a visual encoding. But it was
analogue, and not
computer or telecoms, so my answer was not responsive anyway. :)
Well, if we allow film, then howzbout:
Vannevar Bush and his "Memex" or
Emmanuel Goldberg's "Rapid Selector" (which unlike Bush's, actually
existed)
Both were optical encoding of half a dozen bits accompanying each fram of
film
8:04 p.m.
On 26 May 2012 at 18:23, Fred Cisin wrote:
Well, if we allow film, then howzbout:
Vannevar Bush and his "Memex" or
Emmanuel Goldberg's "Rapid Selector" (which unlike Bush's, actually
existed)
Both were optical encoding of half a dozen bits accompanying each fram
of film
Fred--I'm sure you heard tales of terror from the Lawrence Berkeley
and Livermore crowds about the IBM 1360 Photostores in use there.
Best device since the degausser for destroying archival data. "Your
file has been photostored was something nobody wanted to read."
But no, I was thinking about ink-on-paper, digital read/write
recording.
--Chuck
29 May
29 May
12:29 p.m.
At 08:23 PM 5/26/2012, Fred Cisin wrote:
Both were optical encoding of half a dozen bits
accompanying each frame of film
Nothing exceeds like excess, kids!
Let's mount a digital video camera / webcam with a short focus over
a nice, smooth, safe, paper tape path. The attached computer just
records a movie from the webcam. Maybe there's an adjustable guide to
handle various tape widths.
Along with the image of the tape, it records a roller linked by
friction to the motion of the hand-pull or reel-pulled tape.
The roller has encoding so we can tell how fast the tape is moving.
Presto! With a little smart software, it handles 5-, 6-, 7-, 8-level,
any width, fanfold, roll, mylar or paper, with and without sprockets,
variable-sprocket, and chad-less. It'll handle old printing telegraph
tape, Morse, and Baudot code tapes, too, so this device would have a
wider market than classic computer buffs. Few pesky mechanical parts
to misalign, jam, break, or damage the old tape.
Plus you get an archival video of the actual tape, preserving the ephemeral
handwriting, vintage manufacturer markings, or tape with teletype printing.
Hmm. Maybe the deluxe version has two cameras, front and back.
Does it need front- or back illumination to aid the process? Maybe the
two illuminations are different colors. After all, we're recording video
in color, right? It's like we're making a photocopy of the tape for your
museum. Future emulators can play the video to more fully recreate
the classic computing experience. A future paper-tape punch mated with
an ink-jet printer could make a copy of a tape that looks just like the
original.
Obviously, two full video cameras are overkill if you had very fast
scanner-type linear pixel arrays.
What is the longest paper tape, in feet and bytes?
- John
1:16 p.m.
At 08:23 PM 5/26/2012, Fred Cisin wrote:
>Both were optical encoding of half a dozen bits accompanying each frame of film
On Tue, 29 May 2012, John Foust wrote:
Nothing exceeds like excess, kids!
Actually, the examples that I gave (Bush's Memex, Goldberg's Rapid
Selector) were in response to a tangent, and not about the orginal
subject. Both examples (more than 50 years ago) were metadata for
finding the right frame of microfilm. Bush lifted the idea, without
attribution, from Goldberg (see Buckland). But Bush's description
(that incorrectly implied actual consturction) became his "As We May
Think" paper that was arguably the seminal work of modern IR, and
inspired Ted Nelson, who . . . (watch "Hyperland"!! (BBC pre WWW
feature about future of internet by Douglas Adams and Ted Nelson,
starring Tom Baker) )
30 May
30 May
11:19 a.m.
Obviously, two full video cameras are overkill if you
had very fast
scanner-type linear pixel arrays.
FWIW, linear CCDs are considerably easier to drive than 2-dimanesional
ones. THis might be a factor if, like me, you'd ant to do the whole thing
yourself :-)
What is the longest paper tape, in feet and bytes?
The standard reels of paper tape were, I think, 1000 feet long, That's
120,000 characters (== bytes), since the standard is 10 character/inch. A
little over 117K (for 'real' K :-))
-tony
26 May
26 May
10 p.m.
Running a paper tape based system.
In the early 1970's I worked for a company called C&N Electrical based near
where I lived (and still do live) in Newbury UK.
They were a division of Camper & Nicholson a boat building company. (They
made beautiful wooden yachts.) Anyway the bit I worked for did a large
range of electronic subcontract work.
They had a contract to test and repair core store memory units from computer
systems made by a UK manufacturer called Elliot Automation.
The test computer system was made up of two or three racks, not 19" but more
like the wide lateral filing cupboards you see in offices. There was what I
would call an I/O station that had two high speed tape readers, two tape
punches and two IBM Golf ball typewriters without keyboards.
Guess who got the job of testing and repairing the memory units? The memory
themselves consisted of a unit about three feet long, twelve inches high and
six inches deep. At either end were plug in PCB's and in the middle the core
stack in a sealed box. These were made by a then well known company called
Mullard. I think (forty years is long time) they were 4k each.
I would take a unit, stick it on the bench, put some long cables on it and
connect it in the place of one of the good memory units in the rack.
I had a test tape that would load via one of the readers. The tape was
rolled up into a spool, you put that on a feed unit (a metal pin in a lump
of wood), the end of the tape then went into the reader. There was a tape
load switch that fed the tape to the start position. Then you flipped the
system load switch and the tape read into the system. Finally you hit the
system run switch and the program would start.
Once through the reader the loose tape was caught in a large box. I would
then fish out the start of the tape and rewind it with a manual movie film
rewinder
The program once loaded read and wrote the store with known patterns and you
could diagnose or set up the store read/write electronics (lots of OC71 and
OC81 transistors)
After repairing or setting up the unit I ran a confidence test which
consisted of a program designed to calculate the handicaps for large yacht
racing. You loaded the program tape and then a data tape. The output printed
on one of the golf ball RO typewriters.
?
Rod Smallwood
?
?
-----Original Message-----
From: cctech-bounces at classiccmp.org [mailto:cctech-bounces at classiccmp.org]
On Behalf Of ben
Sent: 25 May 2012 21:49
To: cctalk at classiccmp.org
Subject: Re: Software for OCR'ing paper tape?
On 5/25/2012 12:52 PM, Tony Duell wrote:
_Way_ before that!. There were Elliot optical tape
readers that used this
principle in the 1950s.
Come to think of it, didn;t the Colossus tape reader work this way?
That is a Pre-555 timer era, you don't expect to have paper tape back then.
The Colossus has the fastest paper taper reader out there I expect.
-tony
Ben.
BTW... I love the paper tape quote but I can't find it at the moment
about any faster reading and you got sticky bits of paper tape all
over the place.
4589
days inactive
4594
days old
15 comments
9 participants
participants (9)
-
ard＠p850ug1.demon.co.uk -
bfranchuk＠jetnet.ab.ca -
cclist＠sydex.com -
cisin＠xenosoft.com -
elson＠pico-systems.com -
jfoust＠threedee.com -
rodsmallwood＠btconnect.com -
toby＠telegraphics.com.au -
wdonzelli＠gmail.com