4 Feb
2011
4 Feb
'11
8:42 a.m.
In addition, while I could easily afford a modern
printer to transplant
into the 4631 case, why should I when I have a perfectly good printer
already, plugged into my PC?
I'd prefer to take the raster image and save it in a suitable file
format (TIFF?) for putting on a PC and printing on any old printer.
This would make it much more flexible, and I think it would be of
benefit to do so.
If a graphics file is the end format, it seems to me the easy way out
Is to use a modern PC running Tek 4010/4014 emulation software (anything
From MS-Kermit to xterm) sniffing the serial traffic
and just do
screen dumps from there.
For the vector purists (e.g. me) making SVG or Postscript or PDF or some
True vector format, would be a desirable goal. But with endpoint resolution
Only 1024x768 probably overkill. I still hate seeing what I know are
True vector lines, get rasterized.
Tim.
4 Feb
4 Feb
9:47 a.m.
Tim, quoting me:
On a 4006/4010/4012/4014, you have a point. On a 4051/4052/4054, there
_isn't_ a serial stream to sniff!
In addition,
while I could easily afford a modern printer to transplant
into the 4631 case, why should I when I have a perfectly good printer
already, plugged into my PC?
I'd prefer to take the raster image and save it in a suitable file
format (TIFF?) for putting on a PC and printing on any old printer.
This would make it much more flexible, and I think it would be of
benefit to do so.
If a graphics file is the end format, it seems to me the easy way out
Is to use a modern PC running Tek 4010/4014 emulation software (anything
From MS-Kermit to xterm) sniffing the serial traffic and just do
screen dumps from there. For the vector purists (e.g. me) making SVG or
Postscript or PDF or some
True vector format, would be a desirable goal. But with endpoint resolution
Only 1024x768 probably overkill. I still hate seeing what I know are
True vector lines, get rasterized.
Isn't it 4096*4096 on the 4014? It is on the 4054.
More to the point, though, rasterising the vector screen is exactly what
those old Tek printers did! They scanned the surface of the tube in a
raster fashion, and read out what they found there. I don't know how
well the analogue circuitry preserved the grey scale - my 4631 hasn't
worked since I got it - but done well, this could do pretty good
staircase elimination.
I wonder. How about a PC that sits on the GPIB and emulates a 4662?
That could indeed preserve true vectors!
Just a thought,
Philip.
5 Feb
5 Feb
7:05 a.m.
I wonder. How about a PC that sits on the GPIB and
emulates a 4662?
That could indeed preserve true vectors!
The Tektronix terminals (as opposed to the 405x computers) don't have n
GPIB interface for the 4662. And the analogue storage displays certain;y
ahve no way to drive this plotter.
I didn;t think that consumables for the 4662 were difficult to find. From
what I remember, the pen carriage has a thereded insert, and the thread
will take a standard Rotring drawing pen. It's best to remove the barrel
from the ben and just fit the nib/ink cartridge, of
course.
I have no diea what the multi-pen version takes, I only have the
single-pen one.
-tony
6 Feb
6 Feb
6:44 p.m.
In article <4D4C3BB8.6000207 at axeside.co.uk>,
Philip Belben <philip at axeside.co.uk> writes:
More to the point, though, rasterising the vector
screen is exactly what
those old Tek printers did! They scanned the surface of the tube in a
raster fashion, and read out what they found there. I don't know how
well the analogue circuitry preserved the grey scale - my 4631 hasn't
worked since I got it - but done well, this could do pretty good
staircase elimination.
There is no grayscale in a storage tube monitor. There's charge
stored on the tube, or there isn't.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
9:45 p.m.
Has it not occurred to anybody to take the 4014 serial drawing stream as the source?
There are only a few commands there.
I used it to plot roller coasters in a commercial program, COASTER!
Embedding the tek file was much smaller than including 3d graphics, I just plotted tek
4010 to the screen.
Randy
To: cctalk at classiccmp.org
Subject: Re: Tektronix 4010/4014 terminal printer project
From: legalize at xmission.com
Date: Sun, 6 Feb 2011 19:44:18 -0700
In article <4D4C3BB8.6000207 at axeside.co.uk>,
Philip Belben <philip at axeside.co.uk> writes:
More to the point, though, rasterising the vector
screen is exactly what
those old Tek printers did! They scanned the surface of the tube in a
raster fashion, and read out what they found there. I don't know how
well the analogue circuitry preserved the grey scale - my 4631 hasn't
worked since I got it - but done well, this could do pretty good
staircase elimination.
There is no grayscale in a storage tube monitor. There's charge
stored on the tube, or there isn't.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
7 Feb
7 Feb
11 a.m.
In article <BLU145-W19811820E20016D7E33386BAEB0 at phx.gbl>,
Randy Dawson <rdawson16 at hotmail.com> writes:
Has it not occurred to anybody to take the 4014 serial
drawing stream as the
source?
Yes; that's not the point.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
12:27 p.m.
Has it not occurred to anybody to take the 4014 serial
drawing stream
as the source?
There are only a few commands there.
I used it to plot roller coasters in a commercial program, COASTER!
Embedding the tek file was much smaller than including 3d graphics, I
just plotted tek 4010 to the screen.
Yes, Randy, it occurred to Tim Shoppa a couple of days ago.
The objection is still the same as it was when he suggested it: it's a
great idea, but ONLY IF you're trying to copy the screen of a serial
terminal. If you're using a self-contained computer like my 4052, or an
analogue storage tube, there's no serial drawing stream to take.
Philip.
6 Feb
6 Feb
10:53 p.m.
New subject: Gray scale in storage tube, was Re: Tektronix 4010/4014 terminal printer project
--- On Sun, 2/6/11, Richard <legalize at xmission.com> wrote:
There is no grayscale in a storage tube monitor.?
There's charge stored on the tube, or there isn't.
This may be the case for the Tektronix storage tubes,
but storage tubes that can handle grayscal have been
made, e.g., the Hughes Tonotron. It could capture a
television image.
--Bill
7 Feb
7 Feb
12:33 p.m.
More to the
point, though, rasterising the vector screen is exactly what
those old Tek printers did! They scanned the surface of the tube in a
raster fashion, and read out what they found there. I don't know how
well the analogue circuitry preserved the grey scale - my 4631 hasn't
worked since I got it - but done well, this could do pretty good
staircase elimination.
There is no grayscale in a storage tube monitor. There's charge
stored on the tube, or there isn't.
8 Feb
8 Feb
7:16 a.m.
In article <4D50570A.7080704 at axeside.co.uk>,
Philip Belben <philip at axeside.co.uk> writes:
Well, the whole process in the original equipment is analog. You're
not going to see aliasing due to discrete sampling, because there
isn't any. You will see aliasing due to the bandwidth of the analog
signals, but there isn't anything changing so fast that its outside
the bandwidth of the circuitry.
There is no
grayscale in a storage tube monitor. There's charge
stored on the tube, or there isn't.
Sorry, I wasn't clear. I didn't mean greyscale stored on the tube. I
meant, can it read out how much of a pixel is covered by drawn lines and
interpret that as greyscale? For that matter, how big are the pixels?
There aren't any pixels.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
12:09 p.m.
On 2011 Feb 8, at 7:16 AM, Richard wrote:
Philip Belben <philip at axeside.co.uk>
writes:
Well, the whole process in the original equipment is analog. You're
not going to see aliasing due to discrete sampling, because there
isn't any. You will see aliasing due to the bandwidth of the analog
signals, but there isn't anything changing so fast that its outside
the bandwidth of the circuitry.
I know what you mean here, so not to argue your point about the analog
process in the system under discussion, but I would like to add an
historical footnote: even analog raster-scan systems were characterised
in terms of "picture elements" going back to the very early days of TV.
A book I have here ("Principles of Television Engineering" / Fink /
McGraw-Hill/ 1940) presents the scientific and engineering background
to TV systems at the time. The phrase "picture element" is used
throughout. The then-current transmission standard was characterised as
441 lines by 400-600 picture elements per line, "that 3,000,000 to
6,000,000 elements must be sent in a second", etc. The characteristics
of other mediums such as photographs are also presented in terms of
picture elements.
When "picture element" started getting abbreviated to "pixel", I
don't
know.
There is
no grayscale in a storage tube monitor. There's charge
stored on the tube, or there isn't.
Sorry, I wasn't clear. I didn't mean greyscale stored on the tube. I
meant, can it read out how much of a pixel is covered by drawn lines
and
interpret that as greyscale? For that matter, how big are the pixels?
There aren't any pixels.
2:15 p.m.
In article <d7c81ea7b43ac08503dccfb4a518a78e at cs.ubc.ca>,
Brent Hilpert <hilpert at cs.ubc.ca> writes:
On 2011 Feb 8, at 7:16 AM, Richard wrote:
Sure, but its not a TV system. There is some limit to the resolution
of a 463{1,2} printer. There is also some limit to the resolution of
a DVST in a Tektronix product. The former is more limited to the
printing mechanism and the latter is more limited to the electron beam
width and addressability circuitry in the storage tube.
The way "pixel" is used canonically these days, there aren't any in a
storage tube system. Nor are there any pixels in a refresh vector
display. There is the idea of the smallest addressable unit on the
screen, but that's not the same as a pixel.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
Philip Belben <philip at axeside.co.uk>
writes:
I know what you mean here, so not to argue your point about the analog
process in the system under discussion, but I would like to add an
historical footnote: even analog raster-scan systems were characterised
in terms of "picture elements" going back to the very early days of TV. For that matter, how big are the pixels?
There aren't any pixels.
9 Feb
9 Feb
8:57 p.m.
On 2011 Feb 8, at 2:15 PM, Richard wrote:
Brent Hilpert <hilpert at cs.ubc.ca>
writes:
See next message.
On 2011 Feb 8, at 7:16 AM, Richard wrote:
Sure, but its not a TV system. There is some limit to the resolution
of a 463{1,2} printer. There is also some limit to the resolution of
a DVST in a Tektronix product. The former is more limited to the
printing mechanism and the latter is more limited to the electron beam
width and addressability circuitry in the storage tube.
The way "pixel" is used canonically these days, there aren't any in a
storage tube system. Nor are there any pixels in a refresh vector
display. There is the idea of the smallest addressable unit on the
screen, but that's not the same as a pixel. Philip Belben <philip at axeside.co.uk>
writes:
I know what you mean here, so not to argue your point about the analog
process in the system under discussion, but I would like to add an
historical footnote: even analog raster-scan systems were
characterised
in terms of "picture elements" going back to the very early days of
TV. For that matter, how big are the pixels?
There aren't any pixels.
10 Feb
10 Feb
11:58 a.m.
In article <d7c81ea7b43ac08503dccfb4a518a78e at cs.ubc.ca>,
Brent Hilpert <hilpert at cs.ubc.ca> writes:
Actualy, I would say the 4631 printer is very close to being a TV system.
As I understnad it (and it's been some years since I looked at the
schematics), the printer casues an electron beam (from the flood gun?) to
be scanned over the storage target in a raster. This produces a signal
indicating the charge level on the target for that point. And the printer
has a 1-dimentinal CRT whic his sanned in synchronism to the raster
lines, its image is recorded o nthe apepr. The vertical 'deflection' in
the printer is obtained by advancing the paper.
No matter how the image in the storage tube is put there, the printer is
a raster devvice. There is a limited vertical resolution from the number
of lines in taht readout raster. I assume the horizontal resolution is
limited only by the bandwidth of the system.
-tony
On 2011 Feb 8, at 7:16 AM, Richard wrote:
Sure, but its not a TV system. There is some limit to the resolution
of a 463{1,2} printer. There is also some limit to the resolution of Philip Belben <philip at axeside.co.uk>
writes:
I know what you mean here, so not to argue your point about the analog
process in the system under discussion, but I would like to add an
historical footnote: even analog raster-scan systems were characterised
in terms of "picture elements" going back to the very early days of TV. For that matter, how big are the pixels?
There aren't any pixels. 
11:04 p.m.
The picture element measurement consists of the number of horizontal lines
multiplied by the horizontal resolution determined by the ability to
distinguish the transitions of a square wave as it increases in frequency.
Normally it was done by sweeping a square wave, increasing in frequency
along each line from left to right.
It's really a measure of video bandwidth. My last job as an engineer before
moving to DEC was designing character based VDU's.
Regards
?
Rod Smallwood
?
?
-----Original Message-----
From: cctech-bounces at classiccmp.org [mailto:cctech-bounces at classiccmp.org]
On Behalf Of Tony Duell
Sent: 10 February 2011 19:58
To: cctalk at classiccmp.org
Subject: Re: Tektronix 4010/4014 terminal printer project
In article <d7c81ea7b43ac08503dccfb4a518a78e at cs.ubc.ca>,
Brent Hilpert <hilpert at cs.ubc.ca> writes:
Actualy, I would say the 4631 printer is very close to being a TV system.
As I understnad it (and it's been some years since I looked at the
schematics), the printer casues an electron beam (from the flood gun?) to
be scanned over the storage target in a raster. This produces a signal
indicating the charge level on the target for that point. And the printer
has a 1-dimentinal CRT whic his sanned in synchronism to the raster
lines, its image is recorded o nthe apepr. The vertical 'deflection' in
the printer is obtained by advancing the paper.
No matter how the image in the storage tube is put there, the printer is
a raster devvice. There is a limited vertical resolution from the number
of lines in taht readout raster. I assume the horizontal resolution is
limited only by the bandwidth of the system.
-tony
On 2011 Feb 8, at 7:16 AM, Richard wrote:
Sure, but its not a TV system. There is some limit to the resolution
of a 463{1,2} printer. There is also some limit to the resolution of Philip Belben <philip at axeside.co.uk>
writes:
I know what you mean here, so not to argue your point about the analog
process in the system under discussion, but I would like to add an
historical footnote: even analog raster-scan systems were characterised
in terms of "picture elements" going back to the very early days of TV. For that matter, how big are the pixels?
There aren't any pixels.
11:54 p.m.
On 2011 Feb 10, at 11:04 PM, Rod Smallwood wrote:
The picture element measurement consists of the number
of horizontal
lines
multiplied by the horizontal resolution determined by the ability to
distinguish the transitions of a square wave as it increases in
frequency.
This is the basis or principle laid out in the 1940 book for the
horizontal resolution. For the vertical they give analyses that reduce
the resolution by a factor relative to the number of scan lines.
Various other considerations such as worthwhile H/V resolution ratios
(in addition to the aspect ratio) are included.
Normally it was done by sweeping a square wave,
increasing in frequency
along each line from left to right.
It's really a measure of video bandwidth. My last job as an engineer
before
moving to DEC was designing character based VDU's.
12 Feb
12 Feb
9:09 p.m.
Ohh god stop this discussion.
Tony, I find you kinda offensive, and always brain dumb, with your replies to me.
If the objective is a true printout, the digital data is available up the chain in the
system.
If you want a cool screen print effect, use a digital camera and shoot, permanent, make a
scope camera frame out of plywood.
For those of you that are serious about Tektronix equipment and resources, we are all here
in Beaverton for your serious restoration efforts.
Randy
From: ard at p850ug1.demon.co.uk
Subject: Re: Tektronix 4010/4014 terminal printer project
To: cctalk at classiccmp.org
Date: Thu, 10 Feb 2011 19:58:28 +0000
In article <d7c81ea7b43ac08503dccfb4a518a78e at cs.ubc.ca>,
Brent Hilpert <hilpert at cs.ubc.ca> writes:
Actualy, I would say the 4631 printer is very close to being a TV system.
As I understnad it (and it's been some years since I looked at the
schematics), the printer casues an electron beam (from the flood gun?) to
be scanned over the storage target in a raster. This produces a signal
indicating the charge level on the target for that point. And the printer
has a 1-dimentinal CRT whic his sanned in synchronism to the raster
lines, its image is recorded o nthe apepr. The vertical 'deflection' in
the printer is obtained by advancing the paper.
No matter how the image in the storage tube is put there, the printer is
a raster devvice. There is a limited vertical resolution from the number
of lines in taht readout raster. I assume the horizontal resolution is
limited only by the bandwidth of the system.
-tony
On 2011 Feb 8, at 7:16 AM, Richard wrote:
Sure, but its not a TV system. There is some limit to the resolution
of a 463{1,2} printer. There is also some limit to the resolution of Philip Belben <philip at axeside.co.uk>
writes:
> For that matter, how big are the pixels?
There aren't any pixels.
I know what you mean here, so not to argue your point about the analog
process in the system under discussion, but I would like to add an
historical footnote: even analog raster-scan systems were characterised
in terms of "picture elements" going back to the very early days of TV. 
10:35 p.m.
On Sunday, February 13, 2011, Randy Dawson wrote:
Ohh god stop this discussion.
<various grumbling and ironic statements snipped>
Welcome to cctalk...?
Pat
--
Purdue University Research Computing --- http://www.rcac.purdue.edu/
The Computer Refuge --- http://computer-refuge.org
13 Feb
13 Feb
8:40 a.m.
Ohh god stop this discussion.
Who died and put you in charge of the list?
No I will not stop any discussion because you (or anyone else other than
the list owner) tells me to. IMHO this discussion is very much on-topic
here. It is disucssing a peripheral for a classic computer system, or
possibly a comptuer system itself.
Nobody is forcing you to read or contribute to this discussion. If you
don't like it, read something else.
Actually, if you'd read the messages you would realise that one of your
comments is downright wrong.
Tony=2C I find you kinda offensive=2C and always brain dumb=2C with your re=
plies to me.
I can't find anything that I would consider remotely offensive in that
bit you quoted. Perhaps you can explain how my description (which I
believe to be corrent, if it isn;'t then you are welcome to give factual
correctlions) offends you.
If the objective is a true printout=2C the digital data is available up the=
chain in the system.
And that's the stantement which I consider to be totally wrong. In the
case of a 4014 (or whatever) terminal, sure you vcan grab the incoming
serial data stream. I suppose for a 405x computer you could grab the data
going into the DACs, but I don;t think many 405x owners would want to
make permanent modifications to their machines to do this (not if there's
another solution available).
But as I have mentioend, I have a Tektronix large-screen (12" or so)
storage display. It has analogue X-Y inputs. It has a screen printer
connector such as we are discussing. The input unit I have for this
display is an 8 channel multiplexer and timebease. The thing behaves as
an 8-channel storage 'scope (probbly around audip bandwidth).
There is no digital stream in any part of that unit.
Yes, I could replace the whole thing with an 8-input ADC, etc. But that's
not the same system at all.
If you want a cool screen print effect=2C use a digital camera and shoot=2C=
permanent=2C make a scope camera frame out of plywood.
tht is one way, if you have a digital camera. But I have never believed
that probkems only have one solution,. What you do one way, another
person may do a differnt way.
This is a hobby for most of us. We do it becuase we enjoy it. As such,
its up to the individual how to spend his time and money. If somebody
wants to inteface to the printer connector on a storage display, then who
are you (or I) to stop them. If somevody wants to photograph the screen,
ditto.
Some people might claim that spending hours keeping old hardware running
when there a re perfectly good emulators out there is a waste of time and
money., Maybe it is. But that is not a reason for not doing it. I can
think of a dozen hoobies that I don't wish to spend my time and money
doing. But that certain;y doesn't mean others shouldn't do them if they
so wish.
For those of you that are serious about Tektronix equipment and resources=
=2C we are all here in Beaverton for your serious restoration efforts.
Fien. Do you have a specification of the hard copy unit connector that we
are discussing? Do you have details of the special paper for the 4631
printer (Is it possible to re-create it)?
-tony
9 Feb
9 Feb
2:17 a.m.
On Tue, 8 Feb 2011, Brent Hilpert wrote:
I know what you mean here, so not to argue your point about the analog
process in the system under discussion, but I would like to add an historical
footnote: even analog raster-scan systems were characterised in terms of
"picture elements" going back to the very early days of TV.
That's exactly the point: those Tektronix terminals/computers are not
raster-scan systems, you draw a line from here to there. The only
"limitation" is the addressing range for the start and end point (either
10 or 12 bits). Speaking of the printer, AFAIK it's the printer that
determines the scanning speed and the resolution of the rasterization
process.
For that matter, how big are the pixels?
There aren't any pixels. systems at the time. The phrase "picture
element" is used throughout. The
then-current transmission standard was characterised as 441 lines by 400-600
picture elements per line, "that 3,000,000 to 6,000,000 elements must be sent
... which is not quite correct since the horizontal "resolution" depends
on the bandwidth, the beam spot size and the size of the screen surface.
in a second", etc. The characteristics of other
mediums such as photographs
are also presented in terms of picture elements.
?!? I know that the grain size matters, but where can I find those picture
elements on a film strip?
Christian
8:57 p.m.
On 2011 Feb 9, at 2:17 AM, Christian Corti wrote:
I don't know why you would say it's not correct. The numbers are out of
the book and refer to the abilities of the then-current RMA TV
Committee standards for transmission, they account for things like
channel bandwidth.
It is a slightly different meaning than 'pixel' today, it's a measure
of resolution: that which can be resolved; rather than a fixed grid of
points on the image medium.
It's not an assertion by me, I'm just pointing out how these things
were characterised in 1940: even for analog systems they did so in
terms of a matrix or discrete count of "picture elements". There is a
whole chapter in the book on image analysis and it is more complex than
what I present here, it doesn't correspond 1:1 in the V dimension to
the number of scan lines for example.
"A single frame of 35mm motion picture film contains about 500,000
picture elements when exposed, developed and projected in the usual
manner."
"A fine 'contact' photographic print of 8*10in dimensions contains as
many as 50,000,000 picture elements."
On Tue, 8 Feb 2011, Brent Hilpert wrote:
I know what you mean here, so not to argue your point about the
analog process in the system under discussion, but I would like to
add an historical footnote: even analog raster-scan systems were
characterised in terms of "picture elements" going back to the very
early days of TV.
That's exactly the point: those Tektronix terminals/computers are not
raster-scan systems, you draw a line from here to there. The only
"limitation" is the addressing range for the start and end point
(either 10 or 12 bits). Speaking of the printer, AFAIK it's the
printer that determines the scanning speed and the resolution of the
rasterization process.
(And ultimately that 10 or 12 bits does place a maximum on the V*H
resolution of the drawn image.)
For that matter, how big are the pixels?
There
aren't any pixels. systems at the
time. The phrase "picture element" is used throughout.
The then-current transmission standard was characterised as 441 lines
by 400-600 picture elements per line, "that 3,000,000 to 6,000,000
elements must be sent
... which is not quite correct since the horizontal "resolution"
depends on the bandwidth, the beam spot size and the size of the
screen surface. in a
second", etc. The characteristics of other mediums such as
photographs are also presented in terms of picture elements.
?!? I know that the grain size matters, but where can I find those
picture elements on a film strip?
10 Feb
10 Feb
1:53 a.m.
On Wed, 9 Feb 2011, Brent Hilpert wrote:
No, you've got that wrong. The bits only affect the starting and end point
of a vector but *not* the vector itself!
I don't know why you would say it's not correct. The numbers are out of the
book and refer to the abilities of the then-current RMA TV Committee
standards for transmission, they account for things like channel bandwidth.
Sure, and the newspaper is always right ;-)
Since TV is/was purely analog, the horizontal resolution only depend on
the bandwith and the beam focus.
It is a slightly different meaning than 'pixel' today, it's a measure of
resolution: that which can be resolved; rather than a fixed grid of points on
the image medium.
It's not an assertion by me, I'm just pointing out how these things were
characterised in 1940: even for analog systems they did so in terms of a
matrix or discrete count of "picture elements". There is a whole chapter in
No, they didn't. Maybe in the US, but usually they refered to the number
of scan lines ("vertical resolution" if you like) and the bandwidth
(e.g. 5.5 MHz here or around 10 MHz in France).
That's
exactly the point: those Tektronix terminals/computers are not
raster-scan systems, you draw a line from here to there. The only
"limitation" is the addressing range for the start and end point (either 10
or 12 bits). Speaking of the printer, AFAIK it's the printer that
determines the scanning speed and the resolution of the rasterization
process.
(And ultimately that 10 or 12 bits does place a maximum on the V*H resolution
of the drawn image.) systems at the time. The phrase "picture
element" is used throughout. The
then-current transmission standard was characterised as 441 lines by
400-600 picture elements per line, "that 3,000,000 to 6,000,000 elements
must be sent
... which is not quite correct since the horizontal "resolution" depends on
the bandwidth, the beam spot size and the size of the screen surface. in a second", etc. The characteristics of other
mediums such as
photographs are also presented in terms of picture elements.
?!? I know that the grain size matters, but where can I find those picture
elements on a film strip? the book on image analysis and it is more complex than
what I present here,
it doesn't correspond 1:1 in the V dimension to the number of scan lines for
example.
Well, the vertical dimension depends of the size of the CRT, the number of
scan lines remains the same (625 lines according to the CCIR). Some of
these lines are "drawn" outside the visible area of the CRT as they
contain the vertical sync and the blanking period (the B and S in (C)VBS).
"A single frame of 35mm motion picture film
contains about 500,000 picture
elements when exposed, developed and projected in the usual manner."
"A fine 'contact' photographic print of 8*10in dimensions contains as many
as
50,000,000 picture elements."
Sounds like marketing figures...
Christian
10:56 a.m.
On 2011 Feb 10, at 1:53 AM, Christian Corti wrote:
I don't know why you would say it's not correct. The numbers are out
of the book and refer to the abilities of the then-current RMA TV
Committee standards for transmission, they account for things like
channel bandwidth.
Sure, and the newspaper is always right ;-)
Since TV is/was purely analog, the horizontal resolution only depend
on the bandwith and the beam focus.
It is a slightly different meaning than 'pixel' today, it's a measure
of resolution: that which can be resolved; rather than a fixed grid
of points on the image medium.
It's not an assertion by me, I'm just pointing out how these things
were characterised in 1940: even for analog systems they did so in
terms of a matrix or discrete count of "picture elements". There is a
whole chapter in
No, they didn't. Maybe in the US, but usually they refered to the
number of scan lines ("vertical resolution" if you like) and the
bandwidth (e.g. 5.5 MHz here or around 10 MHz in France).
Geez.
Systems here were described in terms of scan lines and bandwidth too.
But the more detailed analyses being done at the time involved as I
have quoted. The book was not a newspaper, nor is it a marketing
brochure or paperback tract. It's a 500-page hardcover book from a
reputable publisher, targeted at engineers, giving a quite thorough
technical compendium of the state of the art at the time. I would be
surprised if "picture element" originated with this book, not that that
would obviate my point.
You can lead a horse to water...
On Wed, 9 Feb 2011, Brent Hilpert wrote:
No, you've got that wrong. The bits only affect the starting and end
point of a vector but *not* the vector itself!
I am well aware of how vector displays work.
That's exactly the point: those Tektronix terminals/computers are
not raster-scan systems, you draw a line from here to there. The
only "limitation" is the addressing range for the start and end
point (either 10 or 12 bits). Speaking of the printer, AFAIK it's
the printer that determines the scanning speed and the resolution of
the rasterization process.
(And ultimately that 10 or 12 bits does place a maximum on the V*H
resolution of the drawn image.) systems at the
time. The phrase "picture element" is used
throughout. The then-current transmission standard was
characterised as 441 lines by 400-600 picture elements per line,
"that 3,000,000 to 6,000,000 elements must be sent
... which is not quite
correct since the horizontal "resolution"
depends on the bandwidth, the beam spot size and the size of the
screen surface. in a second", etc. The characteristics of other
mediums such as
photographs are also presented in terms of picture elements.
?!? I know that the
grain size matters, but where can I find those
picture elements on a film strip? the book on
image analysis and it is more complex than what I present
here, it doesn't correspond 1:1 in the V dimension to the number of
scan lines for example.
Well, the vertical dimension depends of the size of the CRT, the
number of scan lines remains the same (625 lines according to the
CCIR). Some of these lines are "drawn" outside the visible area of the
CRT as they contain the vertical sync and the blanking period (the B
and S in (C)VBS).
"A single frame of 35mm motion picture film
contains about 500,000
picture elements when exposed, developed and projected in the usual
manner."
"A fine 'contact' photographic print of 8*10in dimensions contains as
many as 50,000,000 picture elements."
Sounds like marketing figures... 
6:54 a.m.
At 10:57 PM 2/9/2011, Brent Hilpert wrote:
It is a slightly different meaning than 'pixel'
today, it's a measure of resolution: that which can be resolved; rather than a fixed
grid of points on the image medium.
And when you're talking about NTSC color, the black-and-white info
has a very different resolution and capability than the color.
- John
2 p.m.
>It is a slightly different meaning than
'pixel' today, it's a measure
>of resolution: that which can be resolved; rather than a fixed grid of
>points on the image medium.
On Thu, 10 Feb 2011, John Foust wrote:
And when you're talking about NTSC color, the
black-and-white info
has a very different resolution and capability than the color.
I've missed parts of this thread, but as I understand it, the argument is
about using "pixels" when referring to a verctor display, with no raster.
("Vector: n. A carrier that transmits a disease from one party to
another")
As such, "pixels" doesn't really apply to the line segments.
BUT, it is not unreasonable to use "pixels" to refer to the accuracy with
which the end-point locations can be specified. And amybe even as one of
several possible ways to measure the width of the line segments. Unlike
in math, in the real world, "points" and "lines" DO have width (even
if
only as the "circle of confusion"), and not very many can dance on the
point of a pin.
--
Grumpy Ol' Fred cisin at xenosoft.com
13 Feb
13 Feb
1:24 p.m.
Fred commented:
I've missed parts of this thread, but as I
understand it, the argument is
about using "pixels" when referring to a verctor display, with no raster.
I've missed parts too, since I only subscribe to cctech, not cctalk; but
we are actually talking about the device that scans the storage tube and
prints what it finds on paper. This is a raster device, not a vector
device.
When I asked how big the pixels were, I seem to have opened a can of
worms. Perhaps I can contain some of the worms by re-stating my question:
How many scan lines does the 436x print per paper inch? How many scan
lines does it read per screen shot? Does the latter figure depend on
the screen being copied (i.e. does it actually scan the large screens at
a higher resolution, to match the higher resolution of the vector
addressing)? How many dots can it resolve in one scan line? Does this
too depend on the screen it is scanning?
I hope this will clear up some of the discussion about pixels. I also
hope it will get me a more useful answer than "there aren't any".
Philip.
10 Feb
10 Feb
8:41 a.m.
In article <alpine.DEB.2.00.1102091106490.29908 at linuxserv.home>,
Christian Corti <cc at informatik.uni-stuttgart.de> writes:
[...] Speaking of the printer, AFAIK it's the
printer that
determines the scanning speed and the resolution of the rasterization
process.
Correct. The analog ramps to scan out the tube are sent by the printer.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
4 Feb
4 Feb
11:47 a.m.
I'd prefer
to take the raster image and save it in a suitable file=20
format (TIFF?) for putting on a PC and printing on any old printer.=20
This would make it much more flexible, and I think it would be of=20
benefit to do so.
If a graphics file is the end format, it seems to me the easy way out
Is to use a modern PC running Tek 4010/4014 emulation software (anything
From MS-Kermit to xterm) sniffing the serial
traffic and just do
screen dumps from there.
11:50 a.m.
For the vector purists (e.g. me) making SVG or
Postscript or PDF or some
True vector format, would be a desirable goal. But with endpoint resolution
Only 1024x768 probably overkill. I still hate seeing what I know are
True vector lines, get rasterized.
Actually, IIRC, the 4631 is raster, not a vector device, It scans an
electro beam across the storage target in the (vecotr display) CRT and
reads out what's stored. And IIRC it does it a scan line at a time and
prints the image using a strange, flat, 1-axis CRT onto photosensitive
paper with the other axis being achieved by the paper movement.
-tony
6 Feb
6 Feb
6:42 p.m.
In article <B136EDE3DF5EC441B6F08E0A7AB872450BADE72CB6 at EX2K7-CMS-1.wmata.local>,
"Shoppa, Tim" <tshoppa at wmata.com> writes:
If a graphics file is the end format, it seems to me
the easy way out
Is to use a modern PC running Tek 4010/4014 emulation software (anything
The point is to make a hardware compatible printer interface for the
existing terminal.
From MS-Kermit to xterm) sniffing the serial
traffic and just do
screen dumps from there. From the perspective you advocate, why bother obtaining
the terminal
at all? Just run the emulator.
--
"The Direct3D Graphics Pipeline" -- DirectX 9 draft available for download
<http://legalizeadulthood.wordpress.com/the-direct3d-graphics-pipeline/>
Legalize Adulthood! <http://legalizeadulthood.wordpress.com>
5061
days inactive
5070
days old
30 comments
12 participants
participants (12)
-
ard＠p850ug1.demon.co.uk -
cc＠informatik.uni-stuttgart.de -
cisin＠xenosoft.com -
hilpert＠cs.ubc.ca -
jfoust＠threedee.com -
legalize＠xmission.com -
pat＠computer-refuge.org -
philip＠axeside.co.uk -
rdawson16＠hotmail.com -
rodsmallwood＠btconnect.com -
tshoppa＠wmata.com -
wmaddox＠pacbell.net