braille output hardware (jf)(long post)

* braille output hardware (jf)(long post)
@  emc.er
   ` Dave Mielke
                   ` (2 more replies)
  0 siblings, 3 replies; 5+ messages in thread
From: emc.er @  UTC (permalink / raw)
  To: blinux-list

Hi blinux-list_ers,

The following is a very long post on the subject of braille output
hardware.

First I will say, I am absolutely in awe of Y'all, installing linux
sightless. Windoze, Icons, Mice, and Pointers=WIMP!
It is still an adventure to me with 20/20.

Q. Why am I here on blinux-list???
A. I have an old idea for a braille tty that has resurfaced.
   I don't do 'windoze' if possible, so here I am on blinux!!!

I'm an inventor (read 'currently unemployed network analyst').
I have a bucketful of ideas for you to ponder, and questions.
I do retain any rights granted by US patent law regarding any ideas
discussed below that are new and/or unique and hence possibly
patentable. After 1 year from today March 9th 2002, when I am making
public disclosure, if any patent is not applied for, any patentable
unique and new ideas(claims) of mine become public domain!
I can prove posession of the ideas a long time into the past also.
Having said that, I have no desire to exploit any rights to the
detriment of the
blinux community. Open Source RULES!
Even a patented idea can be produced by any individual, just not in
damaging quantity or sold for profit. Why does this paragraph exist.
I have a US patent #4926998 for a "coin sensor for pay telephone", and
it started out as a very simple idea, but it was original enough to be
patentable. I really am an inventor, just not a financially successful
one. You might be able to persuade a local HAM or HW geek to construct a

prototype of what I describe for you for the cost of parts.,only a
couple/few
hundred dollars US.
If I am describing old or already patented ideas, please let me know.

Forgive the property rights tangent, I'll get to the main point of the
posting now.

About 20 years ago while working as a computer bench technician I came
up with the idea of putting special braille keys on a regular keyboard
instead of the 'f' and 'j' keys. I thought to use something like the
guts from a bunch of used dot matrix print heads to make scrolling
braille under the index finger keys.
The idea is/was to read without ever moving your fingers from the
keyboard. A thumb mounted scroll speed/cursor control.
This thought train started by thinking it a waste of 8 good pins in a
faulty 9 pin print head!

Now, year 2002, while working on another unrelated electronic widget,,,
I had some new ideas for the old 'new' idea!!!
With the computer technology so advanced today, someone must have
something like this available already, I figured it HAD to be!!!

I can't find anything even close discussed anywhere on the net!!!
Am I looking in the wrong place or are my ideas unique?

I have learned of braille tty's but they appear to me to be only up to
80 static charachers per display and do not scroll past your finger(s).
I have heard of grid graphic output devices with crude 1 bit per pixel
resolution.
this is the same as black and white only output. Up or down, no in
between height.
One thing leads to another, and I have a bunch of new ideas!
Everything available appears to be mostly servo-mechanical in some way.

I am now doing some serious research on the subject.

I have a bunch of questions. I will number them and try to go from
general to specific. I can answer linux questions in return.
After the questions I'll describe the ideas in much more detail and
possibly
put the questions in a better perspective.

1. Is a keyboard with scrolling braille 'home' keys available anywhere
today?
2. Will speech output displace the braille tty completely?
2a. Is braille reading speed faster than listening to audio?
3. Do all braille tty's cost many many thousands of dollars?
3a. If they were one tenth the cost would they become more prolific?
4. What is the smallest size character easily read by most fingers.
5. How tall must each dot be? What is the shortest easily readable?
5a. Would a shorter vibrating dot be reliably detectable.
5b.  Is it easier or necessary to use 2 or more fingers to read???
6. Do the graphic output grids work? There is a 64 by 64? grid for sale.

   (see way at the bottom for more on the subject of graphic output)
7. Is there any braille tty that even scrolls the characters past your
fingers?
8. Where do I find the braille 'standards' and are they reasonable or
   bogus.

Currently available LCD (Liquid Crystal Display) technology
right now, with slight modification can do what I describe.
LCD technology has 2 main components the actual liquid crystal display
(liquid crystals sealed between 2 thin pieces of clear glass and a X Y
grid of conductors to activate individual grid elements), and a seperate

LCD driver chip to translate an ascii character into a correct 'font'
grid pattern.
The LCD driver chips are inexpensive and can display any character set
you put in memory including braille.
There are 2 main flavors of LCD driver chip readily available.
They make character models, typically 40-80 characters by 4 lines.
They also make graphic models that do larger grids but can also do
custom character sets in many sizes.
The 5 column by 7 row matrix of the typical LCD ascii display character
lends itself well to the 2 colum by 3 row braille format.
Top to bottom 4 spaces and 3 dots available. Left to right has options.
Left to right can be customizable to reading speed!.
The actual LCD display elements are in a 5 by 8 grid with the bottom row

8 being for display of cursor position, The driver chips already know
this.  A top row cursor would be easy to impliment, and would be needed
for
inverted usage anyway.
The LCD driver chips can already slide characters across the LCD one row

or column at a time. If you just replace the LCD with something TACTILE
the rest is easy!

The "something TACTILE" is the key component.
Come up with a drop in replacement for the LCD screen and many many
consumer electronic items become braille friendly.
A few important ones are, GPS, wrist watch, clock, thermostat,
thermometer, scale, caller ID, compass, microwave, timer, radio and TV
channel, alarm panel, light meter, voltage and/or current meter,
radar, sonar, speedometer(joke???), many others I take for granted!!!
They even put a silly LCD on a water filter that goes on a kitchen sink.

The drop in replacement for the LCD could have many forms depending on
the application.
If you place a grid of hose ends instead of dots you could use air
pressure or vacume to produce the feeling. They could be baloon covered
or open ended.
If you use a grid of peltier junctions you could use either heat OR cool

to represent dots. The junctions require lots of current so mobility
would require wheels on all the batteries.
Straight high frequency/voltage ultra low current AC electricity could
be used, but this could easily be dangerous. If not outright dangerous,
unpopular sounding enough to be avoided.
If you use a tiny grid of linear solenoids or servo's you just have a
tiny braille
tty but it could scroll.  With servo's multiple heights are possible.
They don't make many dot print heads any more either, so a source of
free or
inexpensive prototype parts is gone.  They are just on or off anyways.
Moving mass takes energy, and this is the highest mass solution.

If you use a grid of piezo crystals you could use height and/OR
vibrations. A piezo is what goes beep in most computers and is also used

in many small speakers.  "extra" piezo elements, those not under a
finger,
might/could even speak what you are reading???

The piezo crystal grid seems like the natural choice as it is lowest
power compared to the other ways. A protective covering could also serve

to silence any unwanted audio frequency noise. The ON vibration
frequency and
amplitude could easily be varied to suit different users or
environments.

Piezo elements change shape when stimulated with electricity, and make
electricity when compressed/expanded. This could possibly be exploited
to detect fingers and only activate the under finger dots reducing power

consumption/heat/noise. The piezo grid elements can also easily be
surrounded by a conductive grid to detect finger presssure capacitance
or resistance.

It may even be possible and/or feasible to use the current LCD
technology with only a special tactile covering added.
I will try to explain.
An LCD creates no light or display energy on it's own.
It is a grid of small elements that are activated in sequence to create
character shapes by the adjacent ON elements forming recognizable
shapes.
It either blocks the reflection of existing ambient light OR energy, or
it blocks the transmission of light OR energy through itself from an
energy source.

Things exist that expand when exposed to energy. Some of them even at
energy levels that are safe for living things. If a thin flexible layer
of such a material were placed on top of a transmisive type of LCD, and
a safe source of correct energy were placed below, the job is done.

The main question I have,,,
Is it possible to read an almost flat vibrating surface reliably?
I see this as a requirement for using off the shelf parts to begin a
piezo version prototype reader now/today. The piezo also seems the only
likely candidate for installation in a semi-regular keyboard key EVER.
A vibrating surface can rest half or more of each vibration extending
usable battery life.

A piezo reader need only be as big as something to slip over a finger or

two with a mini joystick and a few buttons on it for the thumb to
operate. Even with a 9 volt battery only a couple of ounces.
Imagine sighted people learning to read braille so they can legally read

or surf while driving! A palm or CE and a reader the same size.

Both the piezo and LCD technologies described can be engineered to work
optically and tactically at the same time! You could show sighted people
what
you are reading in braille OR ascii.
You could ALSO show them something different or nothing.

Prototyping hardware designed to display serial data from a PC port on a
LCD
is available for about $70 US. The LCD driver chips without the serial
PC interface
are just a few dollars.
A few piezo elements (buzzers) to make a test grid a few
characters wide another $50-100.
Additional electronic components needed to glue everything together
$20-40.
$100-500 is not bad for a frist of a kind prototype!

---------------------------------------------------

Now to change gears a bit and discuss tactile graphical output devices.

Is this worth the cost and effort to produce?
Are there any that do shades of grey or color (by using different
heights) or are they all "black & white".
What brings this to mind is something I am familiar with from computer
ray tracing,
the "height field". A "height field" in ray tracing is a digitized
picture divided into its
grid of pixels and the color of each pixel is translated into a hieght.
A ray tracer is software that shows what such a field would LOOK like to

a sighted person in 3 dimensions given a camera position and source of
illumination. A 3D pin grid is given as an analogy/example in the ray
tracing documentation of a 'height field'. I use POV a free raytracer.

If you ever have an opportunity to try such a device, I suggest running
the original picture through various picture manipulation algorithms to
see which flavor you like best. GIMP and/or xview come with most linux
distros. A lot of the detail in a picture can be modified by changing
the brightness, contrast, sharpness, smoothness ,etc.
Colors can be swapped and or removed/enhanced digitally.
Different enhancements bring out different details in a picture.
I often have to 'mangle' pictures beyond any visible recognition to
produce
smooth height fields for ray tracing. But they can convey the original
data heightwise!
The GIMP can do 3D outline,  blended, carved,  chalk,  chip away,
chrome,
comic book,  crystal,  frosty, gradient bevel,  glossy,  neon,
starburst,
particle trace, textured, flatland, land manipulations on any 'graphic'
image.
It seems like it might work but would take some
fiddleing with each picture to exploit the available data.
Linux DEFINATELY has a lot of command line image manipulation tools
available.
Movies can be made with constantly varying levels of manipulation. on a
single picture.

Please don't write them off because a sighted person tells you that it
feels like what it looks like. Everyone thinks based on personal
experience. Play with the 'focus' knobs via software image manipulation!

None of this would have ever occured to me if I had never used a ray
tracer to
take a picture of a nebula and turn it into a landscape.
I have a couple mpeg examples for the sighted at dynolink.org

For GUI OS usage this may become a workable compromise.
A tactile graphical output device need not be very large.
A trick of visual display terminals is the virtual screen larger than
the viewable screen. When the cursor or pointer reaches the edge of the
visible display area the display is shifted in the opposite direction to

reveal the data that was off screen in the direction of pointer
movement. The graphical LCD display driver chips mentioned above already

do this virtual screen thing. Detecting fingers at the edge of the
screen or multiplying the movements are trivial to impliment and could
save on hardware cost.  The overall height range of a graphical display
can be
expanded through software. This would have the side effect of adding a
little
'tilt' feeling to the display under your fingers. It can be learned!

If each tactile element can be made to exhibit many discernable
patterns,
either many different heights, or 'feel' like many different heights, a
physical
approxamation of greyscale, and or color can be rendered along with many

different 'textures'.  Having multi bit resolution on each pixel
(different heights)
would be best for smooth scrolling of text too. In a 'graphic'
environment, each dot
can be made larger than any single element by combining adjascent dots
in groups.
This could be done easily with plain text LCD displays, but it's not
already built in.
This can be accomplished to some extent within the 5 by 7 grid of the
standard ascii LCD character.

If a 'graphic' device were in use, it seems like a matter of changing
the display FONT
to braille,  and using existing window manager software.  A touch
sensitive tablet would
allow pointer usage. Just tap your fingers to simulate mouse buttons.
This is how a
touchpad on a laptop works under linux. It is cool, tap 1 finger for
button 1, 2 for button 2,
and tap all 3 fingers for button 3. FVWM2 a window manager available
with slackware
comes with instructions to modify the placement , shape and size of the
various GUI buttons
that 'decorate' each window in a GUI.
FVWM2 uses human readable ascii text files for configuration.
Every menu item in the GUI is in the text file for user modification.
FVWM2 us not a resource hog so the X penalty is lower with it as apposed
to Gnome or KDE.
twm is even leaner on resources!!!
I mention the resource issue because one of the Linux benifits is
efficient usage of available resources.

Well, that's it, my latest ideas.
_I_ call it the Finegan method.
I'm a Finegan so everything I think I improve upon uses the Finegan
method to do so.

OOPS! This just in..........(I seem to be on a roll lately)
I have just thought of 2 more ways to make a SIMPLE electro-mechanical
scrolling braille display
with only 6-8 moving parts each. One of the ideas utilizes solenoid
tecnology, and the other one involves small stepper motors.
Lucky me, I was already typing this and did not forget to record the
ideas!

OK I'll attempt to describe the 8 solenoid scrolling braille display.
I'll discribe something GIANT but think smaller.
Take a box about the size of a pack of cigs on it's side and have it
represent 1 braille character space.
Cut 2 slots in the box to represent each column of a braille character.
First seperate a braille character into 6 plus 2 spare segments, so each
column has 4 dot positions.
I'll describe 1 segment, a single row,column position, multiply by 8 for
the whole thing.
The same single moving part can function in either direction so flipping
the display or scrolling backwards through text is possible too.
Take a bi-directional solenoid 'with a center off position' and modfy it
or attach it to a bar that sits below the slot in the box.  The bar can
move left and right just over the width of the slot.  On the bar wider
than the slot are a 'dot' on each side. normally they are hidden when no
power is applied to the spot. When power is applied to the dot position,
the solenoid moves the dot on the correct end of the bar, in the correct
direction under your finger. When power is removed, the bar is made to
drop away from the surface
so your finger does not feel it on the return path.  The bar would
require one cycle per character dot .
This type of mechanical movement is not uncommon in many tools. A
prototype of this design could be quite inexpensive. Plastic would be a
suitable building material. You would need to press in a bit to feel the
dots as they would not lift above the surface.

Now I'll try to describe the #2 'motorized mechanical' scrolling braille
tty design idea with only 8 moving parts! The dots project above the
surface of the reader box.
Use the same box mentioned above, but replace each bar with a wheel with
bumps wider than the slots. A square shaped wheel with rounded corners
would work fine. Each rounded corner could represent a dot when rotated
45 degrees, and when flat to the slot not represent a dot. the wheels
can easily be rotated in both directions. multiply by 8 and there is a
scrolling braille display.
If you didn't round the square the dots would just be pointy.
Put a rubber membrane over the top to hide the slots and improve finger
mileage!!!
Stepper controllers are not dirt cheap, but for the needed power,
definately not over few dollars per stepper in small quantities.
Tiny steppers are a few bucks each! An inexpensive off the shelf 8 or 16
bit microcontroller, with a couple hundred lines of code can interface
the works to a serial OR parallel port. A complete microprocessor
development 'environment' can be had for between $100-200.

You can simulate a scrolling display with a static display by cutting a
character sized hole in a piece of cardboard and have someone move the
braille tty below the cardboard past the hole while you hold your hand
stationary with your finger(s) over the hole.  Or pass a braille book
under the hole.
All you feel is the dots passing. Put different flexible things over the
hole like cloth,plasticbag,rubber,etc...

Well, if you have to listen, isn't that wierd???

Now I make a request for comments/suggestions/ and even flames if you
think it will help in some way. Have I neglected any glaringly obvious
details???
Are they silly ideas as a result?

Thank you for your time.

Jan P. Finegan
jf.blinux@dynolink.org
AKA; Mr Whoopie, The 3D BlackBoard guy. (from old Tennesee Tuxedo
cartoon fame)
P.S.  My brain and/or any unique intellectual property I might posess
are for rent at reasonable rates.

^ permalink raw reply	[flat|nested] 5+ messages in thread