The Space Shuttle Translite Project

The backglass of the Space Shuttle Pinball Machine is a valuable part that is difficult to find.  Prices on units on ebay range from $150 to $250 depending on condition.  As far as I know, reproductions have not been made.  Although I have a spare unit, I decided to try making a translite for the technical challenges.  Translites are the way later pinball machines (post ~1990) decorate their backbox.  Instead of on glass, the artwork is printed on a clear plastic sheet, and then sandwiched to a pane of glass. 

Clearly, there are two major steps: 1) obtain the artwork, and 2) printing.

Obtaining the artwork

The scanning process.  I used an HP4600 scanner, which is specially made to
sit flat on an object for scanning.

I used an HP4600 scanner since it is possible for it to be removed from its frame and be placed directly on the object to be scanned.  This allows it to scan a large object in sections.  I started the scan at various resolutions, but hit a snag because I was obtaining stripes in the scanned image of the backglass.  I realized it was some kind of aliasing artifact of the dots in the silk-screen since they did not appear in the scans of a photo that was placed on the backglass (see below).  I finally found a number of settings on the HP scanning software that eliminated this problem (descreen: ON, highlight setting: 70, shadow setting: 5).

Initial results of scan process showed faint horizontal stripes in the lighter areas of
the backglass (left).  These stripes do not appear on scans of photos (right).

I experimented with various scan resolutions.  It turns out that a scanner will not be able to get a perfectly sharp image due to the thickness of the glass itself.  The reason is that the optical design of scanners presumes the subject is right up against its glass window.  The added distance of the backglass' glass causes a slight defocusing effect.  As a result, I found that scanning at 300 dpi was sufficient.  No gain could be seen when scanning at 300 dpi and doubling the image size vs scanning in native 600 dpi mode.

Full resolution excerpt of one of the scans.  The resolution is 300 dpi.

As can be seen from the above (full resolution) excerpt of the scan, there is some slight blurring in the horizontal direction.  Thus horizontal edges are sharp, but vertical ones have a blur of a few pixels.  I think one explanation is as follows: the scan bar is horizontal, and the scan motion is vertical with respect to the above image.  I think that the internal reflections of the glass causes a pixel of the image sensor to see a small amount of light from an adjacent emitter.  Note that this effect does not occur with dust particles that are on the surface of the glass.  These appear as sharp single-pixel dots.  One example is above the Shuttle windows.

Of course this problem can be addressed by photographing the backglass, but that leads to the possibilities of distortions and mismatch in lighting and color from section to section.

Thumb nail of the eight scan files of the front of the backglass.
Each file is about 10 MBytes in size.

Once they are all scanned it is time to use Photoshop.  I started with a new 300 dpi project with the dimensions slightly larger than the backglass.  One tricky aspect is that the backglass has a lot of mirroring.  These areas scan poorly, and will make the manufacturing more challenging.  Also there are of course clear areas such as the score glass areas.  For this reason, transparent areas need to be defined on the artwork.  Photoshop makes it very easy to later change the color of the transparent areas, and assign a color later if I decide to not reproduce the mirroring.

Photo of the back of the Space Shuttle backglass to show the opaque or mask areas.
Gray is opaque, white is diffuse.

The overall concept of the reproduction translite is to use several layers of artwork.  Starting with the layer closest to the player:

The 'front' layer.  This is the layer that will be printed from the scanned image.

The 'mirror' layer.  This will provide the mirrored border and the mirroring around the words "Space Shuttle".

The 'white' layer.  This will provide the diffuse white layer of the Moon, Shuttle and Earth because the previous layers are printed on clear film, or are only used in strips (mirror layer).

The 'mask' layer.  This layer will obscure the light from specific areas and provide highlighting for others.  It looks like the image above.

The mirror layer may be omitted if it proves too challenging to implement.  However, it is possible that it can be done by using mirrored film such as what is used in one-way mirrors.  I obtained a roll of the material and will be experimenting with it.

The eight front side images stitched together and defects removed.  The checker box indicates the transparent areas.
I think this is a very beautiful backglass.


The mask layer that is at the back of the backglass.  Note the scattered dots to
match the stars, and the gradual fade on the exhaust plume and volcanic
eruptions on the earth.

Simulated backlit effect.  The mask layer is put in front of the front layer and
then set to 75% transparency.  Note the brighter areas where the
mask is transparent.  The words at the bottom of the glass are only
visible when backlit.

Initially, I was going to scan the back (mask) layer also.  I then realized that it would be more precise and easier to derive the mask from the front layer.  I copied the words "Space Shuttle" from the front layer, and pasted it right back in.  This caused the image to align perfectly, but in a new layer.  After a lot of "Select By Color" and "Fill" operations, I was able to turn the red-yellow lettering to transparent on a black field.  For the Moon, Shuttle, background stars, and Earth, I simply drew those large areas in by hand.  This allowed me to precisely control the gradients such as by the exhaust plume and the volcanic eruptions on the Earth.

The words at the bottom of the mask were typed in with a font that looks similar, merged into the mask layer, and then turned to transparent by "Select By Color" and "Cut".

Printing the artwork

As a proof of concept, I decided to make a miniature demo unit composed of various layers.

For the main artwork, I printed the front and mask layers onto clear transparencies using a color laser printer.  Since a color laser does not print white, I can define transparent areas by assigning white to those sections. 

The backglass has a mirrored border all the way around, and the words "Space Shuttle" are framed in mirror.  To mimic this effect, I cut out a band of material from mirror film.  This material, when applied to ordinary glass, converts it to a mirror look.  I decided to skip mirroring the area around the words as I would have to do some very intricate cutting.  Only the outer band is mirrored.

For the white color, I found at the local Michael's craft store some translucent white paper that has a very low amount of blockage to backlit light.  Regular office white paper is very opaque and has a high attenuation factor to light from the back.

These four layers were sandwiched together to form the miniature demo unit.

The front and mask layers printed on conventional transparencies as a test. 
A translucent white sheet of material is sandwiched in between to provide
the white color, and mirror film is used for the mirroring.

The four layers are individually taped to a pane of glass from a picture frame.  This allows me to hang the demo unit as pinball art in my game room.

Backlit version of the sandwiched printed transparencies.  The words
at the bottom can be seen.

The framed demo mini translite next to the full-sized one.  Note the good color match.

Printing the artwork II

After I assembled the above demo unit consisting of the four layers, I decided to build up a simpler unit without the mirroring.  Although the above provides the most accurate reproduction, I think it would be a lot of effort to register all those layers and tape up the mirror film.  It would also be difficult to roll it up for mailing and storing.  For these reasons, I decided to make a second demo unit with a simpler concept.  This consists of simply printing directly on the two sides of the translucent paper.

The front and mask layers printed directly onto translucent paper. 
Easier to roll up and store away.

This result looks good in the above image due to the bright flash.  With the naked eye and ambient light, the version printed on the translucent paper looks dull and lifeless.  Another version printed on white office paper looks bright and beautiful.  The lesson here is that the substrate on which the translite is printed is very important.  It should be bright white in reflection, and be very low attenuation in the backlit mode.

Printing the artwork III

With the involvement of two RGP members, I ordered a full-sized test print of the artwork.

Full-sized printing of the translite (left) next to the original unit (right).

The above photo of the printed translite and the backglass shows that the color match of the red and yellow is very close, but the blue is a shade more purple than the original.  Also, there is a slight haze of the translite when the ambient light is incident at certain angles.  This is especially visible in the top of the translite.  This is shown better in the photos below.

Another angle of the two next to each other.


Portion of the original backglass for reference.  Lit by the overhead fluorecent lights.


The same portion of the printed translite.  Note the haze due to the overhead light.

Using some water, an illustration of the effect of applying a clear gloss
coating onto the translite.
The haze is reduced.

In general, the printed translite looks very good, especially when backlit.  However, as mentioned above, there is a slight haze on the front when the light is reflected at certain angles. I decided to see if it was possible to remedy this by clearcoating it.  Tests showed that the printing is impervious to water and alcohol.  As a simulation of the effect of the clearcoat, I poured some water over the printing, and the result is shown above.  It shows that the contrast can be enhanced by applying a layer of clear.  The black gets darker, and the colors slightly more saturated.

The words at the bottom are printed on transparent plastic, and then
fastened to the back of the translite.  Just like the original backglass,
the words are vaguely visible from the front, but can be seen when backlit.
(This is a backlit image).

The words that are backlit by status lamps in the backbox (Stop & Score, TILT, etc) are printed on transparent plastic, and then fastened to the back.  I found an adhesive method that is invisible from the front, but does not dry hard, which allows the translite to be rolled up for shipment.  It also allows non-destructive removal of the printing in case it needs to be changed or repositioned.  Another important requirement is that the printing has to be stuck tight to the back of the translite, or the words will blur.

Printing the artwork IV

Once the protoype was printed, I had four more done, and they are shown below.  Two went to Australia, and shipping took three weeks.

The production batch of four translites.

Comment from some of the recipients:

Printing the artwork V

I found another printer that could print the translite with a potentially better texture on the front surface.  I decided to try another test print.

The fifth test print.  The haze on the front is much reduced, and the blacks are darker.
Original backglass on left, Print V on the right.


Another comparison.  Print V on the top, Backglass on the bottom.

The result was very satisfactory.  The haze was reduced, and the translite looks much better lit from the front.  The dark colors are darker, and that increased the contrast of the image.  The contrast in backlit mode was as good as the previous print.

Side by side comparison.  Backglass on left.  Print V on right.


Three copies of Print V ready for shipment.

Comment from a recipient of Print V (5/3/2007):

It is most satisfying to me to have a small part in saving a Space Shuttle
machine from being parted out!
See below for a photo of his machine at the show.


The H.A.A.G. show was on Oct 30 2007.  The above recipient of the translite sent me this
picture of his restored machine at the show.

One recipient of the translite had some astronauts sign his:

Charlie Bolden (who later became NASA Administrator), Hank Hartsfield and Jon McBride.


Dana sent me this picture of his restored machine.
He also installed an overlay from me.


Adam sent me the above picture.  His comment below.

Picture from Robert and his restored machine.  I provided several
other items, and the entire machine looks beautiful.


Another translite installed.  This one from Maurizio.


Translite on Jody D's machine.

Comment above from a repeat customer.


Another machine with the translite (11/2014).

STS-125 Crew Visit

In February 2007, the Space Shuttle astronaut crew of the final Servicing Mission

to the Hubble Space Telescope visited our group for crew orientation and
training.  During a social event, I asked them to sign a copy of my translite.



The Crew of STS-125 signing my Space Shuttle Translight.  They were
delighted to see such beautiful artwork on a pinball machine.
It was a nice opportunity to talk pinball with the astronauts
instead of work-related topics.

In order: Scott "Scooter" Altman (commander), Greg "Ray Jay" Johnson (pilot),
Megan McArthur (RMS Operator), John Grunsfeld, Andrew Feustel,

Mike "Bueno" Good and Mike "Mass" Massimino.


The signed translite.  It should look great on my pinball machine.

Backglass Links I have gathered over the years

Backglass restoration article (flippers.com)

Backglass repro (Tuuka)

Backglass sealing (Bill G.)

Touching up Flash Backglass.

Backglass sealing (Clay)

Restoring a backglass with waterslide graphics (Alan Lewis)

Book about how to print playfield and translite artwork (Classic Arcades Grafix).  Discussion/review about them here.

Backglass/Translite Printers (arcadeoverlays.com).

Kinkos will print backlit posters.

Google search

Plastic trim to hold the translite:

Backglass lift channel

Side trim

Top Trim

Backglass restoration.

Project Log

October 28, 2006 - Project started on a rainy day in October.  On this day, backglass was scanned, and photoshop was used to create the front and mask layers.  For reference information, the thickness of the glass is 1/8" (about 0.131" with the paint).  It is 28.5" wide and 21" tall.

November 1, 2006 - Printed demo miniature unit.

November 2, 2006 - Printed second demo unit.

November 14, 2006 - Received full-sized printing of translite.

December 2, 2006 - Production of four more units.

February 6, 2007 - Print V is a success.

Main Pinball Page

Back Home

(c) 2006 Edward Cheung, all rights reserved.