Paper: an upright printer/scanner that uses a continuous roll of paper

[Read the post]

If he could make that work with perforated green-bar paper, then he’d really have something. Also, have it make a loud buzzy noise, just for old-times’ sake.


Can it print in dot-matrix? Just for old times’ sake.

That also reminds me of a scriptwriter I met who’d written for shows like Soap. He said the feeling of having to work to the end of the page always intimidated him so he bought a roll of continuous paper and fed that through his typewriter.


Dot matrix has held on surprisingly well for its age and fairly pitiful resolution. You can get bright, shiny, new models right off the shelf. Some of them have even been updated for USB support.

Aside from sounding cool and producing the definitive typeface of the retrofuture, consumables costs are low and the dot matrix print head is nigh-indestructible. Probably more importantly to its ongoing survival, if you are using NCR-paper forms to push paper in triplicate your options are either dot matrix or typewriters.


I have that for toilet paper. It’s disgusting!


I wonder how it would get paper off a roll and have it come out flat. I also wonder where its gut are, as it doesn’t seem like there would be room for both a moving scan apparatus and a moving inkjet printhead - it simply can’t be CMYK laser at that size.

In short, it looks like a neat concept, but also like it badly misrepresents what the final product would have to look like.


Personally, I was relieved when printers no longer required tractor-feed fan-fold paper. This seems a step backward, I don’t care what color it is.

1 Like

Quite a lot of laser printers can print banners.
The main problem with continuous feed is that unless it is tractor feed like the old printers we knew and loved hated, eventually it gets out of true. Paper is affected by heat and moisture and deforms slightly, and a rubber roller is never 100% true. So eventually it will jam unless there is some clever mechanism to release the roller and realign the paper at convenient intervals - like pages.

1 Like

There’s design, and there’s engineering design. Yes, this has been designed in the aesthetic sense. It has not been designed in the engineering sense. It would not look like that if it was designed to actually work.


He’s a “Big Idea” guy. He’ll leave it to Marketing to work out the kinks.


There are a lot of Inkjet roll printers that keep tracking on 30M rolls quite easily. The key part is having a sturdy feed mechanism. Most laser bypass trays have a lot of play in them.

1 Like

Dot matrix printers are still massive in a lot of industries. We sell a lot to companies in the Airline/cargo business, often used oddball models. My local builders merchant still uses them with duplicate paper.

1 Like

This would certainly not be the first time, by a long shot, that a ‘design’ wins acclaim without being remotely feasible; but I see two possible solutions:

If there is room for at least the moving print head, you could simply add a small imaging sensor into the print head package: in order for moving print heads to work at all(deposit ink in the right place, not have lines overlapping when they should be laid out next to each other, that sort of thing) printers with moving print heads already keep track of both the print head position(using stepper motors to drive the print head in earlier models, using optointerrupters and feedback systems in newer ones) and the paper feed. This position data would allow you to ‘stitch’ together the pictures produced by the sensor on the print head; and since it would only need to take a picture of a very small chunk of the document at any one time, you could get pretty high resolution.

The other option is to use one of the full-width scanner sensors, common in cheap flatbeds; but use the printer’s paper feed to pull the document past the sensor rather than having the document lie flat and moving the sensor across it.
(I apologize for the quality of the below image, I’m currently without non-sucky macro gear and had to resort to a cheap flatbed to scan the scanner, which is meta but not very good for quality).

This is an example(out of a ‘canoscan lide’ that never received Win7 drivers so that a hardware-identical model with a slightly different color plastics kit and a different USB ID could receive Win7 drivers and be sold as a replacement; assholes.) is one of the linear imaging mechanisms.

On the far left, the white line is the top side of an RGB LED cluster. Moving across the board the grey stripe in the center is a series of linear silicon image sensors laid end-to-end and covering the entire width of the board. In use, the RGB LEDs cycle rapidly, allowing the monochrome sensor strip to capture an R, G, and B image of each strip before moving on to the next one; with the color image produced by combining the R, G, and B exposures for each strip and then all the strips into an entire page view.

The device this came out of used a conventional flatbed design, and moved the sensor; but it would work equally well if the paper were moved past a static sensor, so long as the paper feed occurred at known speed.


As a mechanical designer, this “fantasy” stuff drives me batshit. I remember an award winning bike design based on room temp superconductor. There’s a similarly acclaimed and silly idea to put a bike and pedestrian bridge across the Hudson River from Jersey City to Lower Manhattan. Cost/benefit and other “externalities” like viewscape and river traffic are not considered in the least. The engineering is dubious, a cage frame structure spanning >500 ft between supports, according to this rendering.


Architecture, as a field, would probably be better off if it formally recognized the existing split between people who design buildings and people who design large-scale sculptures aimed at commenting on urbanism and related issues.


Are you counting Line printers as typewriters?


Much easier to do if the feed is vertical.

Even better if it uses thermal paper.

I still keep a dot-matrix printer in my office in case I have to produce a ditto master. Spirit duplicators use no power (other than materials production) and are hardy, so remain popular in the 3rd world.

1 Like

I think you could squeeze a thin flatbed mechanism in there, it looks like the scanner is in a separate lid part (I couldn’t watch the video), and those LIDE scanners have almost nothing in them. Sheet fed scanners are sub optimal for everyday scanning as you can’t do anything other than a plain sheet of paper.

As for the print part, I’d use a full width printhead, such as the Memjet or the HP Pagewide tech. It’s not caught on that much for general use, it’s too expensive for the typical home printer. They have some business models but they were practically giving them away a while back, so I don’t think adoption has been as good as HP had hoped.

With some improvements to the paper feed mechanism, it’s not a totally outlandish concept. I’m still not sure that the roll paper functionality is going to be that useful- Reams of paper are available everywhere. You’d need a 150m roll to equal one standard A4 ream, which is about 3x the length of a normal roll of 80GSM paper.

“I see two possible solutions”

Both good descriptions of scanners in the Real World, but given the representation of the scanner as a 1/2" (less?) thick, flatbed assembly outside the path of the paper off the roll, neither is consistent with the device as pictured.

If this thing were ever actually produced, you know there would be a stupid list of steps requiring the user to remove the roll paper from the feed mechanism, then load the document to be scanned in its place to be pulled past the print/scan head. And since it’s virtually impossible to load a sheet of paper perfectly straight, even in a flatbed scanner, you know the source doc will feed crooked, probably wrinkling and/or tearing one edge and creating a crooked duplicate.

1 Like