Tactile graphics on MacOS with a braille embosser

This doesn't address your problem directly, but I'm posting it here as a reference.

After months waiting, a few weeks ago the ZyFuse swell embosser that I ordered finally arrived. This is not a Braille embosser, although it can emboss Braille just fine if I print the dots on paper, but its main function is to emboss very detailed tactile graphics. The way it works is I just print whatever I want on proprietary paper sheets using a regular printer, a laser printer in my case, then I feed the sheet to the embosser, and whatever is darkened in the printed image is swollen by the embosser thus creating the tactile graphics, with darker parts getting more pronounced swelling. Drawing with a marker or even a roller ball pen also works, and even very detailed fine-grained textures that feel like sandpaper can be embossed. The embosser itself does not require and does not actually have any way to connect it to a computer, which also means that it doesn't need drivers or specific operating systems to operate, and that was definitely a huge selling point for me.

One problem that I found, however, is that Safari does not handle printing SVG very well, so images end up getting cropped, sometimes severely. However I haven't spent much time figuring out a workflow yet so I might end up finding a solution or implementing one myself at some point in the future.

Wait, wait, wait. You purchased a machine that is not IOT compatible?
Say it isn't so?! 😱

Wait, wait, wait. You purchased a machine that is not IOT compatible?

Yes, it only has an on-off switch and a knob to control how much heat is applied to the sheets and thus how pronounced the swelling ends up being, nothing else. It's also made of hard metal with sharp corners, and feels like hardware from the 80s, or a battle tank, and the only port on it is for the power cord. There's a paper tray on the front, made of the same metal as the rest of the case, which can either take A3 sheets in a vertical orientation or A4 sheets in a horizontal orientation, and the paper comes out through a slit on the opposite side, falling down on whatever happens to be behind. It's a rather crude machine, but does it's job well.

I am wondering, as a basic means of verification if we could create an app that lets you 'feel' the outline of a 2D object using haptics, IE, haptics when you're touching within the boundaries of the object, and no haptics when moved beyond.

I have considered doing so before ordering my embosser, but quickly realized that it would not work since even with haptic feedback it becomes quite hard to tell the direction of the stroke, which at least I find important. I can still write something for iOS, or maybe even try to find a way to repurpose the trackpad on macOS for people to experiment with that yourselves, but I don't think it will be very useful.

When I output SVGs to my ViewPlus Delta, I just treat it like a printer and send the files directly to it by printing from Safari, Chrome, or Preview. Have you tried that using the latest MacOS Index printer drivers?

I also have a lot of information about outputting SVG and rasterizing images for tactile format output on my BlindSVG site: BlindSVG Output and Embossing

In the 80s, there was some sort of... bake and puff paper that you ran through the copy machines and worked just like what João Santos describes, and it may be the same stuff. It sounds like you can get a crisper resolution now because the braille came out weak back then. This was a gadget a blind school might have, and probably not an individual, in most cases.
Still too rich for my budget, and what I'm working on.
Later in college, I took a ribbon-style, dot matrix printer and rigged it up to scratch the paper, and it was enough to study the shapes of fonts, one or two letters at a time, scaled up and in bold. That's about all I did with it before the pins started failing. Might have worked for pictures if I had experimented more with delicate paper or something.
Now, I just find someone who likes tracing things, like adult coloring books, and give them a mirror printout with one of those raised line frames.
I use the heck out of Seeing AI's explore function for pictures. Not quite haptic, and very vague about what it recognizes, but still useful. You have to do a lot of guesswork on where the edges of the image are, and go back to ask a lot of questions.
It would be nice to look at the layout on a quick printout when I'm doing a collage, instead of trying to rely only on the math, and asking AI about the white space or borders.

Yeah, we talked about this briefly on reddit, where I post as Fridux. I did skim over your website, and then replied back to you some time later, but the cropping problem that I'm facing, which sometimes is so bad that only a tiny portion of the SVG is printed out, happens on currentSafari with Apple's own AirPrint and USB drivers for my Brother printer, which work well in every other case. Also just telling the system to print to a PDF instead results in parts of the image getting cropped as well, so I have little reason to doubt the quality of the drivers or the printer itself.

While writing an earlier comment to this thread, the possibility that Safari could just be covering the edges of the sheet with some kind of white margin crossed my mind, since the viewport of the SVG that I was drawing had just enough room for the whole logo. This might be causing Safari to render it in one of the corners of the sheet resulting in parts of the image getting cropped by the aforementioned white margins. I have a few housekeeping tasks to finish now, but will investigate this possibility once I'm done and report on the results.

Man, this is in my purchase list since like 2 years for literally the exact reasons you mentioned! Gonna buy PIAF from Humanware soon. Thank you for a short email exchange with a guy at blind SVG.
My plan was to use it with SVGs too, thanks for letting me know about problems in Safari.
Also if somebody can try freeform with this thing... :) Will unicode braille do the trick for labeling, with zoom if needed?
Edit: I swear I've just finished reading the whole thread and it looks like this guy is here, greetings!

Back then, those dot matrix printers had an adjustment in the works that moved the printing head closer to the paper. So I moved it as close as it would go. There's a metal platen, I can't believe I remember that term, on the other side of the paper that is a stop for the pressure of the pins. I covered that with electrical tape for a slight give, and make the paper even closer. The ribbon was in a cartridge and got shredded, but as long as the cartridge was in place, the printer still printed. You had to look at the print to tell if it needed to have the ribbon replaced back then.
On the software side, which was Word Perfect, I increased the font size to huge and bold. I'm not sure I needed to make it bold.
Had I tried some of that film for raised line drawings it might have been interesting, but I didn't have any at the time. I'm sure there was other types of paper that would have scarred more easily than regular printer paper.

It's a little too complicated to write out, but offset lithographic printing might be used in the second thing you brought up with the tactile paint. I'm trying to remember the process of getting the print and pictures onto the aluminum sheets, which used an ink that held oil and repelled water. I seem to remember that being slightly tactile, but I was a tiny, little kid when my family members worked at a newspaper. We used the aluminum sheets, with the mirrored print and pictures, as shingles on a back room of the house.
Ditto machines kind of worked in a similar way.
Anyway, all that rambling is to say, maybe some of the specialty inks could provide a path to having a thicker substance stick or not stick onto printed material in a multi step process. Maybe even some of those cloth printing inks.

I spent some time trying to figure this out, and even wasted some expensive paper sheets in the process, but I guess that this is one of those times where sighted assistance will be fundamental for me to understand what's happening.

I doubled the size of the viewport of my SVG, moved everything to the center, and then tried to print out from Safari, which resulted in a tiny portion of the rasterized image actually getting printed to paper. Then I did the same but instead of printing to paper I just opened Preview with the print job, and only part of the logo was visible, but VoiceOver did recognize part of the text in that logo, meaning that more content is visible in Preview than gets printed out, and this time printing from Preview yielded the same results as directly from Safari, with just a tiny portion of the rasterized image appearing at the center of the paper.

I also confirmed something that I had already noticed before, which is that the swell paper comes out of my laser printer with a slightly more bumpy texture, possibly due to being in contact with the drum inside the printer which is heated by the laser. While as I understand it the laser never shines directly on paper, it does shine on the drum to generate an electrostatic charge on it so that once the paper comes in, the ink in the toner gets pulled onto the paper, and I think that this heat is also causing small reactions on the swell paper that are changing its texture. New paper sheets are quite smooth to the touch, with the back side feeling like magazine paper and likely producing blurry specular reflection visual effects, while the front side is slightly less smooth but still smoother than regular paper, and when it comes out of my laser printer, the texture feels more like a rubbery and bumpy surface of an orange, but much denser and less pronounced, so people using this kind of hardware for production may want to opt for inkjet printers instead.

That paper you described sounds like photo paper, which one would use if they were printing up digital photos from something like a memory card. Not necessarily saying that that is what you're using, just that's what it sounds like from your description.

Maybe it's a misunderstanding on my end, but why the heck humanware and other keep calling these devices embossers? They are nothing but that, they are just heaters on steroid with a paper that loves being heated? :) They are not embossers yet hw put PIAF in the same category as actual braille embosser, a marketting laziness?
PS: after VOSH and another dev like thread about macos a11y infra problem, this is in my top teer list of best applevis threads, period. A friend of mine has a 3d printer, completely blind. And he's crazy happy with it.

I was thinking that laser printer are better than inkget things? But I have 0 experience nor the device to test it with (yet) so...

There are cheaper tactile graphics tablets than the Graphiti at £18000. The Dot Pad X is listed at £6495 at the RNIB Store, although you can likely find it cheaper since that price is $8840.67, and in the US the Dot Pad X sells for $5900. The Dot Pad X does not support variable dot heights like the Graphiti does, but it has smaller dots and higher resolution.

It can rastorize an image we feed it, like at double tap they talked about how when focusing on the clock icon dotpad x would show the actual icon of the clock app, which is, surprise surprise, a clock. But I'm only telling what I heard about others reporting it so maybe it's not true or not as I understood and put it.

Just a few more thoughts to throw in the pond.
I was thinking about it, and I wondered to the Google search field, "Is there something that would react with inkjet pigments to cause a texture on paper?" Asking that because the black in inkjet ink is a pigment mainly that may or may not be mixed with a little of the CMY color dyes.
There were interesting results that seemed to be about sighted people adding textures to pictures.
I felt of an extra print of a photo collage I had laying around to find out if I could detect any ink. I couldn't, and I should have been able to at least detect the difference in white boarders around the photos if it could be detected. That is on quality photo paper though. If it were printed on a regular piece of paper, I am told there would be so much ink or dye that it would make a slow-drying mess... Kind of reminds me of water colors which are a bit tactile. Perhaps if a line drawing or something like that were run through the printer several times it would build up, or even expand the paper like water colors.

Ya, I'm sorry that might not have made too much sense. I'm juggling a bunch of thoughts about different printing processes and art stuff I've experienced over the last fifty years that might be combined, and new thoughts pop up as I write.
The photo paper absorbs ink and should be fairly nonabsorbent after the ink is dried, but not where the unprinted white space is. Some other substance might also be absorbed by the paper, so the printed part should be resistant to the new substance. I didn't quite get there in what I wrote because I started thinking about water colors.
Silk screening and photo emulsion might be something else you could look into. That can be very tactile, and you can print on projector-type transparency film for which ever kind of printer you have. It's similar with lithography, but it uses an aluminum plate.
I hope that kind of ties my posts and wandering thoughts together. LOL

The Dot Pad X has equidistant pins. When displaying Braille, it will leave one horizontal column blank between each cell for proper Braille spacing, and usually one or more vertical rows blank in between cells. I'm pretty sure the Dot Pad X is 60 pins wide by 40 pins tall, and that 300 is not the number of pins, but the number of eight dot cells if you do not account for Braille spacing. With proper Braille spacing from the horizontal space between cells, you get 20 cells wide by 10 cells tall, and there ends up being less than 10 lines when vertical space between lines is enabled. I think the Graphiti might be bigger, but it definitely has lower resolution and so does not display Braille as well. The Dot Pad X is 2400 pins, while the Monarch (which has the exact same pin spacing and way as displaying Braille) has 3840 pins (96 by 40 pins, 32 by 10 cells).

You can display Braille or graphics on the Dot Pad X in one of two ways. You can connect it to a screen reader, or use one of the applications created using the Dot Pad SDK. I do not remember what applications there were, but I'm pretty sure there was one for reading books and documents where you can customize the Braille spacing, and one for viewing images. Also, I'm pretty sure the only screen reader that supports displaying graphics on the Dot Pad is VoiceOver on iOS and MacOS; the other supported screen readers will only display multi-line Braille. Windows itself will not interface with the Dot Pad at all; it is always either your screen reader or the SDK.

To label things with swell paper, I have seen well produced documents in highschool but this is probably because they were using dbt or TactileView, all windows. But I understand that it's not as simple as printing then heating unicode braille or those braille fonts right?

Just type the Unicode Braille on a comment and I'll test that. I'll even test multiple font sizes on the same sheet to figure out what is closest to regular Braille printed out to paper. I'd imagine that the font size would have to be huge compared to sight-reading since Braille is quite large compared to print, but have absolutely no idea in numerical terms.

My apologies for your tts @everyone :)
⠠⠓⠑⠗⠑ ⠊⠎ ⠁ ⠞⠑⠎⠞ ⠊⠝ ⠠⠠⠥⠑⠃⠲
As for the font size, from the random documents from APH and others it seems that 24 or slightly above is a good size. I think there's also one font recommended more than others though I can't remember the exact name now.
Thanks so much!
A note on the fonts, I could be completely wrong but from my reading there's two category I think, one to show unicode braille best, and others that take normal text in latin characters or something and visually render it as braille... But I have 0 expertise or way to confirm this so trust 0 of what you've just read, information / correction is most welcome.

It's already 6AM here, I'm implementing an audio routing solution to actually record this because for some reason Apple removed the ability to create all kinds of aggregate devices from Audio MIDI Setup on macOS 26 so I can no longer just use Blackhole as a knee jerk solution to mix microphone and system audio in an audio recording, and I wanted to make an audio recording of this to try to convey the textures embossed on the swell paper through sound. Unfortunately I'm way too sleepy to do anything at this point so I'll go to bed and resume working on the audio routing later, which I will also make available as a Swift library with a sample executable target for others to use, learn from, or contribute to.

I pretty much only use Safari now to render and output tactile or visible SVG images since it handles the SVG elements and text so well, especially with braille. Remember, the viewBox only defines your overall canvas space inside the graphic. Anything placed outside of the viewBox coordinates will get cropped and will not render. You manage the overall output file scale with the width and height attributes on the top SVG element tag.

Conversion to raster is also a topic I wrote out in the BlindSVG Output/Printing](https://blindsvg.com/pages/output/) page. If I need to rasterize an SVG for specific printer output, I use headless Inkscape in Terminal to convert my SVG into the required format plus set the export DPI there.

Also, when embossing, you need to get rid of the margins created by the paper type/handling in the system print dialog. I have a custom paper type set with 0 margins for Letter sized paper, and that is my print preset for my Delta embosser. Otherwise if you have something right near the edge of the paper, it'll get cut into by the margins.

If you want to center a graphic in SVG, you don't use the viewBox for that. You should group all the shapes you want to center and use a transform attribute with the translate function to move the image to the center of your paper size. The translate x value would be set to half of the paper width value minus half of the total size of the image, and the same goes for the y value.

So if you have an 8.5x11" sheet of letter paper in portrait, and think of the SVG units in 100DPI, that means 8.5 inches is 850 units. Half of 850 is 425, so that's the horizontal center of the paper. If your logo is 100 units wide, you'd subtract 50 from 425 to get the x value you need to put in the translate function, so that would look like:

transform="translate(375, 500)"

Assuming your logo is created in the top-left corner of the canvas, of course. :)

When it comes to defining the coordinates of the SVG, I think that I'm doing everything right. Originally the SVG had a viewport of 0 0 origin and 1024 768 size, with everything drawn inside. From what I gather, saving the image to a PNG works just fine, but printing it to the printer or to a PDF results in cropping. Resizing the viewport to 0 0 2048 1536, which is twice the sides or 4 times the area, and then centering the logo, also results in cropping, but in this case it may be user error since I did not check whether exporting the page as PNG worked.

Thanks for the SVG suggestions though! I'm not very familiar with the standard and just wanted to design a logo for one of my registered trademarks so I wasn't expecting to spend much time with it and didn't consider digging more deeply.