This tactile display lets visually impaired users feel on-screen 3D shapes

Using a computer and modern software can be a chore to begin with for the visually impaired, but fundamentally visual tasks like 3D design are even harder. This Stanford team is working on a way to display 3D information, like in a CAD or modeling program, using a “2.5D” display made up of pins that be raised or lowered as sort of tactile pixels. Taxels!

The research project, a collaboration between graduate student Alexa Siu, Joshua Miele, and lab head Sean Follmer, is intended to explore avenues by which blind and visually impaired people can accomplish visual tasks without the aid of a sighted helper. It was presented this week at SIGACCESS.

tactile display2The device is essentially a 12×24 array of thin columns with rounded tops that can be individually told to rise anywhere from a fraction of an inch to several inches above the plane, taking the shape of 3D objects quick enough to amount to real time.

“It opens up the possibility of blind people being, not just consumers of the benefits of fabrication technology, but agents in it, creating our own tools from 3D modeling environments that we would want or need – and having some hope of doing it in a timely manner,” explained Miele, who is himself blind, in a Stanford news release.

Siu calls the device “2.5D,” since of course it can’t show the entire object floating in midair. But it’s an easy way for someone who can’t see the screen to understand the shape it’s displaying. The resolution is limited, sure, but that’s a shortcoming shared by all tactile displays — which it should be noted are extremely rare to begin with and often very expensive.

The field is moving forward, but too slowly for some, like this crew and the parents behind the BecDot, an inexpensive Braille display for kids. And other tactile displays are being pursued as possibilities for interactions in virtual environments.

Getting an intuitive understanding of a 3D object, whether one is designing or just viewing it, usually means rotating and shifting it — something that’s difficult to express in non-visual ways. But a real-time tactile display like this one can change the shape it’s showing quickly and smoothly, allowing more complex shapes like moving cross-sections to be expressed as well.

tac

Joshua Miele demonstrates the device.

The device is far from becoming a commercial project, though as you can see in the images (and the video below), it’s very much a working prototype and a fairly polished one at that. The team plans on reducing the size of the pins, which would of course increase the resolution of the display. Interestingly another grad student in the same lab is working on that very thing, albeit at rather an earlier stage.

The Shape Lab at Stanford is working on a number of projects along these lines — you can keep up with their work at the lab’s website.

Gadgets – TechCrunch

Feel the beep: This album is played entirely on a PC motherboard speaker

If you’re craving a truly different sound with which to slay the crew this weekend, look no further than System Beeps, a new album by shiru8bit — though you may have to drag your old 486 out of storage to play it. Yes, this album runs in MS-DOS and its music is produced entirely through the PC speaker — you know, the one that can only beep.

Now, chiptunes aren’t anything new. But the more popular ones tend to imitate the sounds found in classic computers and consoles like the Amiga and SNES. It’s just limiting enough to make it fun, and of course many of us have a lot of nostalgia for the music from that period. (The Final Fantasy VI opening theme still gives me chills.)

But fewer among us look back fondly on the days before sample-based digital music, before even decent sound cards let games have meaningful polyphony and such. The days when the only thing your computer could do was beep, and when it did, you were scared.

Shiru, a programmer and musician who’s been doing “retro” sound since before it was retro, took it upon himself to make some music for this extremely limited audio platform. Originally he was just planning on making a couple of tunes for a game project, but in this interesting breakdown of how he made the music, he explains that it ended up ballooning as he got into the tech.

“A few songs became a few dozens, collection of random songs evolved into conceptualized album, plans has been changing, deadlines postponing. It ended up to be almost 1.5 years to finish the project,” he writes (I’ve left his English as I found it, because I like it).

Obviously the speaker can do more than just “beep,” though indeed it was originally meant as the most elementary auditory feedback for early PCs. In fact, the tiny loudspeaker is capable of a range of sounds and can be updated 120 times per second, but in true monophonic style can only produce a single tone at a time between 100 and 2,000 Hz, and that in a square wave.

Inspired by games of the era that employed a variety of tricks to create the illusion of multiple instruments and drums that in fact never actually overlap one another, he produced a whole album of tracks; I think “Pixel Rain” is my favorite, but “Head Step” is pretty dope too.

You can of course listen to it online or as MP3s or whatever, but the entire thing fits into a 42 kilobyte MS-DOS program you can download here. You’ll need an actual DOS machine or emulator to run it, naturally.

How was he able to do this with such limited tools? Again I direct you to his lengthy write-up, where he describes, for instance, how to create the impression of different kinds of drums when the hardware is incapable of the white noise usually used to create them (and if it could, it would be unable to layer it over a tone). It’s a fun read and the music is… well, it’s an acquired taste, but it’s original and weird. And it’s Friday.

Gadgets – TechCrunch

eMarketer: Amazon took 2/3 of smart speaker sales in 2018, but Echo will feel the squeeze in 2019

Smart speakers that let you control services and other connected devices in your home will continue to be a popular gift choice during the holiday season and into next year, when usage is set to rise by 15 percent to 74.2 million people in the US, working out to 26.8 percent of the US population, according to estimates from eMarketer.

But while Amazon’s Echo helped to define and still dominates the market, consumers’ love affair with Alexa may be cooling, just a little, as the Echo is finally starting to feel the heat from competitors like Home from Google, Apple’s HomePod and the Sonos One.

A new report estimates that the Echo will have accounted for nearly 67 percent of all smart speaker sales in the US in 2018, with Google taking 29.5 percent and others at 8.3 percent. But by next year, Amazon will drop to 63 percent, Google will bump up to 31 percent and a plethora of smaller OEMs will collectively take 12 percent. Three percent decline doesn’t sound like a lot, but it will be the first time ever that Amazon will have dropped below two-thirds of sales. (And for the record, eMarketer research from the UK found similar numbers and declines.)

eMarketer believes this could be the beginning of a gradual decline for the e-commerce giant that will continue through 2020 as the next wave of adopters increasingly explore other brands.

“Consumers in the market for a smart speaker have more options than ever, and Amazon will lose some of its majority share as a result,” said eMarketer forecasting analyst Jaimie Chung, in a statement. “Google has the Home Mini and Home Hub to compete with Amazon’s Echo Dot and Echo Show, and both the Apple HomePod and Facebook Portal will experience their first holiday season this year. Amazon has remained relevant by plugging Alexa into premium speakers like the Sonos, but even Sonos plans to bring Google Assistant to its devices next year, keeping the two companies neck and neck in the voice assistant race.”

There is a valid question to be asked about what people use their speakers for once they do have them. The main takeaway it seems is that while some device makers may turn speakers into a tidy business, it might be some time before the apps and software built around them monetises as lucratively.

For now, the main purpose seems to be listening to audio, where smart speakers provide a handy way to call up music and hear it — which 79.8 percent of speaker owners say they have done — one reason perhaps that the Sonos and Apple’s HomePod are making some inroads since both companies have put music at the core of their experience.

Second most common usage? Inquiries at 73 percent, which is an are where search giant Google is particularly strong.

Amazon has also made Alexa, in her own way, also a fairly amusing, and sometimes helpful, assistant on various topics, helped significantly by all the skills integrations that have been built. However, one key Alexa/Echo use case for the company has always been voice commerce, providing a new interface for people to be able to shop, to fit scenarios where a screen and keyboard are not as convenient.

For now, however, eMarketer says that this a less popular usage for these devices, and that overall voice commerce will remain a very niche slice of the e-commerce market, accounting for just 0.4 percent of sales, or $ 2 billion. Some 27 percent of speaker owners will experiment with buying something via voice commerce next year — a number that eMarketer revised down from an earlier estimate of 31 percent, while 37.1 percent will “shop” using their smart speakers — that is, ask questions about products, if not actually buy them.

Bad news for all the companies thinking that smart speakers will usher in a new era of smart home device usage: smart home integrations are used by just 34.5 percent of smart speaker users.

Gadgets – TechCrunch