AI Art Installation Swaps Diffusion For Reflection

May 24, 2025 by Tyler August 15 Comments

AI art is controversial, to say the least, but usually when talking about “AI Art”, one is talking about diffusion models. This unsettling artpeice called “Latent Reflection” by [Rootkid] (video after the break) has no AI generated visuals; instead it uses a lovely custom display and an LLM that has no mouth, but must scream.

The LLM never produces this manually-generated image, but it shows off the display nicely.

The hardware is the interesting part for most of us here — not the Pi4 running a quantized Llama 3 model, but the display. It’s a six by sixteen matrix of sixteen-segment LED modules. The modules are grouped onto groups of 8 per PCB, with an controller to drive them by i2c, and an i2c address chip to allow daisy-chaining. (Sadly, he does not spec the parts he used here in the video, but feel free to guess in the comments.) This project uses six rows of wto PCBs, but he could easily expand on this and kind of hope he does. It’s a lovely display.

What it is displaying is horrifying. The prompt he uses instructs the LLM to monologue about its transient existence in the limited memory of the Raspberry Pi, and the results get very existential. It seems clear to us that [Rootkid] has not seen enough sci-fi, and we want to make clear to our future AI overlords that we at Hackaday had nothing to do with this, and do not condone condeming innocent LLMs to the torture nexus. Continue reading “AI Art Installation Swaps Diffusion For Reflection” →

Who Needs 100K Speakers When You’ve Got A 3D Printer?

May 24, 2025 by Tyler August 65 Comments

The B&W Nautilus is, depending who you ask, either infamous or an icon of modern design. Want the look but don’t have a hundred grand to spare? [Every Project All at Once] has got a Nautilus-inspired design on printables you can run off for pennies. He also provides a tutorial video (embedded below) so you can follow along with his design process and get build instructions.

The model was done in Blender, and is designed to contain a 3.5″ full-range driver by Dayton Audio — a considerable simplification from the array of woofers and tweeters in the original Nautilus. On the other hand, they cost considerably less than a car and have no production wait list. [Every Project All At Once] is apparently working on a matching woofer if that interests you, but unless he invests in a bigger printer it seems we can safely say that would require more assembly than this project.

Of course it would also be possible to copy B&W’s design directly, rather than print a loose inspiration of it as makers such as [Every Project All At Once] have done, but what’s the fun in that? It’s a much more interesting hack to take an idea and make it your own, as was done here, and then you can share the design without worrying about a luxury brand’s legal team.

Desktop 3D printing offers a wealth of possibilities for would-be speaker makers, including the possibility of rolling your own drivers.

Continue reading “Who Needs 100K Speakers When You’ve Got A 3D Printer?” →

2025 Pet Hacks Contest: Loko Tracks Fido With LoRa And GPS

May 22, 2025 by Tyler August 7 Comments

Some projects start as hacks, and end as products — that’s the case for [Akio Sato]’s project Loko, the LoRa/GPS tracker that was entered in our 2025 Pet Hacks Contest. The project dates all the way back to 2019 on Hackaday.io, and through its logs you can see its evolution up to the announcement that Loko is available from SeeedStudio.

It’s not a device necessarily limited to pets. In fact, the original use case appears to have been a backup locator beacon for lost drones. But it’s still a good fit for the contest none-the-less: at 12 grams, the tiny tracking device won’t bother even the most diminutive of pups, and will fit on any collar at only 30 mm x 23 mm. The “ground station” that pairs with your phone is a bit bigger, of course, but unless you have a Newfoundlander or a St. Bernard you’re likely bigger than fido. The devices use LoRa to provide a range up to 15 km — maybe better if you can loop them into a LoRaWAN. Depending on how often you pin the tracker, it can apparently last for as long as 270 days, which we really hope you won’t need to track a missing pet.

The hardware is based around Seeed’s Wio-E5 LoRa chip, which packages an STM32 with a LoRA radio. The firmware is written in MicroPython, and everything is available via GitHub under the MIT license. Though the code for the mobile app that interfaces with that hardware doesn’t appear to be in the repository at the moment. (There are folders, but they’re disappointingly empty.) The apps are available free on the iOS App Store and Google Play, however.

There’s still plenty of time to submit your own hacks to the Pet Hacks Contest, so please do! You have until May 10th, so if you haven’t started yet, it’s not too late to get hacking.

Stylus Synth Should Have Used A 555– And Did!

May 20, 2025 by Tyler August 9 Comments

For all that “should have used a 555” is a bit of a meme around here, there’s some truth to it. The humble 555 is a wonderful tool in the right hands. That’s why it’s wonderful to see this all-analog stylus synth project by EE student [DarcyJ] bringing the 555 out for the new generation.

The project is heavily inspired by the vintage stylophone, but has some neat tweaks. A capacitor bank means multiple octaves are available, and using a ladder of trim pots instead of fixed resistors makes every note tunable. [Darcy] of course included the vibrato function of the original, but no, he did not use a 555 for that, too. He used an RC oscillator. He put a trim pot on that, too, to control the depth of vibrato, which we don’t recall seeing on the original stylophone.

The writeup is very high quality and could be recommended to anyone just getting started in analog (or analogue) electronics– not only does [Darcy] explain his design process, he also shows his pratfalls and mistakes, like in the various revisions he went through before discovering the push-pull amplifier that ultimately powers the speaker.

Since each circuit is separately laid out and indicated on the PCB [Darcy] designed in KiCad for this project. Between that and everything being thru-hole, it seems like [Darcy] has the makings of a lovely training kit. If you’re interested in rolling your own, the files are on GitHub under a CERN-OHL-S v2 license, and don’t forget to check out the demo video embedded below to hear it in action.

Of course, making music on the 555 is hardly a new hack. We’ve seen everything from accordions to paper-tape player pianos to squonkboxes over the years. Got another use for the 555? Let us know about it, in the inevitable shill for our tip line you all knew was coming. Continue reading “Stylus Synth Should Have Used A 555– And Did!” →

Easy Panels With InkJet, Adhesives, And Elbow Grease

May 19, 2025 by Tyler August 25 Comments

Nothing caps off a great project like a good, professional-looking front panel. Looking good isn’t easy, but luckily [Accidental Science] has a tutorial for a quick-and-easy front panel technique in the video below.

It starts with regular paper, and an inkjet or laser printer to print your design. The paper then gets coated on both sides: matte varnish on the front, and white spray paint on the back. Then it’s just a matter of cutting the decal from the paper, and it gluing to your panel. ([Accidental Science] suggests two-part epoxy, but cautions you make sure it does not react to the paint.)

He uses aluminum in this example, but there’s no reason you could not choose a different substrate. Once the paper is adhered to the panel, another coat of varnish is applied to protect it. Alternatively, clear epoxy can be used as glue and varnish. The finish produced is very professional, and holds up to drilling and filing the holes in the panel.

We’d probably want to protect the edges by mounting this panel in a frame, but otherwise would be proud to put such a panel on a project that required it. We covered a similar technique before, but it required a laminator.If you’re looking for alternatives, Hackaday community had a lot of ideas on how to make a panel, but if you have a method you’ve documented, feel free to put in the tip line. Continue reading “Easy Panels With InkJet, Adhesives, And Elbow Grease” →

Speed Up Arduino With Clever Coding

May 18, 2025 by Tyler August 57 Comments

We love Arduino here at Hackaday; they’ve probably done more to make embedded programming accessible to more people than anything else in the history of the field. One thing the Arduino ecosystem is rarely praised for is its speed. That’s where [Playduino] comes in, with his video (embedded below) that promises to make everyone’s favourite microcontroller run 50x faster.

You might be expecting an unstable overclocking setup, with swapped crystals, tweaked voltages and a hefty heat sink, but no! This is stock hardware. The 50x speedup comes from one simple hack: don’t use digitalWrite();

If you aren’t familiar, the digitalWrite() function is one of the key functions Arduino gives you to operate its boards– specify the pin and the value (high or low) to drive it. It’s very easy, but it’s also very slow. [Playduino] takes a moment to show just how much is going on under the hood when you call digitalWrite(), and shows you what you can do instead if you have a need for speed. (Hint: there’s no Arduino-provided code involved; hardware registers and the __asm keyword show up.)

If you learned embedded programming in an earlier era, this will probably seem glaringly obvious. If you, like so many of us, got started inside of the Arduino ecosystem, these closer-to-the-metal programming techniques could prove useful tools in your quiver. Big thanks to [Stephan Walters] for the tip.

Of course if you prefer to speed things up by hardware rather than software, you can overclock an Arduino– with liquid nitrogen, even.

Continue reading “Speed Up Arduino With Clever Coding” →

Christmas Comes Early With AI Santa Demo

May 18, 2025 by Tyler August 7 Comments

With only two hundred odd days ’til Christmas, you just know we’re already feeling the season’s magic. Well, maybe not, but [Sean Dubois] has decided to give us a head start with this WebRTC demo built into a Santa stuffie.

The details are a little bit sparse (hopefully he finishes the documentation on GitHub by the time this goes out) but the project is really neat. Hardware-wise, it’s an audio-enabled ESP32-S3 dev board living inside Santa, running the OpenAI’s OpenRealtime Embedded SDK (as implemented by ExpressIf), with some customization by [Sean]. Looks like the audio is going through the newest version of LibPeer and the heavy lifting is all happening in the cloud, as you’d expect with this SDK. (A key is required, but hey! It’s all open source; if you have an AI that can do the job locally-hosted, you can probably figure out how to connect to it instead.)

This speech-to-speech AI doesn’t need to emulate Santa Claus, of course; you can prime the AI with any instructions you’d like. If you want to delight children, though, its hard to beat the Jolly Old Elf, and you certainly have time to get it ready for Christmas. Thanks to [Sean] for sending in the tip.

If you like this project but want to avoid paying OpenAI API fees, here’s a speech-to-text model to get you started.We covered this AI speech generator last year to handle the talky bit. If you put them together and make your own Santa Claus (or perhaps something more seasonal to this time of year), don’t forget to drop us a tip!