Hackers love to make music with things that aren’t normally considered musical instruments. We’ve all seen floppy drive orchestras, and the musical abilities of a Tesla coil can be ear-shatteringly impressive. Those are all just for fun, though. It would be nice if there were practical applications for making music from normally non-musical devices.
Thanks to a group of engineers at Case Western Reserve University in Cleveland, there is now: a magnetic resonance imaging machine that plays soothing music. And we don’t mean music piped into the MRI suite to distract patients from the notoriously noisy exam. The music is actually being played through the gradient coils of the MRI scanner. We covered the inner working of MRI scanners before and discussed why they’re so darn noisy. The noise basically amounts to Lorenz forces mechanically vibrating the gradient coils in the audio frequency range as the machine shapes the powerful magnetic field around the patient’s body. To turn these ear-hammering noises into music, the researchers converted an MP3 of [Yo Yo Ma] playing [Bach]’s “Cello Suite No. 1” into encoding data for the gradient coils. A low-pass filter keeps anything past 4 kHz from getting to the gradient coils, but that works fine for the cello. The video below shows the remarkable fidelity that the coils are capable of reproducing, but the most amazing fact is that the musical modification actually produces diagnostically useful scans.
Our tastes don’t generally run to classical music, but having suffered through more than one head-banging scan, a half-hour of cello music would be a more than welcome change. Here’s hoping the technique gets further refined.
Continue reading “Musical Mod Lets MRI Scanner Soothe the Frazzled Patient”
[Nop head] discovered that cheap multimeter leads costing only a few bucks can come with more than one may have bargained for. The first set had a large amount of useful-looking attachments, but the wires used for the leads were steel with a resistance of about one ohm each. With two leads in use, that means any resistance measurement gets two ohms added for free. More seriously, when measuring current, the wires can heat up rapidly. Voltage measurements would be affected the least, but the attachments and lead design expose a large amount of bare metal, which invites accidental shorts and can be a safety hazard with higher voltages.
Are all cheap multimeter leads similarly useless? Not necessarily. [nop head] also purchased the set pictured here. It has no attachments, but was a much better design and had a resistance of only 64 milliohms. Not great, but certainly serviceable and clearly a much better value than the other set.
It’s usually not possible to identify garbage before it’s purchased, but [nop head] reminds us that if you do end up with trash in hand, poor quality counterfeits can be good for a refund. That goes for electronic components, too.
Microsoft is really leaning into vaporwave these days. Microsoft is giving away knit Windows sweaters to social media influencers. Is it for an ugly sweater contest? Maybe, or maybe Microsoft is capitalizing on the mid-90s ＡＥＳＴＨＥＴＩＣ. Recently, Apple got back in their 90s logo game with the release of a few ‘rainbow Apple’ t-shirts. The spirit of the 90s lives on in tech culture.
Have a Hackerspace? Frack is organizing the great Inter-hackerspaces Xmas goodies swap! Since your hackerspace is filled with weird ephemera and random crap, why not box it up and send it out to another hackerspace? You’ll probably get another random box of crap in return!
Just an observation looking for commentary, but is Thingiverse slow these days? It seems really, really, really slow these days.
The Blockchain makes it to the Apple II! By far, the most interesting thing in tech right now is the blockchain, with AI, at the edge. This will get your Merkle trees tinglin’ with some AI, and 5G is where it’s at. We’re back with cylinder computing this time, and this is the greatest achievement that will synthesize brand new paradigms. Of course, if it weren’t for millennials, we’d have it already.
There’s a new portable console out there, and it’s at the top of everyone’s Christmas lists. The SouljaGame Handheld is a rebrand of what’s available on AliExpress. What makes this one different? It has Soulja Boy’s name on it. If you couldn’t get your hands on the SouljaGame Handheld, don’t worry: Post Malone Crocs are available on eBay for about $300.
One of the most common uses of neural networks is the generation of new content, given certain constraints. A neural network is created, then trained on source content – ideally with as much reference material as possible. Then, the model is asked to generate original content in the same vein. This generally has mixed, but occasionally amusing, results. The team at [Made by AI] had a go at generating Christmas songs using this very technique.
The team decided that the easiest way to train their model would be to use note data from MIDI files. MIDI versions of Christmas songs are readily available and provide a broad base with which to train the model. For a neural network, the team chose to use a Long-short Term Memory (LSTM) architecture. This is a model which is contextually sensitive, which is important when dealing with structured formats like music or language.
The neural network generated five tunes which you can listen to on the Made by AI Soundcloud page. The team notes their time was limited, and we think that with some further work and more adherence to musical concepts such as structure and repetition, it might be possible to generate something a little more catchy.
There are other applications for AI in music, too – like these intelligent musical prostheses.
[Mark Rehorst] has been busy with his Ultra MegaMax Dominator (UMMD) design for a 3D printer, and one of the many things he learned in the process was how not to design a 3D printed belt clamp. In the past, we saw how the UMMD ditched the idea of a lead screw in favor of a belt-driven Z axis, but [Mark] discovered something was amiss when the belts were flopping around a little, as though they had lost tension. Re-tensioning them worked, but only for a few days. It turned out that the belt clamp design he had chosen led to an interesting failure.
The belts used were common steel-core polyurethane GT2 belts, and the clamp design uses a short segment of the same belt to lock together both ends, as shown above. It’s a simple and effective design, but one that isn’t sustainable in the longer term.
The problem was that this design led to the plastic portion of the belt stretching out and sliding over the internal steel wires. The stretching of the polyurethane is clear in the image shown here, but any belt would have had the same problem in the clamp as it was designed. [Mark] realized it was a much better idea to use a design in which the belts fold over themselves, so the strain is more evenly distributed.
[Mark] has been sharing his experiences and design process when it comes to building 3D printers, so if you’re interested be sure to check out the UMMD and its monstrous 695 mm of Z travel.
We’re not allowed to have TV here in the Hackaday Wonder Bunker, but occasionally we’ll pool together the bandwidth credits they pay us in and gather ’round the old 3.5 inch TFT LCD to watch whatever Netflix assures us is 93% to our liking. That’s how we found out they’ve made a show based on, of all things, one of the Castlevania games for the NES. We wanted to play the game to understand the backstory, but since it hails from the era of gaming where primitive graphics had to be supplemented with soul-crushing difficulty, we didn’t get very far.
But thanks to a very impressive project developed by [Michael Birken] maybe we’ll have it all figured out by the time we’ve saved enough credits to watch Season 2 (no spoilers, please). The software, which he’s quick to point out is not an example of machine learning, is an attempt to condense his personal knowledge of how to play Castlevania into a plugin for the Nintaco NES emulator. The end result is CastlevaniaBot, which is capable of playing through the original Castlevania from start to finish without human intervention. You can even stop and start it at will, so it can play through the parts you don’t want to do yourself.
[Michael] started this project with a simple premise: if he could make a bot successfully navigate the many levels of Dracula’s castle, then getting it to kill a few monsters along the way should be easy enough. Accordingly, he spent a lot of time perfecting the path-finding for CastlevaniaBot, which included manually playing through the entire game in order to get an accurate map of the background images. These images were then analyzed to identify things like walls and stairs, so the bot would know where it could and couldn’t move protagonist Simon Belmont. No matter what the bot is doing during the game it always considers where it is and where it needs to be going, as there’s a time limit for each stage to contend with. Continue reading “AI Bot Plays Castlevania So You Don’t Have to”
When we took a biology lab, you had to use a mouth pipette to transfer liquids around. That always seemed odd to use your mouth to pick up something that could be dangerous. It’s also not very efficient. A modern lab will use a liquid handling robot, but these aren’t exactly cheap. Sometimes these are called pipettors and even a used one on eBay will set you back an average of $1,000 — and many of them much more than that. Now there’s an open source alternative, OpenLH, that can be built for under $1,000 that leverages an open source robot arm. You can find a video about the system below.
The robot arm, a uArm Swift Pro, is the bulk of the cost. The Pro can also operate as a 3D printer or a laser engraver with a little work. In fact, we wondered if you could use the arm to make a 3D printer and then print the parts you need to convert it to a liquid handler. Seems like it should work.
Continue reading “OpenLH: Automating Biology for Everyone”