Stable Diffusion, the slick generative AI tool, just launched and went live on GitHub

You’ve almost certainly already seen a torrent of high-quality text-to-image generation pics lately. Now one of the most impressive generative tools has published code and opened up to academic researchers. It’s called Stable Diffusion – and you’ll probably be hearing a lot more soon.

The launch

Emad Mostaque from Stability AI has the news – no doubt timed to this month’s SIGGRAPH event:

Stable Diffusion launch announcement

And then it’s now live on GitHub (though you’ll need to submit academic institutional credentials in order to gain access to checkpoints, if you aren’t already in the now-closed-up beta):

Those of you who have already been following this one probably jumped straight through to the blog and GitHub repository, but here’s the short version for everyone else:

Recipe for Stable Diffusion

First, you start with some of the leading-edge research into high-resolution image synthesis using latent diffusion models. The basic notion there is, someone else takes care of the machine learning “training” and dataset – and the massive amounts of GPU power consumption that involves – and you just type in text and get whole images back, quickly and without much computation.

This particular version of the model is creating images that are stunningly coherent – and that in turn has come out of a lot of human collaboration and a lot of data science (both in terms of the research and the scale of the data being used). As they explain:

The model itself builds upon the work of the team at CompVis and Runway in their widely used latent diffusion model combined with insights from the conditional diffusion models by our lead generative AI developer Katherine Crowson, Dall-E 2 by Open AI, Imagen by Google Brain and many others.

There’s also big data used to weight the output of the big data – much as social media currently works. So this particular CLIP model “filtered LAION-5B based on how “beautiful” an image was, building on ratings from the alpha testers of stable diffusion.”

[embedded content]

Humans, machines, big data, and openness

That’s a big deal. If these look really good, it’s partly because humans are choosing what looks good. Note that that could also shift as tastes change over time – any old-timers remember when Flickr was littered with HDR images? Or check this article from 2006. Part of what we’re seeing in machine learning is an extension of what we’re seeing in social media – large groups of humans deciding collectively what it and isn’t valuable, for better and for worse.

How big is that task? Well:

The model was trained on our 4,000 A100 Ezra-1 AI ultracluster over the last month as the first of a series of models exploring this and other approaches.

We have been testing the model at scale with over 10,000 beta testers that are creating 1.7 million images a day.

To me, these limitations are fascinating, and I think they can deepen appreciation for what you’re looking at and how it might be used. So this isn’t necessarily critical, though it could lead to some criticism.

The flipside of centralizing computation, too, is that your local computer doesn’t have to do much work – a consumer GPU with less than 10 GB VRAM can make a 512×512 image in seconds, the creators say.

Unlike some things that call themselves “open,” the LAION project on which this is based has a mix of members in a non-profit, entirely funded by donations and public research grants. So the ultimate goal they say is “aiming to make large-scale machine learning models, datasets and related code available to the general public.”

[embedded content]

But what are these images?

It’s important to understand that the AI is not “creating” images, despite the “generative” label. It’s more accurate to say that what you’re looking at, convincing as it is, is emergent from patterns in the data, guided by your text prompt. That might sound the same, but it isn’t. It says a lot about the content of the data set and the particular way these convolutional techniques navigate it based on text that you see what you see.

It’s even important not to retroactively apply the way machine learning works to your brain. So no, absorbing images and then spitting them out like this is probably not quite how your brain works. That’s happened a lot in tech, which is why past centuries start to imagine brains like clocks (no, not at all) or computers (still no). To really delve into that and how to understand the way the brain works with images, we should get a conversation going with neurologists and psychologists and maybe an anthropologist. I’m of less use as a musicologist and media art historian or whatever I am but … well, I could probably get them talking.

I do expect we’ll see a lot of these images, and that it will disrupt businesses around stock imagery and illustration. That disruption could create some opportunities – and some pain. I’d have to be really technologically optimistic to imagine it’s all good and “progress.” We could also see so much of this sort of imagery that eventually there’s a counter-reaction and… I’m getting ahead of myself. But this will be one to discuss, including, as it happens, the panel I’m moderating at MUTEK later this month if you’re around in Montreal.

As impressive as the generated images are, it’s also likely that these techniques will be applied to other contexts, as well – like image repair, upscaling, 3D applications, and some really wild possibilities for editing and whatnot.

And yes, I’ve tagged this under ‘motion’ partly because motion artists are deeply involved in exploring machine learning, but also because no doubt we’ll see moving applications of this technique soon, too.

I’m sure we’ll be talking about this, so feel free to argue with me on this – and I’m curious how folks are using these tools artistically (or not), research you’re doing, and how you’re thinking about them.

All images published on GitHub using Stable Diffusion.

Make your own MIDI-to-CV adapter in a skull, with a 3D printer and CircuitPython

Liz Clark over at Adafruit has a great DIY project. It’s compact. It’s beginner-friendly. It teaches you some simple embedded Python with Circuit Python. It’s… a skull.

The basic ingredients:

  • A badass 3D-printed skull enclosure (all nicely sized)
  • CircuitPython code that converts MIDI to voltage
  • Digital-to-analog converter (MCP4725 DAC, which has its own tutorial)
  • A cute little chip to handle the python – aptly named QT Py (“cutie pie!”) – aka the Raspberry Pi RP2040 – check their dedicated guide

If you want to start at the very beginning, see also their awesome MIDI for Makers guide, also by Liz.

This being Adafruit, there’s an immaculately organized project guide:

CircuitPython MIDI to CV Skull

Watch it in action:

[embedded content]

I was going to write “pink” or “red” – seems a dark salmon. It’s colored awesome. Let’s say that.

In any event, what’s equally beautiful is all the parts are in stock, including that 12-bit DAC, which is something no one takes for granted in 2022. (Okay, the hookup wires is out of stock at the moment, but that’s literally the easiest thing here to locate.) I also love that cute little proto-board, which looks handy for Teensy and whatnot, too.

No affiliate links here – sort of the reverse, as I think I might have just sold myself on this, too. And yeah, Adafruit Industries team, you all continue to demonstrate how to do these guides of guides!

Images here (CC-BY-SA) Liz Clark.

Premiere: Stones Taro ‘99’

London’s Sub Merchants label will release a new six-track EP, ‘Columbia Road’, this month, bringing together Main Phase, Daffy and Stones Taro for a late summer suite of warped and wobbly garage, with a few curve balls thrown in for good measure. Listen to Stones Taro’s ‘99’ below. 

The EP features two tracks from each artist, with ATW Records co-founder Main Phase kicking things off with the dubwise snapper, ‘Smoke’, featuring MC Rakjay. Daffy’s ‘Roughneck Sound’ rips its deep, velvety bass and keys apart with a sharp two-step beat, while Kyoto producer Stones Taro turns in an acid-dipped electro cut with ‘Knock My Ear’.

The B-side keeps the same momentum, with Main Phase delivering the aptly titled ‘Big Shaker’ and Daffy going two-for-two on ceiling-snapping UKG. Stones Taro caps things off with a classic breakbeat hardcore bang, launching its beat into the stratosphere with snarling bass and ravey zaps. It’s a recipe for headrush dancefloor peaks, and in a release that’s full of them, ‘99’ is a moment-maker built for this summer’s hot, sweaty and ecstatic festival run. 

‘Columbia Road’ will be released on 19th August. Pre-order it here

George Riley shares new single, ‘Delusion’: Listen

George Riley has released a new single, called ‘Delusion’.

Seeing her take aim at fake friends, the track is the latest in a series of singles to emerge from Riley. All have been produced by UK artist Vegyn, who has also executive produced her forthcoming record, ‘Running In Waves’.

Announced last month, the project is set to be released on 9th September via Vegyn’s PLZ Make It Ruins label. Alongside the record’s announcement, Riley also shared the track ‘Time’.

Earlier this year, her tracks ‘Jealousy‘ and ‘Sacrifice‘, which also feature Vegyn, were released, building on the momentum of last year’s breakout mix, ‘interest rates, a tape’ and her appearance on Anz’s Ninja Tune-released track, ‘You Could Be‘. 

Watch a lyric video for new cut ‘Delusion’ below.

Hudson Mohawke shares new video for ‘Bicstan’, starring Patti Harrison: Watch

Hudson Mohawke has unveiled a video for previously released single ‘Bicstan’.

The video, which you can watch below, is co-directed by Alan Resnick and actor/comedian Patti Harrison, and the latter also stars in it. The visual focuses on a 1v1 basketball match between a player and a dancer, with the latter dominating the game thanks to supernatural talents.

The sharing of the video comes ahead of the release of Hudson Mohawke’s latest album, ‘Cry Sugar’, on Warp Records this Friday, 12th August. The album was first announced in June when ‘Bicstan’ was first released.

The record is inspired by “apocalyptic” film scores and soundtracks, like the works of the late Vangelis and John Williams’ ’90s material. According to the press release, the 19-track double-LP “score[s] the twilight of our cultural meltdown.”

Since announcing the album, the Scottish producer has also gone on to share two more tracks from the record: ‘Stump’ and ‘Dance Forever’.

Revisit Amy Fielding’s cover feature with Hudson Mohawke from September 2020.

Eric Prydz ‘Opus’ vinyl sells for $2,000 on Discogs

A 4xLP limited edition pressing of Eric Prydz’s 2016 album ‘Opus’ has sold for $2,000 on Discogs.

The sale happened in June of this year, and the final amount made it the 24th most expensive item to be sold via the platform that month.

The particular pressing of the album was limited to 500 copies upon its release, and was available only to people who pre-ordered the record. Pressed on 180g vinyl, it also included a hand-numbered and signed art insert. Find the Discogs listing here.

Find out more about the 30 most expensive items sold on Discogs in June 2022 via the platform’s blog post.

Earlier this month, Tomorrowland announced that it would be promoting the only performance of Eric Prydz’s HOLO show in Europe for the remainder of the year at Amsterdam’s Ziggo Dome on 20th October.

The Chemical Brothers to headline Amnesia Ibiza’s 2022 closing party

The Chemical Brothers will headline Amnesia Ibiza’s closing party this October with their only DJ set on the island this year.

The club’s closing event will go down on 15th October, kicking off at 6pm and running through late into the night. The rest of the line-up for the party is set to be announced in the coming months. 

An Instagram post shared by the duo, which you can see below, says they will be “spinning an array of sonic classics – both old and new”.

Tickets for the party are on sale now via Amnesia’s website.

In June, the Chemical Brothers shared a previously unreleased track, called ‘Cylinders’, as part of the 25th anniversary reissue of their classic album ‘Dig Your Own Hole’.

They were also due to DJ at Glastonbury in June, but were forced to pull out just ahead of the event due to COVID-19 cases within the duo and their touring crew.

Exit Festival 2022: five of the best performances

The way into Serbia’s EXIT Festival is unlike any other. Every year, tens of thousands cross the Danube — Europe’s second largest river, running from Germany’s Black Forest through Ukraine to the Black Sea — into the Petrovaradin Fortress, a 17th-century citadel built during the Ottoman Empire. As we pass through the outermost walls and slowly ascend the site proper, waves of sound work their way down the winding cobbled streets. For four days in early July, this sleepy fort is transformed into a sprawling network of music stages. Rock, metal and indie artists play alongside dance, reggaeton, psy-trance and hip-hop acts. This year Calvin Harris, Iggy Azalea, Nick Cave and grindcore pioneers Napalm Death play the same stage; as one of Europe’s biggest outdoor music festivals, there really is something for everyone.

But there’s more to EXIT than its historic location and delightfully random line-ups. The roots of the festival date back to a darker period in Serbian history, a decade into the Yugoslav wars, where conflicts along ethnic lines led to serious war crimes, among them genocide. A product of the student movement, the first EXIT — or Zero EXIT as it was then known — emerged in the wake of a series of protest concerts.

One concert held in 1999, Šakom u glavu (Fist to the head), showed clips of Yugoslavian President Slobodan Milošević’s clear abuse of powers. The crowd’s angry response sparked an idea among student union reps, and a year later EXIT was born. The annual event has since morphed from a grassroots, counter-cultural movement into a well-oiled production of over 40 stages, pulling top-tier headliners from all manner of music genres and revellers from over 70 countries. But that spirit of basic goodwill, togetherness and the belief that music can bring about social change, remains. Here are five key performances from EXIT 2022.

Printworks announces schedule for second half of 2022 following news of forthcoming closure

Printworks announces schedule for second half of 2022 following news of forthcoming closure139345470 1956384607834074 6173666686082533127 N

Printworks won’t be around forever but it has officially confirmed that it’ll still be a part of London through the rest of 2022. This past July, news broke that London’s Southwark Council had approved renovation plans that would swap out Printworks and the surrounding area in favor of office buildings and retail locations.

There hadn’t been a concrete date given for the permanent closure of Printworks—which initially opened in 2017 with plans to only keep its doors open for five years—but on September 23, it’ll resume all show activities for the impending fall and winter seasons. Beginning with a grand reopening evening with DJ Snake, Printworks will play host to Aluna, Charlotte de Witte, Paul van Dyk, Solardo, Wilkinson, Bonobo, Afterlife, Gorgon City, and ultimately a winter closing party with Disciples, SIDEPIECE, Noizu and more.

While we’ll just have to wait and see if this will ultimately be Printworks’ final run of shows, the full AW22 schedule can be found here.

Featured image: Printworks

The post Printworks announces schedule for second half of 2022 following news of forthcoming closure appeared first on Dancing Astronaut.

Northern Ireland’s Free The Night campaign launches nightlife survey

Northern Ireland-based nightlife advocacy organisation Free The Night has launched a new nationwide survey on the nightlife industry.

It’s hoped that information shared via the survey will help the group to gain better insight into people’s experiences and perceptions of nightlife in Northern Ireland. The group also says the survey data will serve as a useful tool in advocating for more progressive licensing policies.

Areas covered by the survey include safety, transport infrastructure, social mobility and the importance of late-night culture. It forms part of a larger project of research by Free The Night and Queen’s University Belfast PhD candidate Ciara Power, with hopes existing that it can be put to use during an upcoming licensing review panel, which is due to be set up by the Northern Ireland Assembly later this year.

That review will look into the impact of licensing laws on Northern Ireland’s nightlife industry, and focus on areas such as the economic effect, post-pandemic recovery, policing and public health, and what exact effects might arise from additional licensing hours across Northern Ireland’s clubs and bars.

Those who particularly find themselves involved as a punter or worker across Northern Ireland’s nightlife industry are encouraged to fill out the survey, which you can find here.

Speaking about it, DJ and Free The Night co-founder Holly Lester said: “This is a really important opportunity for those who feel like they have not had a say in issues affecting the night time economy to step forward and make their voice heard. We hope that the outcome of this survey and the larger body of research that we are undertaking will pave the way for a more progressive, diverse and equitable landscape in Northern Irish nightlife.”

Nightlife opening times in Northern Ireland are among the most restrictive across the UK and Europe, with pubs, bars and nightclubs only currently allowed licenses to serve alcohol until 11pm, 12am or 1am respectively. A review of this matter has long been promised, with Free The Night at the forefront of advocating for such a review since its founding in 2021.

Find out more about the perspectives of Holly Lester and other DJs on how nightlife can be improved via our 2021 feature.

Revisit another DJ Mag feature from 2021, assessing the complicated reality of the nightlife industry here.