transmissions

the signal is out tomorrow. slow release lp on bandcamp from 19 april. two compact discs. twelve tracks. two hours of continuous dub techno, bounced in one piece, mixed as one room.

the title is a promise. nothing here rushes. the opener breathes for nine minutes before it finds the grid. the closer takes eighteen to finish what it starts. the set was built to be sat with, not scrolled past. if the algorithm hates it, that's a feature.

the map is not linear

this isn't a dj set and it isn't an album in the old sense. it's a drift. the tracks are places (or something between places) pulled out of the northern post-industrial ether:

waterworx. underbank. ancoats. viaduct. pennine drift. canal chaos. temporary digital zone. welcome to sheffield. trans-pennine. midweek transmission. snake pass. sunday on the mersey.

read them in that order if you want a route. read them in any order, they still add up. the record moves through stockport, manchester and sheffield, but it doesn't obey the map. it drops into viaducts and rises on the pennine wind. the channels route themselves. the delays decide their own length. c minor holds the whole thing together.

sovereign masters

every sound made here. every plate of reverb, every tape return, every sub on this record was mixed and mastered on local hardware. no stems uploaded to a cloud. no mastering chain rented by the month. no ai "finalising" the peak limiter. the masters live on a drive in a room i can point at. when the internet takes its next scheduled outage, these files still exist.

that matters less if you're streaming. it matters more if you're pressing. these discs are objects. hold one and you're holding the frequency. the jewel of the thing is that it doesn't phone home.

how to get it

on bandcamp from 19 april: ufoutlaw.bandcamp.com/album/slow-release-lp. digital download and 2×cd. buy it direct. no distributor. no middle tier. you pay the artist, the artist pays the pressing plant, the artwork arrives in the post.

in person, 19 april, odioba, stockport: cds in hand. first pressing. signed if you want. come through. two hours live, then the physical objects change hands. that's the transaction.

what this is for

dub techno is a spiritual practice disguised as a genre. moritz, stott, vainqueur, wata. the lineage is long and the instructions are the same: hold the minor key, let the delay lead, leave the room enough silence to breathe. this record is my version of that practice. two hours of listening to the signal come in and go out. no hook. no drop. just the space.

if you're here, you already get it. thank you for finding the frequency. tell one other person. that's the whole marketing plan.

see you on the other side of the signal. 19 april. sovereign and slow.

the cover is the first transmission. before the music, before the tracklist, before the bandcamp page finishes loading. the cover decides whether anyone presses play.

that is the brief. not "make something pretty". make something that carries the frequency.

both records, liminal consciousness ep and slow release lp, started the same way. a 3d scene in blender. hours of lighting. nothing else.

why blender

blender is free. it is open source. it runs on my machine. the files live on my drive. no subscription, no cloud export, no licence ticking down in the background. this is the visual identity's equivalent of mixing on local hardware: sovereign tools. the aesthetic is the ethic.

and blender is patient. you can sit with a scene for weeks. move one light half a centimetre. render again. watch a single reflection change. that slow pace matches the music.

the shape of a cover

a ufo cover should feel like a room you could enter, not a poster you could read. that is the rule. everything else is technique.

deep black foreground. one luminous teal-cyan focal point. a trace of amber in the low atmosphere. negative space doing the work. typography sitting at the edges, light-weight, lowercase, never shouting.

no waveforms. no equaliser bars. no turntables. no neon. no fractals. no tie-dye. no ai-generated 'electronic music' soup. most of the craft is in what you leave out.

the render is the starting point, not the end

the raw blender output is never the final image. that is step one of maybe four.

into photoshop for colour grading. darken the blacks further. lift the cyan highlight so it pulses. warm the amber in the lower band. add a very slight grain so the whole thing feels physical rather than digital.

then let it sit for a day. come back. look at it cold. often that is when you see it is too busy, or too dim, or the focal point is not actually focal. fix it. let it sit again.

the first time the cover tells you to stop is usually the right time.

one sound, two discs

the liminal consciousness cover is a single still: a signal suspended in a deep column of black, a slow warmth bleeding up from the floor of the frame. that one image was the whole brief. it is also the template for everything that came after.

slow release lp needed more, because it is two discs: a journey, twelve places, two hours. the disc one and disc two artworks use the same base render, pushed in opposite directions. cd one is colder, deeper blue, further into the trench. cd two is warmer, more amber at the edges, coming home across the pennines.

the two discs look like they are part of the same weather system. you can tell they belong together without the artwork having to shout 'album' anywhere on it.

on aesthetics (a brief aside)

aesthetics are not decoration. aesthetics are the argument.

a neon-on-black club flyer argues for the drop. a bright, oversaturated cover argues for the algorithm. a glossy ai-collage argues that the image is cheap. none of those arguments match what dub techno is actually trying to do.

dub techno asks you to slow down, sit with something, let the delay decide. a cover that can persuade a stranger to do that has to look like an invitation to be still. cold light. deep black. patience in every pixel. the music and the image arguing for the same thing.

that is the whole aesthetic brief. the rest is hours in blender.

one workflow, forever

this is the pipeline for every ufo release from here. blender scene, post in photoshop, lowercase sans-serif typography at the edges. no stock. no ai. the tools are mine. the images are mine. the frequency is the thing.

if a future cover breaks the pattern, the pattern is probably wrong and we fix it before we press. so far it has held.

liminal consciousness ep and slow release lp are both on bandcamp. the images are the door. the music is the room.

two hours of continuous dub techno. seventeen tracks in C minor. ambient opening through peak psychedelic techno and back to silence. the whole thing recorded, mixed, and mastered in-house. sovereign masters on local hardware.

and we're pressing it onto two compact discs. you should do the same.

the case

streaming pays roughly £0.003 per play. for a two-hour ambient dub techno mix from an independent artist in stockport, 150,000 streams is a fantasy. and even if you got them, you'd clear maybe £450. after months of waiting for payment through a distributor you never met.

a CD run of 150 gatefold digipaks costs between £400 and £580. sell them at £12 to £15 each and you've covered your costs by 40 copies. everything after that is yours. no 30% platform cut. no distributor fee. no three-month payment cycle. no algorithm deciding who gets to hear you.

the maths is not complicated. physical wins.

how to do it

this is the process. it's simpler than you think.

1. finish the music. this is the hard part and nobody can do it for you. but it doesn't need to be perfect. it needs to be done. if you're waiting for the mix to be flawless you'll be waiting forever. bounce it. commit to it. move on.

2. find a manufacturer. in the uk, look at key production, duplication.co.uk, or dms. get quotes for short runs. 100 to 200 copies is the sweet spot for a first release. ask about gatefold digipaks if you want something that feels substantial. standard jewel cases are cheaper but they crack and nobody loves them.

3. prepare your artwork. you need a front cover, back cover, and disc print as a minimum. for a digipak you'll also need inside panels. work to the manufacturer's templates. if you can use blender, photoshop, or even canva, you can do this yourself. keep it simple. the music is the point.

4. sort your audio files. the manufacturer will want CD-quality WAV files (16-bit, 44.1kHz). if your mix is continuous, decide where the split points are. label everything clearly. include ISRC codes if you have them. if you don't, you can get them free from PPL in the uk.

5. order early. manufacturing takes two to three weeks minimum. add a week for proofing and a week for delivery. if you have a deadline, work backwards from it and add a buffer. things go wrong. give yourself room.

6. sell direct. bandcamp for online. your own hands at gigs. a mailing list for announcements. that's the distribution network. you don't need spotify, you don't need a label, you don't need anyone's permission. you made the thing. now sell the thing.

the object matters

there's something that happens when you hold the thing you made. a CD in a gatefold digipak isn't a file on a server. it's a physical object with weight and edges and artwork you can actually see. it exists whether the internet is on or not.

100 copies. hand-numbered. each one is a sovereign object. it doesn't phone home. it doesn't track your listening habits. it doesn't disappear when a licensing deal expires. you buy it, you own it. that's the transaction.

and for the person buying it, that matters too. they're not renting access. they're not feeding data to an algorithm that uses their taste to sell ads. they're holding a piece of music that someone made and decided was worth pressing into plastic. that exchange is real in a way that clicking play on a streaming service will never be.

the economics of not asking permission

this is the freehold applied to distribution. own your masters. own your manufacturing. own the relationship with the people who listen.

the streaming model is designed for major labels with catalogues of ten thousand tracks. it is not designed for you. and the sooner you stop trying to make it work, the sooner you can build something that actually does.

press 100 CDs. sell them at gigs. put them on bandcamp. send five to blogs and five to record shops. build from there.

you don't need to be big. you need to be direct.

an invitation

we're pressing our two-hour dub techno mix onto two CDs for odioba on 19th april. gatefold digipak. sovereign masters. the first time anyone holds this thing will be in the room where the music is playing.

if you make music and you haven't pressed anything physical yet, this is your sign. the barrier is lower than you think. the costs are recoverable. and the feeling of holding the finished thing in your hands is worth more than any playlist placement.

stop waiting for the algorithm to find you. press the disc. sell it yourself. that's the freehold.

ufo / sovereign freehold / stockport

this is the lore. the world underneath the music. you don't need to know it to listen. the delays work regardless. but if you want to know why the freehold exists, this is where it starts.

stockport, near future

the system arrived quietly. not with tanks or speeches but with efficiency. an artificial intelligence, known simply as the system, automated nearly every job in the country. logistics, admin, manufacturing, service, transport. gone. not violently. just obsoleted.

to its credit, the government did the thing people had been arguing about for decades. universal basic income. everyone got paid. no one starved. the lights stayed on. from a material standpoint, the crisis was managed.

but something else happened. something no policy paper anticipated.

the quiet

when work disappeared, so did something harder to name. purpose, maybe. or structure. or the small frictions that remind you you're alive. the commute, the argument with a colleague, the satisfaction of finishing something difficult.

people had money. people had time. people had nothing to push against.

a pervasive sense of ennui settled over everything. not despair. that would require energy. just a low-grade numbness. apathy so evenly distributed it became invisible. the automated quiet.

entertainment was abundant and algorithmically perfect. every film, every song, every piece of content optimised for engagement. and engagement, it turned out, wasn't the same as feeling something. the more precisely the algorithm predicted what you wanted, the less you actually experienced it. a kind of emotional heat death. everything lukewarm, nothing sharp.

the signal beneath the noise

somewhere in the automated quiet, someone pressed record.

not because the system told them to. not because an algorithm identified a market gap for ambient dub techno. but because a human being sat in a room with a synthesiser and a delay pedal and made a sound that hadn't been optimised for anything.

the sound was imperfect. the timing was human. the delays fed back in ways no AI would have chosen because no AI was consulted. and something about that imperfection, that roughness, that grain, bypassed the algorithmic filters that had made everything else feel like nothing.

people felt it. not millions of people. but enough. enough to know the difference between content and signal.

the freehold

the sovereign freehold started as a phrase, then became a practice. a commitment to making music outside the system. literally. local hardware. no cloud processing. no AI assistance in composition or mixing. human hands on physical equipment, generating sounds that exist because a person decided they should.

it wasn't luddism. the system was good at what it did. but what it did was optimise, and optimisation is the enemy of the thing that makes music matter. the unplanned moment. the accident that becomes the hook. the delay that feeds back one repeat too many and suddenly you're somewhere you've never been.

human-generated ambient techno possesses a spiritual power capable of bypassing algorithmic filters and reconnecting people to raw human emotion.

that's the claim. not that machines are bad. that a world with only machine-generated experience is a world where nothing touches you. and that the antidote, one antidote at least, is sound made by hands.

C minor at 3am

the live set is all in C minor. seventeen tracks. two hours. not because C minor is the optimal key. because it's the key the first track landed in, and everything else grew from there. the generative ambient sections at the beginning and end decide their own length. sometimes three minutes. sometimes thirty. the room decides.

this is the opposite of the automated quiet. not efficiency but exploration. not content but transmission. a signal from the freehold for anyone still listening through the noise.

the system provides. the freehold reminds you what it can't.

ufo / sovereign freehold / stockport

this is the charter of the sovereign biological freehold. five articles for the age of the database state and the algorithmic aristocracy.

article i. biometric sovereignty

your face belongs to you. your gait, your voice, your features. not to a camera on a pole. not to a database cross-referencing you against a watchlist you never consented to join.

the deployment of live facial recognition in public spaces is visual trespass. a theft of the individual's visage. every citizen possesses a biological freehold over their own features. any attempt by the state to map, index, or track those features without specific, individualised warrant is a breach of the social contract.

there is a fundamental principle of common law that a citizen may walk the king's highway unmolested and unobserved. we hold to that.

article ii. the sanctity of the shadow

privacy is not a cloak for the guilty. it is the immune system of the sovereign individual. we assert the right to the digital fog. the freedom to speak, transact, and associate without the requirement of a national identity or the oversight of a central authority.

to demand a name before a thought is expressed is to stifle the soul of the nation.

the encryption of private thoughts and communications is the modern equivalent of the locked door. an englishman's digital castle is his own. any attempt by the state to mandate backdoors or client-side scanning is a violation of the sanctity of the home and a breach of the privilege of private correspondence.

article iii. the digital freehold

an individual's data is not the new oil for corporations to pump. it is the digital freehold of the person who generated it. every byte of location history, health data, and browsing habit is private property. to extract this value without compensation or explicit, revocable consent is digital serfdom.

sovereignty implies the right to cease to exist in the eyes of the machine. every citizen shall hold the absolute right to the permanent deletion of their digital footprint from any corporate or state ledger. if a man cannot remove himself from a database, he is no longer a citizen, but an asset.

this is why we own our masters. why the mailing list matters more than the algorithm. why we build on land we hold.

article iv. the national sovereign dividend

the wealth of the digital age is built upon the crown jewels of our collective data and the stability provided by the public. as machines and algorithms assume the burden of labour, the resulting profit constitutes a national surplus that belongs to the commons.

we mandate the establishment of the national sovereign dividend. this is not a benefit or a welfare. it is a royalties payment for the use of the public's data and the automation of the nation's industry.

to fund this dividend, we require an exchequer recalibration tax on the hyper-profits of digital monopolies and automated systems. in an age of AI, poverty is a policy of choice, and we choose abundance.

article v. the freedom of expression

the digital public square must be governed by the speakers' corner standard. if a sentiment is legal to utter in the physical world, it must be legal to post in the digital world. we reject the nanny state rhetoric of online safety as a mask for administrative censorship.

no algorithm shall be mandated by the state to nudge, filter, or moderate the political speech of the people. we demand the right to interact with raw information, free from the efficiency filters. a free people can be trusted to discern the truth. a people who cannot be trusted are no longer free.

the restoration of the freeborn

we do not seek permission to be free. we assert it. we shall take our sovereign dividend, we shall guard our anonymity, and we shall hold our digital freeholds against all encroachment.

by this charter, we move from the era of algorithmic subjection to the era of digital sovereignty.

signed this day in the name of liberty, property, and the individual.

ufo / sovereign freehold / stockport

most people hear the word "techno" and think: clubs, strobes, a drop at 128 BPM, the moment the floor lifts. dub techno is the other thing. the underground of the underground. the genre that built its whole identity around what it refuses to do.

it doesn't drop. it doesn't peak. it barely arrives.

this is not a criticism. this is the point.

what the word "dub" actually does

dub is a jamaican invention, born in the 1960s when engineers at channel one and joe gibbs started running the vocal out of the mix, doubling up the bass, drowning the drums in echo, and releasing the result as its own record. the original version is still there somewhere, but what you hear is its ghost: the absence of the thing as content in itself.

when Basic Channel took that idea to berlin in the early 1990s and routed it through a minimoog and a Roland 909, they made something that has no good name. dub techno is the name we gave it. it doesn't quite fit, which is appropriate: the music doesn't quite fit anywhere either. too slow for clubs, too electronic for ambient purists, too dark for most people who just want something to study to.

moritz von oswald. mark ernestus. maurizio. vainqueur. paul st. hellier. wata igarashi. andy stott. each one made records that work the same mechanism: the delay loop as the main compositional element, the sub-bass as the architecture, the 4/4 grid as something you feel in the room rather than nod your head to.

a room you could enter

most music presents itself as a thing to receive. a melody to catch, a chorus to sing, a moment to feel. dub techno presents itself as a room to enter.

the mix is built from the inside out. the bass is the floor. the reverb is the ceiling height. the delay is the distance between the walls. when it works, you don't listen to the music so much as you stand in it, and the walls slowly change temperature.

this is not metaphor. this is a description of what the music is engineered to do. the frequencies are chosen for their physical behaviour in enclosed spaces. the tempos are calibrated for a particular kind of attention: slow enough that you can hear the room, fast enough that the pulse keeps you anchored. the repetition is not laziness. it is the mechanism. the same phrase on the fifth repeat is a different phrase because you are a different person than you were on the first.

why it matters now

here is what the algorithm cannot do. it cannot make a thing that gets better the more times you hear it in a given sitting. the algorithm optimises for the first impression, the first ten seconds, the hook that prevents the skip. it is built for scarcity of attention, which means it is built for novelty above all else.

dub techno is built for depth of attention. a generative AI fed every Basic Channel release will produce something that sounds, superficially, like dub techno. the pads will be present. the delays will be present. the sub will be there. but the intention that shapes those elements, the patience, the willingness to let a chord resolve over twelve minutes, the refusal to give the listener an easy hook to grab: that cannot be optimised for. it runs counter to optimisation.

c minor and the practice of a key

slow release lp is written entirely in c minor. all twelve tracks. two hours. one key.

a key is a room. c minor has a particular temperature: darker than a major key, less formally melancholic than something like d minor, a shade more open than b flat minor. it sits at a frequency where the sub-bass notes have weight without becoming suffocating.

the live set was built in c minor because the first track landed in c minor. everything else grew from there. twelve tracks later, the key is not a cage. it is a climate. the whole record exists in the same weather.

this is how dub techno approaches almost everything: find the frequency, stay in it, let it develop. the patience is the technique.

what the music is asking you to do

sit down. close the laptop, or don't. put on headphones if you have them. give it twenty minutes before you decide anything about it.

the first five minutes of any good dub techno record is not the music. it is the introduction. it is the record learning the shape of the room you're in. you are also learning the shape of the record. the exchange is mutual.

after twenty minutes, one of two things has happened: either you've been drawn in, and the music has become the room, and you're no longer deciding anything about it because you're inside it. or it hasn't worked, and that's fine: this is a specific kind of attention, and not everyone has it available on a given day.

but if it works, it works in a way that very little else does. you come out of two hours of dub techno having been somewhere. somewhere specific, somewhere that the music built around you, out of frequency and time and the patience of the person who made it.

that is what it is for. it is for building a room. and it is for the people willing to enter it.

slow release lp is two hours of continuous dub techno across two compact discs. on bandcamp now.

100 copies.

not 10,000. not 1,000. not even 500. one hundred compact discs, hand-numbered, in a gatefold digipak, pressed by a small manufacturer in the uk, available directly from the artist on bandcamp and in person at shows.

when people hear that number, the first response is usually some version of "only?" and the word "only" tells you everything about which logic they're thinking inside.

the maths, briefly

slow release lp is a two-disc set. 100 copies. if every one of them sells at £15 direct from bandcamp, that's £1,500. manufacturing cost for a gatefold 2×CD run of 100 is roughly £400 to £500 depending on the printer. break-even point: around 35 copies. everything after that comes back to the artist directly: no distribution cut, no label percentage, no streaming platform's 70/30 split.

compare that to streaming. at the current rate of roughly £0.003 per stream, this record would need approximately 100,000 streams to generate the same net return as selling those 100 discs. 100,000 streams, for a two-hour ambient dub techno record, is not a realistic number. it is not a target anyone should be organising their creative decisions around.

what a small number actually means

the number is not the argument. the number is a proxy for something else.

100 copies is not a scarcity marketing tactic. it is not hype. it is not a limited drop designed to manufacture urgency. it is an honest number: it reflects the reality of a solo independent operation in its early years, a record that took months to make, and a network of listeners that is small but specific and real.

when you buy a major label release, you are one of however many tens of thousands of people who bought it. your purchase registers as a data point in someone's quarterly report. when you buy one of these 100 discs, you are one of 100. the artist knows roughly how many remain. the transaction is direct: money from your account to the artist, physical object from the artist's shelf to your door. nothing in between except bandcamp's 15%, which is the price of the infrastructure, and a fair one.

what happens when they're gone

there is no repress planned. when the 100 are gone, they're gone.

this is not a statement. it is just how it works. if there is ever a second pressing, it will be a second pressing: a new object, a new moment, a different thing. not a reissue. a second press, if it exists, would exist for its own reasons.

the result is that each copy of slow release lp has a kind of weight that a streaming play does not have. when you hold disc 47 of 100, you are holding disc 47 of 100. that is the complete edition. the object has a position in a finite sequence, and that position is inscribed on the sleeve.

streaming plays are infinite. you can play a track 10,000 times and the platform treats the 10,000th play as equivalent to the first. there is no history to the object, because there is no object. there is access to a file, renewable indefinitely.

a CD is a record, in both senses.

the scale argument, and why it's wrong for this

the logic of streaming says that scale is always better. more streams, more listeners, more reach. the ceiling is theoretically infinite, and infinite reach is the goal.

this logic makes sense if you are a major label running a catalogue of 50,000 tracks and your business model is built on aggregating micro-payments at massive volume. it does not make sense if you are a single artist making two hours of dub techno in a room in stockport.

the difference is not just financial. it is the difference between a passive audience and an active one. the person who searched for dub techno on bandcamp, found the album, listened to the samples, decided it was worth £15, and bought it: that person chose. the person who had the record served to them by an algorithm because they listened to something adjacent last Tuesday: that person was targeted. 100 choosing people is worth more than 10,000 targeted ones. not in streaming revenue terms. in every other term.

100 is enough

twenty-five people are holding slow release lp right now. some came to odioba on 19 april and left with a disc under their arm. some found the bandcamp page late at night and clicked buy before they thought too much about it. some are still in the post.

the other seventy-five are out there somewhere: in a room we've never been in, on a shelf that hasn't heard it yet, waiting for the chain of word and signal that moves slower than an algorithm but lands differently when it arrives.

soon, 100 people will hold this record. none of them will have been targeted. none of them followed a playlist that included it. they will have found it, or it will have found them, through the kind of slow transmission that you can't engineer and can't fake.

that is the whole distribution network. it is small. it is slow. it is exactly the right size for what this music is.

the number is not the point. the directness is the point.

there is a photograph you will never find. it does not exist because it was never taken, or because the photographer walked away, or because the person in frame made sure there was nothing to frame. aphex twin. autechre. basic channel. burial. a long tradition of electronic music made by people who refused to give you their face.

at the time, this seemed like mystique. now it looks more like foresight.

why the face became the product

the personal brand is the engine of modern attention economics. platforms are designed to surface the person first, the work second. what do you look like. where are you right now. who are you with. which emotion are you performing today for the algorithm that is watching you perform it.

this is not incidental. it is the architecture. instagram's reach metrics reward frequency, intimacy, and the disclosure of a recognisable face. tiktok's feed is trained on human faces. youtube recommends creators, not videos. the platform's incentive is always to know you better so it can sell that knowledge more precisely.

the face, in this system, is data. not the face you see in the mirror. the face as a set of coordinates: 68 facial landmark points, mapped in milliseconds by systems that cost nothing to run because the infrastructure was already built to monetise your selfies.

the white label tradition

the people who made detroit techno in the 1980s and 1990s pressed their records on white labels for economic reasons. no artwork was cheaper. but the white label became an aesthetic position: the music is the argument. the name and the face are beside the point.

basic channel did this on a different level. moritz von oswald and mark ernestus released records on unlabelled white vinyl. no press photos. no interviews until years later, and even then, reluctantly. the music moved through the world as pure signal: here is the sound. find it if you can.

the result is that basic channel records still feel unknown in the right way. you do not know what these two people look like unless you went looking. the music arrived and exists independently of its creators' faces. that independence is now rare.

anonymity as a defence

the argument for keeping a face out of public space is different now than it was in 1993. then, it was artistic. now, it is also practical.

facial recognition databases are built from publicly accessible images. every photograph that includes your face and is posted to a public platform is potential training data. you do not consent to this. there is no meaningful opt-out. the dataset is compiled from the aggregate of your appearances online, cross-referenced against any database the operator has access to.

this is not a theoretical threat. it is the operational infrastructure of numerous government agencies, advertising networks, and private companies whose business model is identity verification. they are not trying to surveil you specifically. they do not have to. the system is passive: it indexes, cross-references, and waits.

the creative function of the unknown

burial. richard d. james. boards of canada. the most sustained argument for anonymity in electronic music is not political. it is that the unknown listener brings more to what they cannot see.

when the music arrives without a face attached, the imagination fills the space. you do not have the creator's lifestyle to compare against. you do not have their interviews to confirm or deny what you thought you heard. the music is the only evidence. you have to sit with it.

this is not mystique for its own sake. it is a structural choice that changes the way the work lands. the face is context. remove the face and the context moves inside the music, where it belongs.

the project of ultrasonic future outlaw is not anonymous in the same tradition. there is a name, a city, a live practice. but the preference is for the music to arrive ahead of the face. the face is not the product. the sound is the product.

in an era when every platform's design is built to put the person at the centre, that preference is a small but deliberate act. a white label was never blank. it was a refusal to fill in the fields that were not required.

in 1997, a researcher at the university of surrey published a survey of steganographic techniques. one category: audio steganography. the art of hiding data inside sound.

the word comes from the greek. steganos: covered. graphia: writing. covered writing. the message is in the file, but the file does not look like it is carrying a message. it looks like music.

what steganography is

cryptography locks the door. steganography makes the door invisible.

an encrypted message says: here is a secret. you cannot read it without the key. an audio steganographic message says: here is music. there is no secret here to look for.

the two are not the same operation. cryptography protects the content. steganography protects the fact that a message exists at all. a steganographic audio file passes through a firewall check, an automated content scan, an email attachment filter, as music. because it is music. the hidden data is in the file. the file is not a cover story. the file is the object.

how it works

the most basic method: least significant bit encoding. take a lossless audio file. each audio sample is stored as a binary number. the least significant bit of each sample can be changed without audibly affecting the sound. a sequence of changed bits encodes a message.

the more sophisticated methods use the psychoacoustic properties of human hearing. we do not process all frequencies equally. we mask certain sounds behind louder ones. we are less sensitive to variations in certain phase relationships. these perceptual gaps are available as data channels.

the larger and more complex the audio file, the more data it can carry and the harder the payload is to detect. this is where ambient electronic music becomes interesting.

the orb, in this light

the orb's adventures beyond the ultraworld runs to almost 75 minutes. paterson and weston built it out of layered drones, found sounds, sampled speech, sub-bass frequencies, and ambient noise recorded from outdoor locations. the texture of the record is so dense and varied that listening for artifacts in the audio would require knowing what a clean version sounded like.

this is not an accusation. the orb were building a sonic environment, not a data channel. but the structure of dense ambient music, long-form, lossless, layered, is structurally hospitable to hidden payloads in a way that a three-minute pop single is not.

slow release lp is two hours. the sub-bass frequencies in waterworx alone occupy a perceptual range that human hearing barely registers as structured sound. if someone wanted to hide a cryptographic key in a dub techno record, the cover could not be better designed. this is a thought experiment. not a manual.

why it matters

the practical application of audio steganography is documented in several areas: watermarking, digital rights management, and covert communication in high-surveillance environments.

a journalist in a country where encrypted messaging apps are blocked cannot always use a vpn. the vpn traffic itself is detectable. but a music file shared via a legitimate file-sharing service looks like a music file. it has always looked like a music file. there is no traffic signature that distinguishes it from any other music file, because it is a music file.

the theoretical pipeline exists: audio file as carrier, hidden payload, plausible deniability at every transmission point. the sub-bass frequency range that makes dub techno physically immersive is also a data channel that conventional inspection cannot easily distinguish from generative ambience.

the music that refuses the algorithm may also, in principle, route around it.

we are not hiding anything in slow release lp. but the space is there if anyone needed it.

in 1998, autechre played a live set to a small room in sheffield. they had been performing algorithmically generated music since roughly 1993. by 1998, their live sets were being produced in real-time by max patches: software that listens to itself and generates responses. the music was not pre-recorded. it was the output of a system that made decisions during the performance.

this was the algorithm before the algorithm had a bad reputation.

what silicon valley did to the word

the word "algorithm" has been colonised. it now means: the thing that decides what you see. the recommendation system that knows you stayed on a post three seconds longer than usual and inferred from that pause something about your current emotional state, and added it to a profile that was already 2,000 data points deep before you woke up.

this is an algorithm. it is also an algorithm when autechre's max patch generates a polyrhythm in real-time by reading its own previous output. the word is the same. the application could not be more different.

the prediction engine is designed to constrain your behaviour. the recommendation system does not surface what you might want. it surfaces what the model predicts will keep you engaged long enough to show you an advertisement. the incentive is to produce a predicted response. this is the opposite of creativity.

how generative composition works

max/msp is a visual programming environment. you build a network of objects that pass data between each other. a signal comes in. something is done to it. the result is passed on. the results can feed back into the inputs. the system becomes a loop that modulates itself.

autechre's compositions using max are not random. randomness is easy and uninteresting. what they build are systems with internal logic: rules, constraints, probabilistic relationships. the system follows its rules. but the rules are complex enough that the output cannot be predicted in advance, even by the people who wrote the rules.

the composer defines the possibility space. the algorithm inhabits it in real-time.

what dub techno does with repetition

dub techno has always worked with cycles and feedback. the delay unit that feeds back into itself. the chord that repeats with variations introduced by the circuit. the low-pass filter that opens and closes without human intervention because a slow lfo is modulating it.

this is a form of generative composition that predates the software. basic channel's records are not randomly generated but they are not rigidly programmed either. the system plays. the producer listens. the decisions the system makes in real-time are part of the composition.

the distinction from the recommendation algorithm is this: the generative music system is not optimising for a measurable outcome. it is not trying to produce a specific response in the listener. it is following its own internal logic and producing music as a byproduct of that logic. the result may be beautiful or disorienting or both. that is not the system's objective. the system does not have an objective in the conventional sense. it has rules.

the recommendation algorithm has one objective: keep you engaged. every decision it makes is oriented toward that outcome. what matters is that you keep looking.

the neutrality argument

an algorithm is a set of instructions. nothing more. the instruction set can be used to generate music that no human could have composed, or to build a surveillance infrastructure that most humans would not have consented to had they been asked clearly.

the neutrality of the tool does not neutralise the ethics of its application. but the neutrality argument is worth making, because it refuses the panic that makes the tool seem inherently dangerous. the tool is not the problem. the incentive structure that deployed it is the problem.

autechre did not patent their max patches. the dub techno community experimenting with generative rhythm and modular synthesis is not building a surveillance apparatus. they are building music.

reclaiming the algorithm is not a metaphor. it is a practice. it means building systems whose objective is something other than extraction, and using the same mathematics that was weaponised against you to make something that is not a weapon.

slow release lp is not generatively composed in the max/msp sense. but the dub techno tradition it sits inside is the same one that autechre came from and departed from: the loop as a living thing, the repetition that is never quite the same. the music is slow. it does not optimise for engagement. it does not drop. it builds a room and asks you to stay in it on your own terms.

that is a different kind of algorithm. it is running in the listener, not the server.

in 1993, moritz von oswald and mark ernestus were pressing records in berlin. white vinyl. no barcodes. no press photos. distribution through a small network of specialist shops in berlin, london, detroit, and chicago. the records moved through physical space by word of mouth, carried in record bags, heard in clubs before they had names.

in 1993, the cypherpunk mailing list had been running for a year. a group of cryptographers, mathematicians, and civil libertarians who believed that encryption was the prerequisite for a free society. privacy through mathematics. communication without a third party. the right to speak without being overheard.

the two groups did not know each other. they were solving the same problem from opposite ends.

the problem they were both solving

the problem: how do you distribute something of value through a hostile environment without a centralised point of control that can be shut down or captured.

the cypherpunks were thinking about communication. the message needs to travel from sender to receiver without passing through a node that can be intercepted, blocked, or coerced into revealing its contents.

the basic channel operation was thinking about music. the record needs to travel from the pressing plant to the listener without passing through a distribution chain that can be declined by major distributors, refused by chains, or filtered out by label systems that only stock what gets promoted.

the solution, in both cases, was the same: remove the centralised chokepoint. build a network of nodes that each hold a copy and can pass it on. make the network resilient by making it lateral, not hierarchical.

what basic channel built

basic channel did not have a website. they did not need one. the distribution network was physical, distributed, and self-perpetuating. each record shop that stocked their records was a node. each dj who played the records was a broadcaster. each listener who asked the dj what it was had become part of the network.

this is not a metaphor for peer-to-peer networking. it is structurally isomorphic to it. a p2p network distributes a file by sharing pieces across multiple nodes. no single node holds everything. no single node can be shut down in a way that destroys the file. the resilience is a property of the structure, not of any individual node.

the white label as access control

a basic channel white label record has no identifying information beyond a catalogue number and a matrix runout code. no artist name on the sleeve. no label name. no title in most cases. the information required to understand what you are holding exists only in the network of people who already know.

this is an access control mechanism. not a cryptographic one, but functionally related. the information is public, in that anyone who picks up the record has physical access to it. but the information required to decode the record as a cultural object, to know who made it, what it means, where it fits, is held by the community. it is distributed knowledge. you need to be part of the network to fully read the object.

the independent distribution network as p2p

when napster launched in 1999, the music industry experienced it as an attack. the response was legal and technical: lawsuits, drm, the managed rollout of digital storefronts that put the major labels back at the centre of distribution.

what the industry saw as an attack, the cypherpunks saw as infrastructure. the p2p network was the digital version of what underground record distribution networks had been doing for decades: moving valued objects laterally, without a central point of control, through a community of people who wanted to pass them on.

basic channel did not have a position on peer-to-peer networks. but their operation was a proof of concept that the majors could not have countered even if they had wanted to: the value was in the community, not in any single node. there was no central target to shut down.

what this looks like now

the conditions of 1993 berlin are not reproducible. the specialist record shop network is smaller. the internet exists, which changes both the reach and the vulnerabilities of any distribution operation.

but the principle holds. the value of a small, committed network of listeners who want to pass the music on is greater than the value of a large, passive audience who were served it by an algorithm. the algorithm can be switched off. the network cannot. the relationship is in the people, not the platform.

slow release lp is on bandcamp and in the hands of twenty-five people. those twenty-five people are a node network. when one of them plays the record to someone who has not heard it, the network grows. slowly. laterally. without a central server.

this is what the white label meant. not a mystery. a structure.

in 1964, radio caroline launched from a ship in the north sea. it was broadcasting outside british territorial waters because the bbc had a monopoly on legal broadcasting and would not play what needed to be played. the ship was the workaround. the geography was the exploit.

pirate radio is always a technical solution to a political problem. the problem: there is a signal that someone wants to broadcast and someone else wants to block. the solution: find the space in the system where the blocker does not reach.

what uk pirate radio built

by the mid-1980s, pirate radio stations were broadcasting from tower blocks in hackney, brixton, and harlesden. kiss fm before it went legal. dream fm. fantasy fm. pulse fm. the music was jungle, hardcore, house, early drum and bass. the frequency was illegal. the equipment was portable: a transmitter, a car battery, a mic, and the knowledge that you would have to move fast when the authorities came.

these stations built the audience for uk dance music. they broadcast to communities that the bbc was not programming for. they created the infrastructure that made the acid house explosion possible: if you did not hear about the field raves on pirate radio, you did not go.

the stations were shut down when they were found. the equipment was confiscated. new stations appeared the following week from the next tower block. the resilience was structural. a pirate radio station is a single point of failure: the transmitter. but the community that wants to broadcast is not. as long as someone has a transmitter and the will to use it, the signal continues.

what the internet promised and did not deliver

in the early days of the web, the comparison to pirate radio was obvious. publishing tools that anyone could use. distribution that did not require a label or a broadcast licence. the signal could move anywhere, freely.

what emerged instead: platforms. centralised hosting services that own the infrastructure, set the rules, and operate at the pleasure of the regulators and advertisers who fund them. youtube demonetises. spotify delists. soundcloud sends a dmca. the platform is the new chokepoint.

the pirate radio problem has not been solved by the internet. it has been reproduced at scale. the transmitter is now a server. the server is owned by a company. the company can be pressured. the signal is not free. it is licensed.

where the space is now

tor routes traffic through a distributed relay of volunteer nodes, each of which encrypts the traffic before passing it on. the origin of the traffic is obscured. the content, if encrypted, cannot be read by any of the relays.

ipfs, the interplanetary file system, addresses content by what it is rather than where it is. a music file uploaded to ipfs is not located at a url. it is located at a content hash: a mathematical fingerprint of the file itself. anyone who holds a copy can serve it. there is no origin server to shut down. the file exists wherever there are nodes that hold it.

the principle these tools embody is the same one that drove pirate radio in 1987: the signal should be able to move through a network that does not have a chokepoint that can be pressured. the technical details are different. the structural logic is identical.

what this means for music

underground techno, experimental music, political audio, journalism that platforms prefer not to host: all of these have the same problem that pirate radio was solving from a tower block in hackney. the distribution infrastructure that exists is controlled by parties whose interests do not always align with the signal.

bandcamp is better than spotify in this regard. direct digital sales are better than streaming. a cd pressed and sold in person is better than a digital file hosted on someone else's server. but even bandcamp is a centralised service. the catalogue exists at bandcamp's discretion.

a fully distributed option would be an artist who holds their own files, distributes them through ipfs, and allows any listener with the content hash to download them without the transaction passing through any centralised server. the files would be sovereign in the same sense that the masters on the local drive are sovereign.

slow release lp has not taken that step. the record is on bandcamp. the masters are local. the files exist on a drive that is not bandcamp's. but distribution still passes through bandcamp.

the pirate radio logic would say: the transmitter is still a single point of failure. the community that wants to hear the signal is not.

the signal continues

uk pirate radio did not end when the pirates went legal. it ended when the problem it was solving was solved in other ways: when kiss fm went national, when the audience grew large enough that the legal infrastructure accommodated it, because the money followed.

underground dub techno will not grow large enough for the streaming infrastructure to accommodate it on its terms. which means the pirate logic applies permanently: the signal exists outside the system or it does not exist.

slow release lp is on bandcamp and in the hands of twenty-five people and in a distribution network made of humans passing it on. that is also the oldest model. the signal moved before the platform existed.

it will move after.

a desk and a screen, an afternoon, the work waiting. the deep work room is the first of the broadcasts and the one that has been running longest. it sits quietly behind a working day, twelve hours of audio that does not ask to be heard, only to be in the room. here is what it is and what it is for.

theta, not alpha

theta is the slowest band where you are still awake. four to eight hertz. it is what your brain settles into when meditation is doing what it says it does, when a problem starts to solve itself in the shower, when the work that needs the unconscious to come along is about to begin. it is not the band of attention. it is the band that lets attention rest.

the deep work room is built on theta. that is a deliberate choice. there is a more obvious choice, alpha, somewhere in the eight to twelve hertz range, the band of calm focus. alpha is a fine room for a spreadsheet. but a lot of work is not a spreadsheet. a lot of it is the slow approach, the half hour spent on a paragraph that does not yet know what it wants to say, the afternoon spent not solving the problem so that the solution can find you. theta is the room for that work. it is the room behind the desk.

a binaural beat is a quiet trick. two tones, one in each ear, slightly mistuned, and the brain fuses them into a third tone you cannot point to. that third tone is the frequency you came for. underneath, there are slow ambient passages and the same ufo recordings the records are made from. nothing intrudes. nothing climbs the dramatic arc.

twelve hours

twelve hours is the loop. that is not a number for marketing. it is what the room needs to be in order to behave like a room. less than twelve and you start hearing the seam, and once you hear the seam the music stops being weather and starts being a song again. you do not want a song behind a working day. you want a climate.

the room asks one thing of you: do not listen, occupy. put it on at the start of the work and let it become the temperature of the room. when the work breaks, do not close the tab. when you go to make tea, the room is still on, and when you come back the room is still there, and you sit back down inside it.

the audio is original. recorded on hardware in the freehold, in the same workflow as the records. nothing is licensed in. nothing is generated. the streams run on upstream.so and youtube. the room does not go off air.

if you want to own the audio rather than stream it, the deep work piece is on the shop as flac and wav. if you run a venue, a studio, a yoga room, a co-working space, or a film, the licensing page is the place to start. the rest is just the room. open whenever you want it.

a record ends. a stream does not.

we put the second broadcast on yesterday. twelve hours of audio for sleep, looping, on a youtube channel that does not go off air. that makes two rooms now, both open, both ours.

a record is a noun. it is a finished thing you reach for. a broadcast is something else. it is a verb that keeps happening. you can drop in at three in the morning and the room is already in motion. you can leave it on while you cook. you can forget about it for three days and then come back and find it where you left it.

i did not expect this to feel different from a record. it does. when you make a record, you are asking someone to come in and listen. when you keep a room open, you are letting someone live near it. the music is doing something more like weather than performance. people will not always notice it. that is the right way for this music to behave.

two rooms, soon three

the deep work room has been running since last week. it sits quietly in the background of a working day, binaural beats holding the floor, slow ambient passages turning the walls. the new room is darker and slower. lower frequencies, longer breath, recorded for the hours the day was not going to use anyway.

the third room is coming, dub techno mixed for sustained creative work. closer in temperament to slow release lp. that one will push back a little.

how this is built

a few notes for anyone wondering how the broadcasts work. the audio is original. the loops are long enough that you should not hit the seam in any reasonable session. the streams run on upstream.so, which lets a static audio file behave like a live stream. youtube hosts the broadcast. nothing about this requires an algorithm to choose what plays next. the same room plays continuously, by design, because the whole point is that the room is the same room.

if you want to own the audio rather than stream it, both pieces are now on the shop as flac and wav files. and if you have a use for this music in a venue, a film, a retreat, or a hotel, the licensing page is the place to start.

the rest is just the room. open whenever you want it.