Signals, by ear.

01 · sampling

Sound becomes numbers.

Sound is a wave in the air. It's smooth and continuous. Computers can't store smooth things, so they do what a movie camera does: take snapshots. A film takes 24 pictures a second and your eye sees motion. Digital audio takes 44,100 snapshots a second and your ear hears music. That number is the sample rate.

Bit depth is the other half. It's how precisely each snapshot gets measured. Think of a ruler marked in centimeters versus one marked in millimeters. More bits means a finer ruler, which means cleaner sound.

There's one big rule here: you can only capture pitches up to half your snapshot speed. Go past that and things get weird. You know how wagon wheels in old westerns look like they're spinning backwards? That happens because the camera can't keep up. Sound does the exact same thing. Pitches that are too fast fold back down into false tones that were never in the music. That defect is called aliasing, and you can hear it below.

音In Oto: your library is thousands of these snapshot streams, recorded at different rates. Before Oto can blend two tracks, it puts them on one clock. Carefully, so no wagon wheels wake up.

02 · spectrum

Every sound is a recipe.

A violin and a flute play the same note. Same pitch, totally different sound. Why? Because a note isn't one ingredient. It's a stack of frequencies: one main tone, plus a bunch of quieter ones above it. Every instrument stacks them differently. That stack is the sound's recipe, and the recipe is what we call timbre.

The math that reads the recipe out of any sound is called the Fourier transform. In code, it's the FFT. Those bouncing bars on an old stereo? That's an FFT. And here's a fun one: your inner ear does the same job with a spiral of tiny hairs, each tuned to its own frequency. You've been doing spectrum analysis your whole life.

音In Oto: when Oto reads your library, this is the reading. Tempo, key, energy, brightness. All of it comes out of the spectrum. It's how Oto knows two songs will get along before either one plays.

03 · filters

Sculpting the recipe.

Stand outside a club. Through the wall, all you hear is bass. That wall is a low-pass filter: low frequencies pass through, highs get blocked. A phone speaker is the opposite, all sizzle and no weight. That's the whole idea of a filter. It turns some frequencies down and lets others through. An EQ is just a row of gentle filters with knobs.

One extra ingredient makes filters fun: resonance, a boost right at the cutoff point. Sweep a resonant filter up and you get that rising squelch that has opened every dance record since disco. It's waiting for you below.

音In Oto: say "bring it in softer" and a filter is what moves. Blends that hand the bass from one track to the next. Intros that arrive from behind the wall. In Oto, plain words become filter moves.

04 · dynamics

Taming loud and quiet.

Ever notice a radio host's voice never gets quieter? Lean in, lean back, same loudness. That's a compressor at work. Think of it as a tiny listener with its hand on the volume knob. The moment the sound crosses a line called the threshold, it pulls the volume down. It's why a whisper and a chorus can share one pop song and you hear both.

Push a compressor hard and it stops being invisible. It grabs. It breathes. That's the pumping in dance music: the whole track ducking every time the kick hits.

And a compressor with zero patience is a limiter. A hard ceiling nothing gets past. One stands guard at the end of nearly every mix on earth.

音In Oto: a limiter guards the master at all times. Ask for anything mid-set (louder, now, everything) and the room gets energy, never distortion. It's why "take the room" can't end in a blown speaker.

05 · delay

Sound, remembered.

Shout into a canyon and it answers. Delay is a canyon you can carry: keep a copy of the sound and play it back a beat later. That's it. Every echo, chorus, flanger, and reverb ever made descends from this one move.

Feed a little of the echo back into itself and it repeats, quieter each time, like the canyon answering the canyon. That's feedback. Musical delays lock the echo time to the tempo, so the repeats land right on the beat.

Feedback also has a wild side you already know. Point a microphone at its own speaker and it screams. That scream is a feedback loop running away. The one below has a safety leash, but you can walk it right up to the edge.

音In Oto: the graceful exit. When a track leaves your set, it can dissolve into tempo-locked echoes instead of just stopping. And catch me leans on exactly this to land a lost moment softly.

06 · pitch + time

The hard problem.

Play a record too fast and the singer turns into a chipmunk. Slow it down and they turn into a ghost. On vinyl and tape, speed and pitch are welded together. Change one and you drag the other. Basic digital playback has the exact same flaw, and you can hear it below.

Cutting that weld is called time-stretching: changing the tempo while the pitch stays put, or the pitch while the tempo stays put. To do it, software takes the sound apart into its recipe, adjusts it, and rebuilds it. Doing that without smears or robot noises is one of the genuinely hard problems in audio.

It hides inside every DJ sync button, every slowed-and-reverb edit, and Auto-Tune. When two tracks lock to one beat, this is the machinery underneath.

音In Oto: your library wasn't recorded at one tempo, but your night runs at one pulse. Time-stretching is how a 122 and a 126 meet in the middle with nobody turning into a chipmunk. It's also why Oto's drafts warn you when two tracks are too far apart to meet honestly.

07 · the mix

Two songs become one.

Here's the secret at the bottom of all mixing: it's addition. Two signals become one by adding their numbers together, sample by sample. That's the entire trick. It just has two consequences worth knowing.

First, peaks stack. Two full tracks added together can be louder than either one alone. So a limiter watches the door.

Second, and this one's lovely: fading between tracks hides a little trigonometry. Loudness follows power, which is amplitude squared. So a straight-line crossfade actually dips in the middle, like a handshake where both people let go early. The fix is the equal-power crossfade. One side follows a cosine curve, the other a sine. And since cos² + sin² always equals exactly 1, the energy in the room never dips. There's a quarter circle hiding inside every good fader.

音In Oto: every blend Oto draws on the canvas ends here. Two signals, one sum, cosine against sine, a limiter standing guard. Same math whether the fader is your hand or your sentence.

08 · how Oto uses this

Seven blocks. One night.

Everything you just heard is Oto's working vocabulary. When you tell it what tonight should feel like, your words get translated into exactly these seven blocks. Nothing more exotic than what's on this page.

"Read my library" is spectrum work. Oto listens to every track the way chapter 02 listens to a sawtooth: tempo, key, energy, brightness. That analysis is what the drafting is built on. It's also the only thing connected accounts ever share. A taste signal, never audio.

"Draft the night" is pitch and time. Two tracks can only share a pulse if time-stretching can carry them there cleanly. So the draft doesn't just order songs. It checks that every blend is physically honest, and tells you when one isn't.

"Change anything" is filters and the mix. Softer here. Longer blend. Hand the bass over later. Each of those sentences lands as a filter move, a fade curve, or an EQ handoff, scheduled to the phrase.

"Take the room" is dynamics and delay. Raise the energy and compression leans in. Lose the thread and catch me dissolves the moment into tempo-locked echoes and brings it home. And under all of it, always, the limiter. The worst thing that can happen to your night is a lesson, never a blown speaker.

That's the honest shape of it. No black box, no magic. Just a signal chain you can now hear all the way through. Your music, your calls, and seven old ideas doing exactly what you ask.