Personalized music: how mood, genre, era and language shape your station

“Personalized music” often gets treated as one feature — usually whichever one a product is best known for. On FeelFlow Radio that’s the heart-rate integration, so it’s easy to assume that’s the whole story. It isn’t. The heart-rate reading is one input into a larger set of choices — mood, genre, era, and language — that all shape the station together, and most listeners will use several of them without ever putting on a heart-rate monitor at all. This is about those choices: what they are, how they combine, and why the combination matters more than any single one of them.

The mood scale, with or without a heart-rate monitor

How heart-rate-based music actually works covers the mechanism behind converting a BPM reading into a number on a -100 to 100 activation scale. What’s less obvious is that this scale isn’t exclusive to the heart-rate feature — it’s the one scale the whole system runs on, and a manually chosen mood lands on exactly the same line.

FeelFlow Radio offers 16 moods to pick from directly, running from Sleepy (-60, right before sleep takes you) through Relaxed, Focused, and Curious, up to Restless, Elated, and Euphoric (85, maximum intensity). Picking “Focused” from that list and having a heart-rate monitor read a comparable state produce the same input to the music engine — the system doesn’t know or care which one you used. That matters practically: if you switch from wearing a heart-rate monitor to choosing manually, the station doesn’t reset to some default — it starts from the mood closest to your last measured reading, so the transition is continuous rather than jarring. It’s also the reason a mood-based approach like this holds up better than a fixed mood playlist: the station is generated fresh from wherever your mood actually sits, not picked from a finite pre-made list.

Genre: fewer choices, on purpose

Genre selection looks like the obvious personalization lever, and it is — but the constraint on it is less obvious. You choose between one and three genres from 21 available (pop, jazz, ambient, techno, Middle Eastern, and 16 others), and the cap at three isn’t an arbitrary UI limit.

Every genre gets combined with your mood, language, and any seasonal or era settings you’ve turned on. Add a fourth genre and the number of distinct combinations the system has to generate for climbs fast enough that none of them collect enough tracks to actually get reused — instead of one station with continuity, you’d get one that jumps between styles without settling into any of them. Capping it at three keeps each combination populated enough to feel like an actual station rather than a slideshow of unrelated tracks. And the floor is deliberately just as unusual: one genre is enough. A listener who only wants lo-fi doesn’t get pushed into picking two genres they don’t actually want to hear — a narrow station they like beats a broad one they didn’t ask for.

Two more dials: decade and seasonal layers

Beyond genre and mood, two more settings shape the sound without changing what genre it’s officially in. You can restrict a station to specific decades — anywhere from ’70s analog warmth to 2020s contemporary production — or leave it on “all” for full variety. And separately, seasonal and effect layers can be switched on: a Christmas setting that shapes both the sound and the lyrics of every track toward the holiday, or an Autotune effect that can be layered onto any genre rather than existing as its own genre. That second point is a real distinction in how the system is built — autotune used to be listed as a genre, which meant choosing it meant giving up pop or rock; as an effect layered on top, you can have autotune-pop or autotune-techno instead of losing the genre to get the effect.

Language, including the parts most apps don’t offer

FeelFlow Radio supports 17 languages, including an instrumental-only option for listeners who’d rather skip vocals entirely. But language personalization goes a step further than picking one: you set a primary and a secondary language, plus a slider for how much of the mix leans toward your primary (from fully primary down to an even 50/50 split), and — separately — whether the two languages are allowed to appear mixed within the same track. That mixing option is intentionally rare when it’s on (roughly a 1-in-7 chance per track, not a coin flip), because code-switching within a single song works naturally in some genres and sounds forced in others. Turning it on adds occasional variety; it isn’t the default behavior even once enabled.

How all of it fits together without repeating itself

Here’s the part that makes the personalization actually mean something rather than just being a settings screen: every track generated is tagged internally by the exact combination of genre, mood, decade, language, mixing choice, and a few other factors that a listener can actually hear the difference between. That’s the mechanism what AI-generated music actually is describes at the database level — no listener is ever served a track they’ve already heard, but if two different listeners land on the exact same combination of choices, a freshly generated track can be shared between them the first time it’s created, so the generation cost is paid once rather than per listener. Personalization and non-repetition aren’t in tension here; the same system that keeps your station matching your choices is what makes it never loop.

Why this is also the business setup, minus one input

Everything above except the live heart-rate reading is exactly what a business account configures too — a shop or café picks genres, a language, and a fixed mood the same way a personal listener does, just without wiring up a heart-rate monitor to a shared space (a consistent atmosphere across a room makes more sense than one that shifts with one person’s pulse). That’s also most of why setup stays quick: it’s the same handful of choices either way, not a separate configuration path for commercial accounts.

The rest of business setup is even shorter — sign in with Google, look up the business, and confirm the floor area, and pricing is calculated automatically from that one number: 46,78 €/month for up to 240 m², plus 0,134 € per m² above that, no separate licensing call to make. For what that actually saves against a traditional performing-rights licence and a separate music service, see AI music vs. traditional background music: what it actually costs a shop. For the personal side, Standard is 13,27 €/month for 5 hours a day, and Unlimited is 26,68 €/month for up to 18 hours — both include every genre, mood, era, and language setting described above, not a limited subset gated behind a higher tier.

If you’re setting up for a business, the business page has the full pricing walkthrough and FAQ. If you’re setting up to listen personally, the mood and genre choices above are exactly what you’ll see on signup — there’s no separate “advanced settings” screen hiding any of it.