How to Make Lofi Songs with Soft Vocals

Quick answer: To make lofi with vocals, choose the soft vocal setting (not lead), describe the voice as hushed, breathy and low in the mix, and keep lyrics short: four to eight syllables per line, with lines repeated. Start with a wordless hum or one repeated phrase; only add full verses for tracks meant for relaxing, not studying.

This is for people who like lofi but want a little more humanity in it: a voice drifting in over the Rhodes, a phrase that comes back every eight bars, a sleepy duet. It is also for anyone who has tried adding vocals and found that the singer suddenly dominated the track and turned it into a pop song.

You will learn which vocal style suits which use, how to write lyrics that stay in the background, and the exact settings and prompt words that keep a voice soft in the lofi music generator, which starts with vocals off by default.

When a vocal helps and when it hurts

Instrumental lofi is designed to disappear. A vocal pulls a track back into the foreground, which is good in some situations and bad in others.

If your goal is a focus playlist, the plainer route is better, and lofi beats for studying covers it. Vocal lofi is the right choice when the listener has some attention to spare.

Three vocal styles, from lightest to heaviest

1. The wordless hum or "ooh"

A voice with no words, used as another instrument. It adds warmth without asking the listener to follow language. This is the safest place to start.

How to ask for it: set vocals to soft, leave the lyrics empty or use a single line of humming syllables, and describe it: "hushed female voice humming the melody, no words."

2. The repeated phrase

One short line, three to seven words, that returns every few bars. Think of it as a hook that never becomes a chorus. It gives the track an identity without a story.

How to ask for it: paste only that line with a structure tag, and describe the voice as distant.

3. The sparse verse and hook

A short verse of four lines and a two-line hook, with plenty of instrumental space around them. This is closest to a song and needs the most care to stay lofi.

The scale is a relative guide, not a measurement. The point is that each step up adds noticeably more pull, and a full pop vocal is nearly the opposite of what lofi is for.

Settings that keep the voice soft

MusicVivo offers three vocal levels: none, soft and lead, with a choice of male or female voice and 10 vocal languages. For lofi, soft is almost always right.

In the prompt, these words help push the voice back: hushed, breathy, whispered, distant, airy, low in the mix, behind the piano, half-sung, sleepy. These push it forward and should be avoided: powerful, belting, soaring, emotional, lead, front and center.

Writing lyrics that stay in the background

If MusicVivo writes lyrics for you when none are given, it will happily produce full verses. For lofi, writing your own short lyrics gives you much more control. The rules are simple.

Few syllables per line. Four to eight. Long lines force the singer to move quickly, which sounds busy.

Open vowels. Words like "low", "stay", "slow", "night", "home" can be stretched and sung softly. Clusters of consonants ("strengths", "crisp sprint") get choppy.

Images, not plots. A cup going cold, a streetlight blinking, a coat on the chair. Lines that set a scene, not lines that ask the listener to follow a story.

Repeat instead of adding. If you have a line you like, sing it twice. Lofi rewards repetition.

Leave space. Use instrumental sections between vocal parts so the voice comes and goes. A good ratio is roughly one bar of voice to one bar of music or more.

Before and after

Here is a typical first draft and a lofi rewrite.

Before (too dense): "I've been thinking about the way you used to call me late at night and tell me everything was going to be alright / now I'm staring at my phone and wondering why I never told you how I felt before you went away"

Each line is over 20 syllables. It tells a story the listener has to follow. A soft voice cannot fit it without rushing.

After (lofi):

Every line is five or six syllables. The verse is four images. The hook is one phrase sung twice. Between the parts, the music carries the track. Structure tags like [Verse], [Hook] and [Instrumental break] tell the generator where to sing and where to stay quiet. For the general rules on tags, see formatting lyrics for AI.

Three more lyric sketches to adapt

These are short, original sketches you can paste and change. Each one is built for a different setting, and each keeps lines short and repeats on purpose.

Morning sketch (repeated phrase, major key, 82 BPM). Use one line only: "Slow light, slow start." Sing it four times across the track with instrumental breaks between. The two stressed words, "light" and "start", fall naturally on the beat, and the repeated "slow" gives the singer an easy open vowel to stretch.

Night drive sketch (verse, minor key, 74 BPM). Four lines: "Red lights in the rain / Radio down low / Nowhere I need to be / Nowhere I need to go." The last two lines repeat the same opening, which gives the verse shape without adding new ideas. Follow it with a long instrumental section before the verse returns.

Kitchen sketch (duet, 80 BPM). Voice one: "Kettle on." Voice two: "Stay in." Both: "Sunday, Sunday." Three tiny lines, traded back and forth. It works because each part is only two or three syllables, so neither voice ever takes over.

A useful habit is to read your lyric aloud at a whisper before generating. If you run out of breath or stumble, the line is too long for a soft lofi delivery. Cut it down until you can whisper it comfortably in one breath, and the generator will usually sing it gently too.

Worked example: hum to phrase to verse

Let us build the same track three ways so you can hear the difference in what each step adds.

Version A: hum. Lyrics field: "Mmm, ooh, mmm." Prompt: "Rainy bus window at night, warm Rhodes, soft round bass, lazy swung drums, faint tape hiss, hushed breathy female voice humming without words, low in the mix, 78 BPM." Result type: a warm, almost instrumental track where the voice feels like a pad.

Version B: phrase. Lyrics field: "[Verse] Stay a little while" repeated three times with "[Instrumental break]" between. Same prompt, with "humming without words" replaced by "softly repeating one phrase". Result type: the phrase becomes the track's identity, and it still sits behind the piano.

Version C: verse and hook. Lyrics: the full sheet above. Same prompt, with "half-sung, sleepy" added. Result type: a short lofi song. Best for late-night listening.

Generate all three and you will hear clearly where your preference lies. Most people find A or B works for daily listening and C for evenings.

Duets, languages and male voices

Duets. Two soft voices trading lines can sound lovely in lofi. Mark each line in the lyrics with who sings, or describe it: "soft male and female voices trading lines quietly". Keep each part very short.

Other languages. Lofi with French, Japanese, Spanish or Korean vocals is popular, partly because listeners who do not speak the language hear the voice as pure texture. That makes non-native-language vocals easier to work to. MusicVivo sings in 10 vocal languages; if you want to write in another language, the tips in AI song in another language apply here too.

Male voices. A low, close male voice, half-spoken, works well for night-time lofi. Ask for "low warm male voice, half-spoken, close and quiet".

Mixing problems and fixes

The singer is too loud. First cut words. A shorter lyric gives the voice less to perform. Then add "low in the mix behind the piano". If it is still loud, check that vocals are set to soft, not lead.

It sounds like a pop ballad. Energy is too high or the lyric is too emotional. Lower energy, replace emotional words with images, and add "lazy swung drums" so the groove pulls it back to lofi.

The voice is muffled into mush. You stacked too many softening words with heavy texture. Remove "whispered" or reduce the texture to "faint tape hiss".

Words are unclear. Some blur is part of the style. If you want clarity, use fewer syllables per line and simpler words, and drop "whispered", which tends to soften consonants until words run together.

Where vocal lofi fits in a bigger plan

If you run a channel or stream, vocal lofi works well as a contrast: a few vocal tracks on break screens, instrumental tracks during talking. For channel planning, see how to start a lofi music YouTube channel. For lofi's more upbeat relatives, where vocals and samples are more common, read chillhop and jazzhop. And if you want to understand how AI vocals are chosen in general, the guide to choosing vocals in AI songs goes deeper into voice types.

People also ask

Can lofi music have vocals? Yes. Vocal lofi is common, usually with a soft, breathy voice low in the mix, singing a hum, a short phrase or a sparse verse.

How do I keep AI vocals quiet in a lofi track? Use the soft vocal setting, describe the voice as hushed or distant, and keep lyrics short. Fewer words keep the voice back.

What lyrics work for lofi songs? Short, image-based lines with open vowels, four to eight syllables each, with repetition instead of new lines.

Should study lofi have vocals? For reading and writing, usually not. Vocal lofi suits relaxing, cooking, walking and late-night listening better.

Related reading

• How to make lofi music with AI
• Lofi chord progressions and how to prompt them
• Lofi music for sleep
• Choosing vocals in AI songs
• Formatting lyrics for AI
• Write a chorus that sticks

Product and company names mentioned in this article are trademarks of their respective owners. MusicVivo is an independent product and is not affiliated with, sponsored or endorsed by any of them. Comparisons are based on publicly available information at the time of writing; check each provider's own site for current features, pricing and terms.