A lo-fi or noise record translates on streaming when three things are true: the vocal and the snare survive a phone speaker, the low end still has a shape in mono, and the loudest moment sits around minus 14 LUFS without a limiter flattening the noise into a hiss. That is the whole trick, and none of it requires cleaning the record up. At Tapetown we mix post-punk, garage, noise and shoegaze records that arrive deliberately dirty, from cassette four-tracks to blown-out practice-room recordings, and the job is never to remove the dirt. It is to decide which dirt is the song and build the mix so that dirt is heard the same way on a laptop, in earbuds and on a club system.
What "translate" means in 2026
Most people will hear your record on a phone, in earbuds, or through a Bluetooth speaker, and every streaming service turns the volume of each track to a reference level before it plays. Spotify plays at about minus 14 LUFS by default, with Loud and Quiet settings at minus 11 and minus 19, and it asks for one stereo master per track with true peaks under minus 1 dB. Apple Music's Sound Check works to a similar reference, around minus 16 LUFS, and YouTube sits near minus 14. The practical meaning: a noise record mastered to be the loudest thing on the internet gets turned down to the same level as a folk record, and what is left is the balance. A lo-fi record that was balanced for a laptop speaker in the bedroom often has no low end and a vocal buried under tape hiss, and on a phone the hiss is the loudest element. Translation is making the balance survive the move.
The midrange is the record
A phone speaker reproduces very little below about 200 Hz and not much above 10 kHz, so a lo-fi record is judged almost entirely on what happens between 200 Hz and 5 kHz: the vocal, the snare crack, the guitar's body, the bass's harmonics. We mix noise and lo-fi records on small speakers first, at low volume, and the question is always whether the song reads. If the vocal is meant to be buried, it still has to be locatable; a vocal you can follow at a whisper on a small speaker will be buried in the right way on a big one. We usually find the vocal's intelligibility band, often 1.5 to 3 kHz, and make sure the guitars leave it alone, not by carving a hole with EQ but by deciding which guitar track owns that region and turning the others down there.
Keep the noise, move its centre of gravity
Tape hiss, amp hum, room noise and the crackle of a bad cable are part of the recording's identity, but their energy is mostly in the top octave and in the sub, where streaming playback does least good with it. We keep the noise and shift where it lives: a gentle shelf that pulls the noise floor down above 8 kHz on the tracks that carry it, a high-pass on the noisiest sources below 60 or 80 Hz so the hum does not eat headroom, and then the noise stays audible as texture in the midrange, where the phone can reproduce it. The record sounds as dirty as before. It just stops being quieter than everything else after normalisation, because the limiter no longer has to work on hiss.
Mono first, then width
A lot of lo-fi character comes from stereo tricks: wide chorus, pseudo-stereo from a doubled cassette, reverb that lives only on the sides. Phones, Bluetooth speakers and most clubs play in mono or near it, and anything that lives only in the side channel disappears or, worse, cancels. We check every noise and lo-fi mix in mono and make sure the song is complete there: kick, bass, vocal and snare in the centre, the width added on top as a bonus. The widest records on the credits list, the shoegaze and dream pop sessions, were built this way; the walls of guitar are stereo, but the record stands up when you sum it.
Low end has to have a shape, not a size
Lo-fi records tend to have either no low end or an undefined boom from a bass recorded through a small amp in a room. Neither translates. We do not add a clean sub to a dirty record; we find the bass's note in the harmonics, around 100 to 250 Hz, and make sure it is clear there, which is where earbuds and phones hear bass anyway. Below that, a gentle, consistent low end that the kick and bass share is enough, and on vinyl it keeps the cut clean too. If the bass was recorded badly, a parallel distortion that adds harmonics does more than EQ: it gives the phone speaker something to reproduce.
Loudness: let the dynamics do the work
Noise records are often mastered into a wall because the band wants them to feel violent. Normalisation undoes that: a master at minus 8 LUFS is turned down six decibels on Spotify and ends up quieter in impact than a master with real peaks. We mix for a master around minus 12 to minus 14 LUFS integrated, with true peaks under minus 1 dB, and we put the violence in the transients and the arrangement, not the limiter. The loudest moment of a noise rock song should be the moment the band hits, not every moment. If the mastering engineer has room to work, the record ends up louder in effect at the same measured level. Our thoughts on the choice of mastering engineer for this kind of record are in the AI mastering versus a mastering engineer post.
What we ask for before we start
The rough mix you love, even if it is a bounce from a cassette. The individual tracks if they exist; if the recording is a stereo file from a four-track, say so, and we mix from that. Two or three reference records and a sentence on what each is for. And a clear statement of what must not be cleaned: the hiss on the vocal, the pop on the kick, the feedback at the end. We write that list down and work around it.
What it costs
Professional mixing in Europe is commonly quoted at roughly 400 to 1,200 euro per song. Tapetown quotes per project, because the scope, the number of songs and any studio days change the price, and a lo-fi record mixed from a handful of tracks is a different job from a 40-track noise rock album. Send the rough mix through the contact page and we come back with a number and a plan.
FAQ
Does mixing for streaming make a lo-fi record clean?
No. Translation is about balance, not cleanliness. The hiss, hum and distortion stay; we move their energy into the range phones and earbuds can reproduce and make sure the vocal and snare stay audible at low volume.
What loudness should a noise record be mastered at?
Spotify plays at about minus 14 LUFS by default and asks for true peaks under minus 1 dB; Apple Music's Sound Check works around minus 16 LUFS. A master around minus 12 to minus 14 LUFS with real peaks keeps the impact after normalisation; a master crushed to minus 8 is simply turned down.
Can a cassette four-track recording be mixed professionally?
Yes. We mix from stems when they exist and from the stereo file when they do not. A four-track recording can be balanced, given a low end that works in mono and a vocal that reads, without losing what made it a cassette recording.
Why does my lo-fi mix sound fine on my laptop but thin on a phone?
Laptop speakers already lack low end, so a mix balanced on them is often tuned to compensate, and a phone, with even less range and in mono, exposes it. Mixing on small speakers in mono at low volume, then checking on good monitors, catches it.
If your record is deliberately dirty and you want it to stay that way on every speaker, start with online mixing at Tapetown and send us the rough mix.
