Your loudest moment sets the level for the whole record.
On vinyl, every part of the record has to be cut at a level the loudest peak can survive. One clipped snare, one hot kick, one over-compressed chorus — and the entire side cuts quieter. Here's why, and what to do about it.
How vinyl loudness actually works.
When audio is cut to vinyl, a stylus physically carves a groove into the disc. The louder the signal, the wider and deeper that groove needs to be — because the stylus has to move further laterally with each cycle of the waveform. A quiet passage has a narrow groove. A loud peak has a wide groove.
The problem is that grooves have to sit next to each other without touching. If the groove from one revolution is too wide, it collides with the adjacent groove — a groove collision — which sounds like a loud crack and can damage the stylus on playback.
To prevent this, the cutting level has to be set so that the single loudest moment in the entire recording stays within safe limits. Everything else on the record — every quiet section, every verse, every breakdown — gets cut at that same level or below. The loudest peak is the ceiling. It doesn't matter if that peak is one snare hit in bar 64. The whole record comes down to accommodate it.
Think of it like setting the volume on a PA before a gig. You can only turn it up as loud as the loudest moment you're expecting — if the headline act hits a peak that overloads the system, the whole show sounds distorted. You set headroom for the worst case. Vinyl cutting is the same. The cutter sets a level the loudest peak can survive, and the rest of the recording sits inside that.
The loudest instrument dictates the cut
This isn't just about overall level — it applies to individual elements in the mix. Whichever instrument or sound is the loudest in the mix is effectively setting the volume for the entire record. Everything else has to come down to the level that element can be safely cut at.
A few common examples:
Hot hi-hats. If your hi-hats are sitting 3–4 dB louder than the rest of the mix — maybe from heavy compression or a bright sample — the entire record cuts at the level those hi-hats require. Your kick, your bass, your vocal, your pads — all come down with them. The hi-hats are not the loudest thing you want the listener to focus on, but they're the loudest thing the lathe has to deal with.
A single clicky snare transient. One particularly sharp snare hit — especially if it has a very fast attack and high-frequency click — can be significantly louder than everything else for a few milliseconds. That single hit sets the ceiling for the whole side. Pull that transient down and the entire record can go up.
A resonant synth peak. Synthesis can create narrow-band resonances — a filter sweep that spikes at a specific frequency, a pad with a sharp formant — that measure louder than the intended loudest element. These are easy to miss when listening but show up immediately on a meter and on the lathe.
A vocal that spikes on one word. Even a well-mixed vocal can have one phrase or consonant that sits significantly hotter than the rest. A hard 'T' or 'K', a note that catches the top of a resonant peak in the room — if it's the loudest thing in the mix, it sets the level for everything.
The fix in all these cases is the same: identify what's loudest and address it. Sometimes that's a few dB of gain reduction on one element. Sometimes it's a short-term limiter on a specific track. Sometimes it's just a level adjustment in the session. A well-balanced mix where no single element is significantly louder than the others will always cut louder than one with hot outliers.
This is why a well-produced, dynamic mix will often sound louder on vinyl than a heavily limited streaming master — even if the streaming master measures louder on a meter. The limited master has used up all its headroom constantly maintaining high amplitude. A dynamic mix hits its peaks hard, which the cutter can encode at full level, while the quieter sections give the groove room to breathe.
The things that force the whole record down.
Several specific things in a mix cause the cutter to pull the level down to protect the groove. Understanding these helps you avoid them — or at least know what trade-offs you're making.
1. Clipping and hard limiting
A brickwalled master — one that's been pushed through a limiter until the transients are completely flattened — is one of the hardest things to cut loud. The issue isn't just the level, it's what limiting does to the waveform. Hard clipping creates high-frequency distortion artifacts that aren't audible on a phone speaker but become very obvious on a stylus tracking a groove. The cutting head has to work harder to reproduce them, the HF content pushes against the physical limits of the groove geometry, and the result is either distortion on playback or a reduced overall level to compensate.
A record cut from a heavily limited master will often sound harsh, flat and lacking in punch — not because vinyl can't handle loudness, but because the transients that give music its impact have been removed before it even reaches the lathe.
2. Wide stereo bass
Vinyl is cut using a 45/45 system — the left channel is encoded in one groove wall, the right in the other. Stereo information creates lateral (side-to-side) groove movement. Bass frequencies that are wide in stereo — a bass guitar with heavy stereo widening, a kick drum with a wide reverb tail, a sub synth panned hard — create massive lateral movement at low frequencies. This takes up enormous groove space and, at high levels, causes the groove walls to collapse inward.
Bass below roughly 120 Hz should be in mono. This converts low-frequency stereo content to vertical groove movement, which takes up far less space. It doesn't affect how the bass sounds on playback through a decent system — mono bass below 120 Hz is inaudible as stereo anyway — but it gives the cutter significantly more headroom for everything else.
3. Extreme sub frequencies
Content below around 30 Hz is inaudible on almost every playback system — even large club systems don't reproduce much below 35 Hz. But a lathe cuts it anyway, because the cutting chain doesn't know what you can or can't hear. Sub-30 Hz content takes up groove space for no audible benefit. A high-pass filter at 30 Hz before sending your master removes content you can't hear and gives the cutter more room to work with.
4. High-frequency harshness and sibilance
Excessive high-frequency content — harsh cymbals, sibilant vocals, fizzy synth textures — creates very rapid lateral groove movement at high frequencies. The cutting head has resonant characteristics around certain frequencies (typically above 10 kHz) and excessive HF energy in those ranges creates distortion. A de-esser or gentle HF shelf reduction above 12–14 kHz often makes a significant difference to how a record sounds and how loud it can be cut.
5. No headroom — peaks at 0 dBFS
If your master is peaking at 0 dBFS or clipping, there is literally nowhere for the cutting chain to go on peaks. The cutter has to pull the level down to give itself room. Peaks should sit below –1 dBFS, ideally with 2–3 dB of true peak headroom. This isn't about making your record quieter — it's about giving the cutting chain room to encode the dynamics accurately at the highest possible level.
A mix that cuts loud.
There's no single specification that makes a perfect vinyl master — it depends on the genre, the format, the side length and the specific material. But these are the characteristics that consistently allow the loudest, cleanest cut.
Cuts loud
- ✓Dynamic range preserved
- ✓Peaks below −1 dBFS
- ✓Bass centred below 120 Hz
- ✓Sub rolled off below 30 Hz
- ✓Transients intact, not crushed
- ✓Gentle limiting or none
- ✓HF not harsh or sibilant
Forces level down
- —Brickwalled, clipped peaks
- —Peaking at or above 0 dBFS
- —Wide stereo bass
- —Heavy sub below 30 Hz
- —Transients destroyed by limiting
- —Harsh high-frequency content
- —MP3 or lossy source
The counterintuitive truth about vinyl loudness: a record cut from a dynamic, relatively uncompressed mix at −14 LUFS will often sound louder on a turntable than one cut from a brickwalled streaming master at −9 LUFS. The dynamic mix hits its peaks hard, which the cutter can reproduce at full level. The limited master is already at the ceiling everywhere — there are no peaks to cut loud, just constant high amplitude that translates to a flat, lifeless cut.
Longer sides mean quieter cuts.
Mix dynamics and side length interact directly. The more music you fit on a side, the tighter the grooves have to be, which means the maximum groove amplitude has to come down. A side that's running close to the maximum recommended time will always cut quieter than a shorter side — not because of any fault in the audio, but because physics.
This means the mix-level trade-off compounds with running time. A brickwalled master on a long side is a double hit against cut level. The combination of a dynamic mix and a shorter side length will always give the loudest, most impactful result.
See the running time and quality guide for recommended side lengths by format and speed.
A practical note on DJ cuts: if you're cutting for sound system or club use and volume is critical, the two most impactful things you can do are (1) keep the side under the sweet spot length and (2) send a mix with genuine dynamic range rather than a streaming master. A 12″ at 45rpm with a side under 7 minutes and a properly prepared master will cut significantly louder than a longer, limited version.
What I do when a file isn't right.
Files are assessed before anything is cut. If there's an issue with levels, clipping, bass management or something else that will affect the cut, I'll come back to you before the lathe is switched on.
In most cases there are three options:
Option 1 — You pull a new export
If you have access to the session, pulling a new master export with better headroom and gentler limiting is almost always the right answer. I'll tell you exactly what to adjust. It doesn't require a full remaster — often it's as simple as pulling the limiter ceiling down a few dB and re-exporting.
Option 2 — Vinyl cut preparation
If going back to the session isn't possible, vinyl cut preparation is available at an additional fee of £20 for 7″ projects and £25 for 10″ and 12″ projects. This covers the treatments needed to get your file cutting well — level adjustment, bass management, de-essing, HF treatment. It's applied once per run, not per copy.
It's worth being clear about what cut preparation can and can't do. It will control the dynamics to some extent, tame problem frequencies and make the file safer to cut. But it doesn't make up for a bad mix. If the fundamental balance is wrong — if the hi-hats are consistently 5 dB hotter than everything else, if the bass is heavily limited with no transient information left, if the whole mix is brickwalled — cut preparation can only do so much. The best version of a bad mix is still a bad mix on vinyl. Going back to the session and fixing the source is always the better option when it's available.
Option 3 — Cut as is at reduced level
If the file is borderline and you want to proceed without changes, it can be cut at a reduced level to stay within safe limits. The record will be quieter than it could be. I'll always flag this before cutting so you know what to expect.
Nothing goes to the lathe without your sign-off. The aim is always to get the best possible result from what you've got.
Common questions.
A vinyl lathe cuts grooves physically. The louder the signal, the wider the groove needs to be. The cutter has to set a level that the single loudest moment in the recording can survive without the groove walls colliding with the adjacent groove. Everything else on the record sits at or below that level — so one hot peak brings the whole thing down.
Below −1 dBFS on true peaks, with no clipping. Integrated loudness of around −14 to −16 LUFS gives good results. The most important thing is dynamic range — a mix with natural peaks and valleys will cut louder than a brickwalled master even if the brickwalled version measures louder on a meter.
Yes, significantly. A hard limiter crushes peaks and creates a signal that's constantly at near-maximum amplitude. On vinyl this means the groove is always at near-maximum excursion with no room for dynamic peaks to breathe. The result is a quieter cut, increased distortion on transients, and a record that lacks punch. Gentle limiting that preserves the shape of transients is fine. Brickwalling is not.
Low frequencies require the most groove movement. A loud kick or bass note at 60–80 Hz moves the cutting stylus significantly with each cycle. Wide stereo bass compounds this by creating lateral groove movement at low frequencies, which takes up enormous space. Centring bass below 120 Hz and rolling off below 30 Hz gives the cutter significantly more room to work with across the rest of the frequency range.
It depends on the problem. Vinyl cut preparation is available at £20 for 7″ projects and £25 for 10″ and 12″ projects, and covers level adjustment, bass management, de-essing and HF treatment. It's applied once per run, not per copy. For technical issues — a file that's a few dB too hot, wide stereo bass, a harsh high end — cut prep will make a real difference. But if the mix has fundamental balance problems, no amount of preparation at the cutting stage will fix them. The best version of a bad mix is still a bad mix on vinyl. If you have access to your session, going back to the source is always the right answer. Send the file and I'll tell you honestly what's fixable and what isn't.
Often yes, depending on how it was mastered. A lot of commercial mastering for streaming is too limited for vinyl, but it varies. Send the file and I'll assess it before anything is cut. If there's an issue I'll tell you exactly what it is and what the options are. Nothing goes to the lathe without you knowing what to expect.
How to prepare your audio
File formats, sample rates, loudness targets and everything else before you send.
Running TimeRunning time & quality
How side length affects cut level, bass response and overall sound quality.
DistortionInner groove distortion
Why the last track on a side sounds different — and what you can do about it.