LUFS for Tidal
Published · updated · by Loopin
Tidal normalises like everyone else, so the loudness target is unchanged. What differs is that a lossless stream shows you more of whatever you did.
The target is the same: −9 to −14 LUFS
Tidal applies loudness normalisation on playback around the same figure as its competitors. Mastering louder buys nothing except the limiting damage.
Pick by genre and hold it. There is no Tidal-specific loudness and any advice claiming otherwise is describing a setting, not a requirement.
Lossless means your decisions are audible
On a lossy service, heavy limiting and encoder artefacts blur together. On lossless, they do not — a squashed master arrives exactly as squashed as you made it, and a listener on decent headphones hears the difference between a master with dynamics and one without.
This is the one genuine argument for mastering less aggressively: the audience most likely to notice is disproportionately on this platform.
The practical consequence: a target chosen for how the record should sound, rather than for how loud it can be made, is the one that survives this platform intact. Everything you would have spent on the last three decibels is audible here as the thing you gave up to get them.
Deliver 24-bit at your session rate
Give your distributor the best file you actually have. 24-bit at 44.1 or 48 kHz is right for most sessions, and higher if you genuinely recorded higher.
Do not upsample. Rendering a 44.1 kHz session at 96 kHz adds no information, doubles the file, and some delivery pipelines reject or flag it.
Where this bites in practice is sample rate conversion. A 48 kHz session delivered as 48 kHz stays untouched; the same session converted to 44.1 goes through a resampler whose quality you do not control. Deliver at the rate you worked at and one conversion disappears from the chain.
Ceiling at −1 dBTP anyway
Lossless delivery removes the encoder overshoot argument for the lossless tier — but Tidal also serves lossy streams on lower tiers and on constrained connections, and your file is the same file.
Keep −1 dBTP, or −2 if you want one master that satisfies every platform including the strictest. It costs nothing perceptually because normalisation sets the level.
Do not dither twice
If you deliver 24-bit, do not dither — dithering belongs only on the final reduction to 16-bit, and the platform handles its own conversions.
Dithering a 24-bit master adds noise for no reason, and dithering twice compounds it. Master once and let the distribution chain do its own conversions.
If your distributor asks for 16-bit specifically, dither once at that final step and never again. Two dithers in a chain is audible as a raised noise floor on quiet passages, which on a lossless platform is exactly where somebody will hear it.
Frequently asked questions
What LUFS should I master to for Tidal?
The same -9 to -14 LUFS integrated range as everywhere else, chosen by genre. Tidal normalises playback like its competitors, so louder gains nothing.
Does lossless streaming change how I should master?
It makes your decisions more audible rather than changing the target. Heavy limiting that blurs into encoder artefacts on a lossy service arrives exactly as squashed as you made it on lossless.
Should I dither a 24-bit master?
No. Dithering belongs only on a final reduction to 16-bit. Dithering a 24-bit file adds noise for no reason, and doing it twice compounds it.