What Streaming Codecs Do to Your Master

Spotify converts to OGG Vorbis, Apple to AAC. Both introduce inter-sample peaks and subtle artifacts. Here is what happens and why True Peak matters.

Spotify converts to OGG Vorbis, Apple to AAC. Both introduce inter-sample peaks and subtle artifacts. Here is what happens and why True Peak matters.

When you upload a WAV to your distributor, they send it to streaming platforms. Each platform converts it to a lossy format for delivery. This conversion is not transparent - it changes the audio in measurable ways.

Platform | Codec | Quality Spotify Free | OGG Vorbis | 128 kbps Spotify Premium | OGG Vorbis | 320 kbps Apple Music | AAC | 256 kbps YouTube | AAC/Opus | 128-256 kbps Tidal HiFi | FLAC | Lossless

The Inter-Sample Peak Problem

Your DAW shows sample peaks - the loudest digital sample in your file. But the continuous waveform is rebuilt between samples by every DAC and resampler, and that curve can overshoot 0 dBFS even when no individual sample does. These inter-sample (true) peaks already exist in your PCM file; lossy codecs add further overshoot when they encode it. Both can cause real distortion on the listener's device.

We verified this with real data: Suno MP3 files consistently show True Peak at 0 to +0.3 dBTP, while the WAV version of the same track sits at -2.5 to -3.0 dBTP. The MP3 encoding added about 2.7 dB of peak overshoot. LoudCraft's oversampled true-peak limiter holds every master at its job's ceiling (-1 dBTP, or -2 dBTP for cinema and ATSC targets), and lossy downloads are measured after encoding and re-encoded if needed to stay under it.

Did you know? The common -1 dBTP streaming ceiling exists largely because of codec conversion (EBU R128 broadcast also uses -1 dBTP, while ATSC and Netflix ask for -2 dBTP). That headroom is not wasted space - it is a safety margin that keeps your master from distorting on millions of phones and speakers worldwide.

Key takeaway: Your listener never hears your WAV. They hear a compressed version. A -1 dBTP ceiling gives the codec room to work; for masters louder than -14 LUFS, Spotify advises -2 dBTP because loud material overshoots more after encoding.