What we refuse to do
Poetry Audio is a native macOS music player for people who own their music. It plays local files through the DAC you chose, without putting a server between you and the music.
That is the ordinary part. The important part is simpler: Poetry exists to avoid lying about audio.
We do not claim bit-perfect when we are not
When volume control changes the samples, the label changes. When EQ is active, the label changes. When ReplayGain applies gain, the label changes. When sample-rate conversion is required because the selected device cannot accept the source path, the label changes.
There is no magic "bit-perfect mode" toggle. There is a state machine reading what actually happened, and a label derived from that state.
Green requires proof. Amber means transformed. Red means unproven or blocked.
We do not do MQA
Poetry will not implement MQA badges, MQA authentication theater, or post-MQA rebrands as a trust signal. If a format cannot be defended transparently, it does not belong in the core product narrative.
We do not enhance your masters by default
If you give Poetry a well-mastered 24/96 file, the default job is to play it, not improve it.
Restoration can be useful for compromised material. Neural bandwidth extension can generate plausible missing content. That does not make it original content, and it should never be framed as transparent recovery of information that was not in the source.
We do not fill empty space with generated taste
If MusicBrainz, Wikipedia, Wikidata, Cover Art Archive, Discogs, ListenBrainz, LRCLib, TheAudioDB, or fanart.tv can provide sourced metadata, Poetry should show it with attribution.
If a fact does not exist, Poetry should show the gap. Generated filler text is not criticism, context, or provenance.
We do not turn a tool into a cloud dependency
Poetry is local-first. Your library is yours. Playback should continue when the Mac is offline. Optional network calls should be visible, inspectable, and tied to metadata or integrations the user explicitly enabled.
What we measure and show
Poetry's product is not "another audiophile player". It is a player that tells you what is happening to the audio in real time.
- Provenance ledger: source, transforms, output device, quality state.
- Loudness and true peak measured locally.
- Room-correction methodology, not vague "better sound" claims.
- Source-backed metadata with visible attribution.
What this adds up to
Poetry is not trying to be the player with the loudest claims, the largest editorial database, or the most magical DSP.
It is trying to be the player that does not lie about your audio.