iOS UA reports a devicePixelRatio no real iPhone/iPad uses (not 2 or 3)
br.ios_dpr_incoherent · convicts
What it catches
Spatial DEVICE<->DPR coherence — the B1 joint constraint, sibling of br.ios_screen_oversized on the same iOS device manifold. iOS renders at a device-FIXED backing scale: every real iPhone/iPad reports window.devicePixelRatio of EXACTLY 2 or 3 (2 for iPad + SE/older iPhones; 3 for modern/Plus/mini iPhones — the iPhone 6-Plus family and 12/13 mini physically downsample but REPORT an integer 3; Apple ships no fractional-DPR or 1x iOS display). Grounded in the Apple device corpus (iosref.com/res, ios-resolution.com). An iOS UA whose devicePixelRatio is not in {2,3} is off the manifold: a DESKTOP faking iOS reports DPR 1 (the common desktop-scraper default), a sloppy emulator a fractional value (1.5/2.5). FP-SAFE by the hard Apple hardware bound (the DPR set is device-fixed and integer-valued on iOS); distinct from br.devicepixelratio_anomaly (which only flags a non-finite/<=0 DPR, not an OS-incoherent one) and complements br.ios_screen_oversized (screen) — TOGETHER they force a coherent iOS spoof to get BOTH the screen geometry AND the backing scale right, not just the UA + touch. The joint-coherence ratchet, one more rung. GROUNDED LIVE: a Chromium given an iPhone UA at the container-default DPR 1 (KS_DPR unset) fires ios_dpr_incoherent; the same UA at a real iOS DPR (KS_DPR=3) is SILENT — the coherent-randomizer evasion.
Signals it reads
browser.ios_dpr_incoherent
How it fires
present
Bypassed by 10
Frontier evaders that reach the detector uncaught (scored only suspicious, defeating every convicting tell) — this check is not one that stops them. The red-team frontier this detection still has to convict.