Mobile Adaptation of Silver Conformance: Reliable CSS px → Android dp Conversion Patterns

[CATche3r]

As a mobile OS UX researcher focusing on typography scaling, I'm investigating the applicability of the Silver lookup table (primarily designed for web CSS pixels) in native Android environments. Specifically, I'm seeking insights on two key aspects:

Unit Conversion: What's the most accurate method to convert CSS pixel-based font sizes to Android's density-independent pixels (dp), considering varying screen densities and viewing distances?

Underlying Models: Are there established mathematical models or conversion formulas that inform the Silver lookup table's recommendations? I'm particularly interested in understanding how it balances readability standards with device adaptation principles.

Any references to conversion frameworks or research papers addressing cross-platform typography scaling would be greatly appreciated!

[Myndex]

Hi @CATche3r

The CSS px is not pixel-based. The CSS px is a screen-density independent measure, as relative to the canvas abstraction layer. The CSS px is not to be confused with device pixels (though to be sure, some manufacturers/developers make it this way to some degree).

The px is defined as 1.278 arc-minutes of visual angle.

The reason APCA Readability Criterion uses px, is because the px is the canonical reference for size relative to visual angle. At least that was the theory. Where it gets a bit dodgy, is the influence of retina displays on design. The way small fonts are aliased on a 72ppi display, vs on a 3:1 retina-type display is not to be underestimated. The values we chose relate to the low-resolution display, due to the impact of aliasing on contrast for low res displays.

The values chosen for size are based first on x-height, then adjusted for a reference x-height ratio of 0.52 (Helvetica/Arial).

That said, here's a chart of equivalent device pixel density vs viewing distance, to maintain a px with a visual angle of 0.0213° or 1.278ʹ (minutes of arc).

For accessibility, what we are more concerned about is the ability of the user to scale-up small text 200% for mobile (400% for desktop).

Does this answer your question? Please follow-up if you'd like to discuss further.

[CATche3r]

Hi @Myndex

Thank you for your detailed response. The reference chart has been particularly valuable, as it has systematically guided me in applying silver conformance principles to mobile platforms.

I would like to further inquire whether there exist established mathematical models or computational formulas that could comprehensively reconstruct the complete silver lookup tables. Specifically, my research focuses on determining contrast thresholds for Chinese typography on mobile devices. Given the substantially different display characteristics and usage patterns compared to desktop environments, I hypothesize that contrast compliance requirements may need mobile-specific adjustments beyond current silver conformance standards.

As my experimental capacity is limited to testing specific font parameters (e.g., 14sp/400 & 24sp/700), I seek methodological frameworks that could enable responsible extrapolation of results to intermediate values (such as 17sp/400). Additionally, given my use of Chinese character sets as experimental stimuli, I wish to investigate whether such models could demonstrate script-agnostic applicability beyond Latin-based typography systems.

[Myndex]

Yes! While not in release yet, the next iteration for this year will be a pure algorithmic approach, and include line thickness or dataviz.

The lookup tables were an easy way to explore the concepts of glyph weight in development.

[Myndex]

Also: regarding mobile, we are working on additional, and separated, guidelines for certain devices types, due to the unique needs.

Regarding "Silver" -- Silver was a exploratory project, leading eventually to WCAG 3. Silver is over, and no aspect of Silver should be referenced. WCAG 3 is evolving, but far from release.

For beta use right now, we (Inclusive Reading Technologies) have the APCA Readability Criterion, as a public draft. We are planning a large upgrade later this year, pending some current research.

[CATche3r]

Looking forward to your great works!!!