apca-introduction

commit

01642b31d6f181b14c0589541a2adae3850c2b3b

parent

23d736da71aa8f0ffe7a6af75b57e09276404da3

Author

Tobias Bengfort <tobias.bengfort@posteo.de>

Date

2025-05-10 20:06

analysis: tweak threshold section

Diffstat

1 files changed, 16 insertions, 21 deletions

diff --git a/analysis.md b/analysis.md

@@ -272,41 +272,36 @@ possible exponents, while APCA goes to the other (lighter) extreme. This
  272   272 is consistent with the observation that APCA reports lower contrast for
  273   273 darker colors.
  274   274 
  275    -1 ## How much contrast is enough contrast?
   -1   275 ## Thresholds: How much contrast is enough contrast?
  276   276 
  277   277 ### WCAG 2.x
  278   278 
  279    -1 Smaller text is generally harder to read than bigger text. Generally
  280    -1 speaking, smaller text requires more contrast to be legible.[^3] And
  281    -1 body text can be read faster with better contrast.
   -1   279 Smaller text requires more contrast to be legible[^3], so the thresholds should
   -1   280 probably depend on font size.
  282   281 
  283    -1 WCAG 2.x requires a contrast of at least 4.5:1 for regular text and 3:1
  284    -1 for large text. If you aim for AAA conformance, the values are 7:1 and
  285    -1 4.5:1.
   -1   282 WCAG 2.x only distinguished two font sizes The thresholds are 4.5:1 for regular
   -1   283 text and 3:1 for large text. If you aim for AAA conformance, the thresholds are
   -1   284 7:1 and 4.5:1. How these thresholds were derived is not completely clear:
  286   285 
  287    -1 How these values were derived is not completely clear:
  288    -1 
  289    -1 > There was some user testing associated with the validation of the 2.0
  290    -1 > formula. I could not quickly find a cite for that. My recollection is
  291    -1 > that the hard data pointed to a ratio of 4.65:1 as a defensible break
  292    -1 > point. The working group was close to rounding that up to 5:1, just to
  293    -1 > have round numbers. I successfully lobbied for 4.5:1 mostly because
  294    -1 > (1) the empirical data was not overwhelmingly compelling, and (2)
  295    -1 > 4.5:1 allowed the option for white and black (simultaneously) on a
  296    -1 > middle gray.\
   -1   286 > My recollection is that the hard data pointed to a ratio of 4.65:1 as a
   -1   287 > defensible break point. The working group was close to rounding that up to
   -1   288 > 5:1, just to have round numbers. I successfully lobbied for 4.5:1 mostly
   -1   289 > because (1) the empirical data was not overwhelmingly compelling, and (2)
   -1   290 > 4.5:1 allowed the option for white and black (simultaneously) on a middle
   -1   291 > gray.\
  297   292 > -- <https://github.com/w3c/wcag/issues/695#issuecomment-484187617>
  298   293 
  299   294 [Large text] is defined as anything above 18 point or 14 point bold. The
  300    -1 definition comes with a lot of notes that explain the limits of that
  301    -1 approach though, e.g. that some fonts are extremely thin and that font
  302    -1 size depends on user settings.
   -1   295 definition comes with a lot of notes that explain the limits of that approach
   -1   296 though, e.g. that some fonts are extremely thin and that font size depends on
   -1   297 user settings.
  303   298 
  304   299 WCAG 2.x also comes with some rules that allow users to adapt font size
  305   300 to their needs: [1.4.4] requires that users can resize the text,
  306   301 [1.4.10] requires that they can zoom the whole page, and [1.4.12]
  307   302 requires that they can adjust text spacing.
  308   303 
  309    -1 So in a way, WCAG 2.x side-stepps the issue by handing control over to
   -1   304 So in a way, WCAG 2.x side-steps the issue by handing control over to
  310   305 the users who have all the facts.
  311   306 
  312   307 ### APCA