Quick answerRe: One Simple Infographic that explains Spatial Frequency

Hi Alastair!

Sorry just saw this: I’m about to go into a meeting, so very briefly:

Contrast Constancy indicates that once above a certain critical contrast level, further increases won’t have much of a noticable effect.

However, at high frequencies (i.e. a 1px line) contrast constancy does not behave as many expect. The threshold CSF curve then does become more instructive.

Assuming a standard observer and the CSS reference pixel of where 1px = 1.278 arc minutes visual angle, then on my CSF chart that is 23 cpd (cycles per degree). (there are 2 px per cycle) That is the body text at contrast sensitivity 20 to “smaller thinner” text at contrast sensitivity 107. So 1px to 2 px is (more or less) increasing viewer sensitivity about 5 times, for a normal view, and other things being equal, and being adapted to the white point, and actual contrast down in the 3:1 area or less, etc. 

But yes the upshot is 1px to 2px will most likely make a much bigger impact than increasing contrast from 3:1 to 4.5:1, which is the color are dark, the WCAG math makes barely a bllp of increase. 

MOREOVER:
 

Lowering frequency (increase thickness) is the BEST way to assist impaired vision. Impairments shove peopls CSF way down in terms of frequency.

All I have time for right now SORRY!

Will follow up later

Andy




> On Oct 16, 2019, at 2:34 AM, Alastair Campbell <acampbell@nomensa.com> wrote:
> 
> Hi Andrew,
>  
> For images in emails, I suggest putting in a link to that image on your website (or a website), as a secondary method for people to see it. For LVTF email the images do get stripped, on the AG list they get included at the bottom, I don’t know why it’s different!
>  
> Also, I think the special frequency aspect might provide guidance for a question I have about the new focus indicator SC.
>  
> We’re currently setting the minimum bar at:
> A change in contrast of at least 3:1.
> A size in CSS pixels that is at least 1px multiplied by the longest edge.
> If the indicator it doesn’t contrast with the adjacent color, it should be at least 2px thick.
>  
> E.g. these [1] pass, default then focused:
> <image002.jpg>
> <image008.jpg>
> <image010.jpg>
>  
> <image012.jpg>
> <image013.jpg>
>  
> These fail:
> <image015.jpg>(1px increase next to a non-contrasting color)
> <image022.jpg>(light grey replaces white)
>  
> The problem is that this [2] also passes:
> <image025.jpg>
>  
> In terms of ratcheting up the requirement, my impression is that it would make more sense to increase the surface area than the contrast requirement.
>  
> E.g. A size in CSS pixels that is at least 2px multiplied by the longest edge
>  
> So this would pass:
> <image028.png>
> With a 2px thickness line.
>  
> Whereas increasing the contrast to 4.5:1 wouldn’t make as much difference?
>  
> Cheers,
>  
> -Alastair
>  
> 1] https://alastairc.uk/tests/wcag22-examples/focus-more-visible-2.html  <https://alastairc.uk/tests/wcag22-examples/focus-more-visible-2.html%20%0d2>
> 2 <https://alastairc.uk/tests/wcag22-examples/focus-more-visible-2.html%20%0d2>] http://3needs.org/en/testing/code/focus-visibility.html <http://3needs.org/en/testing/code/focus-visibility.html>
>  
> From: Andrew Somers <me@andysomers.com <mailto:me@andysomers.com>> 
> Sent: 11 October 2019 05:28
> To: Jim Allan <jimallan@tsbvi.edu <mailto:jimallan@tsbvi.edu>>; Wayne Dick <wayneedick@gmail.com <mailto:wayneedick@gmail.com>>; Jeanne Spellman <jspellman@spellmanconsulting.com <mailto:jspellman@spellmanconsulting.com>>; Shawn Lauriat <lauriat@google.com <mailto:lauriat@google.com>>; Laura Carlson <laura.lee.carlson@gmail.com <mailto:laura.lee.carlson@gmail.com>>; jjuett@salesforce.com <mailto:jjuett@salesforce.com>
> Cc: public-low-vision-a11y-tf <public-low-vision-a11y-tf@w3.org <mailto:public-low-vision-a11y-tf@w3.org>>
> Subject: One Simple Infographic that explains Spatial Frequency
>  
> Hello everyone,
>  
> I am endeavoring to write shorter emails! Here is a simple infographic:
>  
> <image003.png>
>  
> Assuming a standard viewer distance (CSS reference pixel visual angle):
> a stroke width of between 8px to 20px is at the peak of contrast sensitivity, while 
> a 1px stroke width drops contrast perception down an order of magnatude x2 from peak.
> Increasing stroke width (by increasing font size and/or font weight) increases perceived contrast
> especially closer to threshold, and
> especially at the higher frequencies (over 16cpd).
> impairments offset this curve, typically down and to the left.
> The implication is that impairments come closer to threshold sooner, and need additional contrast
> Please let me know if there are questions/comments.
>  
> I do have one question: is there a way to save graphics/attachments for emails to the list archive? I see they are always stripped? Perhaps I should just double post them to the Wiki?
>  
> Thank you!!
>  
> Andy
>  
> 
> Andrew Somers
> Address Redacted for List
>