The other day, the topic of gifs and photosensitive epilepsy came up on DS106 Twitter, so I thought I’d find out more. I went straight to the medical journals and asked a friend of mine who works at Jisc TechDis.
Types of light sensitivity?
There are a range of ill-defined sensitivities to light, from facial twitching in response to bright light, actual induced changes in brain electrical activity, and changes dramatic enough to induce a seizure. The latter is referred to as photosensitive epilepsy.
What objects or aspects of light can evoke a seizure?
Many of the medical papers report the most common stimulus is a FLASHING LIGHT, and can also be PATTERNS OF LINES, GRATINGS or CHECKERBOARDS. I think we all recognise sometimes we can watch stripes or squares on the TV or PC screen that make our eyes feel funny.
How many people does this affect?
The figures are a little unclear because the data relies on individual’s own reporting, and tests actually done in the lab measuring brain activity (electroencephalograms EEGs) are usually on selected individuals and not necessarily those at risk so are likely to be lower. Nethertheless, the figures range from 5 to 9 % of the population. SO IT IS STILL A RISK FOR A NUMBER OF INDIVIDUALS. (Fisher 2005). Studies suggest it is more common in young people than in adults.
There have been cases of TV programmes and video games inducing seizures. Wilkins and colleagues (2005) described a case relating to a TV advertisement (Golden Wonder Pot Noodles) where seizures were reported in three viewers in the UK. In Japan in 1997 a children’s program “Pokemon” reportedly led to 685 admissions to hospital, with 560 identified as epilepsy. The cartoon displayed alternating red and blue backgrounds. What was startling was most of these people had never had an episode before. I’m not sure of the follow-up, meaning, whether this was just a one off event or actually primed the people into being sensitive, I do not know.
So what colours and patterns are thought to be worst?
Wilkins 2005 paper led to guidelines being introduced in the UK and Japan, and now also internationally. Although they are devised for TV and not PC screens where proximity, resolution and environmental conditions may be different, they are a useful guide for our understanding.
1. Frequency. Flashes with frequency >3 Hz are prohibited.
2. Opposing changes in luminance. Flashes ≥20 cd/m2 are prohibited.
3. Area of flashes. Flashes greater in area than one fourth of the screen are prohibited.
4. Color. Flicker from saturated red light is prohibited.
Frequency – in Hertz (Hz), the number of cycles of the pattern per second. Fisher (2005) reported that the majority of people in their study responded at a peak of 12–30 flashes per second, but Wilkins’s 2005 paper reported this lower at 3Hz or above.
Luminance – relates to the contrast between light and dark. The bigger the contrast, the bigger the risk.
Areas – stationary patterns seem worse than ones that move slowly. Stripes, gratings and patterns are implicated more than spots for example. The paper explains this relates to area covering the screen. On a PC we can assume the pattern might occupy the entire screen.
Colour – red is thought to be more provoking, although the effect of colour is not really well known. The importance of luminosity and contrast is much greater as a consideration.
So they recommend as a correction for you to look at the screen and ask these questions:
Are there more than five stripes?
If so, do they last longer than 0.5 s?
If so, does the brightness exceed the stated limit? If so, categorize the motion of the pattern
Are the guidelines contravened?
If so, reduce brightness.
So what can we do?
- TV broadcasts are now monitored, but computers and video games are a bigger risk.
- People with known epilepsy will safeguard themselves using screen protectors and maybe coloured glasses, but many people just won’t know.
- Whilst the guidelines are written for TV, the effect of digital screens is less well known.
How to practice safe gifs?
Well not just our beloved animated gifs, but video and animations. I would use the criteria above as a guide – avoid flashing geometric patterns of contrasting colours.
But as my friend at TechDis commented:
Just don’t do it. Its not good for accessibility or usability.
As reported in a specialist workshop in 2004, flashing images on websites and roll-over-images are not good for usability. (NGfL Accessibility workshop 2004). I guess if we are making “art” we might want to think about it a bit more carefully.
And some additional notes on accessibility from Terry at TechDis regarding visual formats:
The accessibility of an image depends on technical factors (colourless, contrast, sharpness) as well as pedagogical factors (how effectively it is labelled, how well it is described, how it is integrated into a text / audio narrative).
For moving images, video and gifs:
However visual formats can create barriers to people with poor sight and they tend to be far less searchable and navigable than text.
So, we should certainly think about the art we are creating, and maybe we should be more careful about accessible alternatives to go alongside it?
Or maybe we should consider having a clear warning if we are unsure?
WARNING. CLICK ON THESE LINK ONLY TO GET EXAMPLES OF WHAT MIGHT BE RISKY.
ARE YOU AFFECTED BY THE WEB IN THIS WAY? I’D LOVE TO HEAR FROM YOU? I AM A MEDICAL SCIENTIST AT THE UNIVERSITY OF THE WEST OF ENGLAND IN BRISTOL. I ALSO LOVE MAKING COMPUTER ART!
Fisher RS et al (2005). Photic- and Pattern-induced Seizures: A Review for the Epilepsy Foundation of America Working Group. Epilepsia, 46(9). Available: http://onlinelibrary.wiley.com/doi/10.1111/j.1528-1167.2005.31405.x/full
Jisc TechDis (2014). Technology matters – visual. Available: http://www.jisctechdis.ac.uk/techdis/technologymatters/visual
NGfL National Grid for Learning Accessibility Workshop (2004). Flashing, flickering and blinking. Available: http://www.websemantics.co.uk/tutorials/accessibility_workshop/sessions/session2/05.colour_usage/06.flashing/
Wilkins A et al (2005). Characterizing the Patterned Images That Precipitate Seizures and Optimizing Guidelines To Prevent Them. Epilepsia, 46(8):1212–1218. Available: http://www.ncbi.nlm.nih.gov/pubmed/12504202