Top	[This Network is not currently active and cannot accept new posts]
<- Previous	Next ->

295 hits

Nov 25, 2003 2:52 pm		re: re: Death of Websafe Palette - or worse?
Frank Collyer		Haim, I think you miss the point of the article. "Ugly dithering" isn't the only problem we have to worry about - how (supposedly) websafe colors get rendered by other-than-24-bit monitors is the concern here! The author mentions a test he did by making a table with cells of different "websafe' HTML colors, then making .gif files with the same "websafe" color, but smaller than the cell, and placing it into the cell. In many cases where the monitor was not displaying in 24-bits (regardless of the graphics card) the browser interpreted the same HTML and .gif color differently. Obviously this is a problem. There still are a lot of 8-bit systems out there, and many more that for different reasons (some games run faster in 8-bit) aren't displaying 24-bit color. Is this a problem? Maybe. Even if 5% - 10% of the target audience is looking at an incorrectly displaying page, it may very well be a problem, generating one very displeased client. One of the solutions the writer of the article suggests is adding transparency to .gif files and matting them against the HTML background color so that even if the supposedly same .gif and HTML color displays differently, it will not be distractingly obvious. This is just a sensible design step to avoid a potential problem. It's not about avoiding the use of certain colors, but avoiding situations where display problems can occur. This is a sound design principle for both web designers and usability people. The websafe palette probably is irrelevant to anyone viewing on a 24-bit monitor. What is relevant is designing pages that display well regardless of bit-depth considerations. Best- Frank Collyer > Haim Hirsch wrote: > I was always led to believe that the web-safe palette was all about avoiding ugly dithering. The article you mentioned explains the browser-color-rendering issue very well with respect to background colors defined by the HTML. Apparently the palette is now down to 22 colors. > >Web designers do have other alternatives for background color rendering, and can usually get around that issue pretty easily. What is typically unavoidable for a general web audience is the color rendering of the actual monitor. > >I have had to explain to a number of clients that even if we matched the website color EXACTLY to their brochure on their screen, it would look different on my screen...and it would look different on their laptop screen... > >Rarely does color need to be so precisely defined (Hunter Green vs. Forest Green). When such precision is necessary (see, for example online clothing companies such as Lands End) an offline sample is usually the best solution. > >For the purpose of applications, there are more than enough distinct colors to effectively differentiate an appropriate informational resolution for human-readable needs. Doppler radar map coloration that demonstrates this to a small extent can be found at Weather.com. > >With newer graphics cards and monitors, dithering is rarely a problem. Our computers can display more colors than most people can identify (even using modifiers such as "light" and "dark"). > >Is the Websafe Palette dead? No. But, I would definitely argue that it is Irrelevant from a Usability standpoint. > >-Haim > > Private Reply to Frank Collyer (new win)

© Ryze Limited. Ryze is a trademark of Ryze Limited.  Terms of Service, including the Privacy Policy