Gallery

Making the cut

Posted on 30 Mar 2009 at 16:44

Jon Honeyball discusses the future of Microsoft following its revenue upset and reveals his concert pianist roots.

Screen resolutions

A couple of musings on the subject of screen resolution are in order, I think. A few months ago, I was looking at the rather fetching Eee Top device from Asus, which is basically one of the firm's netbooks rewrapped into a desktop box. It has a flashy touchscreen, but after playing it was immediately clear that something wasn't quite right. A few mouse clicks later I found the problem - its display was running with the 120lpi (logical pixels per inch) screen furniture rather than the usual 96lpi items, and that was why everything looked a bit clunky.

Windows has supported both 96lpi and 120lpi screen objects ever since I were a lad so there's nothing new there, but the 120 setting was introduced to cater for high-resolution displays running on 15in monitors, and by "high resolution" we're talking 1,024 x 768, which was indeed a lot back then. It was felt that 96lpi was too few and would make text look undersized and thus hard to read - hence the arrival of that "8514A" specification 120lpi setting (8514A was the model number of an IBM monitor that first used that resolution). After a flurry of visual weirdnesses everyone finally decided to stick with the 96lpi, so it was a bit of a jolt to see 120 on this Eee Top display. Unfortunately, most application writers have nowadays completely forgotten about lpi settings and just assume that 96 is what their program will be using, so it was no surprise to find that some apps look gawky on the Eee Top.

And why was 96 chosen? Well, on an originally sized, VGA resolution screen it results in a typeface that's about 50% larger than typical printed matter, and this was deemed to be a good thing because the screen would be 50% further away from your eyes than a book or paper memo. But things become more complicated when you consider the use of big TV screens hanging off a PC. Nowadays, it isn't unusual to see a 50in plasma or LCD panel running from an HDMI output on a media-oriented networked computer. And then there's the issue of "HD ready" versus "True 1080p" - the former being around 1,360 x 768 pixels compared with a proper 1,920 x 1,200 or thereabouts for 1080p.

A question arises about what resolution is visible, and this applies not just to big screens but also to the current fad for tiny laptops with relatively high-resolution displays. So what is actually visible? Well that depends to some extent on the user's eyesight - there's variability here in the same way that there's variability in hearing - but most authorities suggest the average person can resolve detail down to around 1 arcminute, which is 0.0166 degrees.

All we need now is a spreadsheet to do the calculations. Here's one that was emailed to me by a colleague who found it somewhere (but he couldn't remember where). This spreadsheet bears the name of Jens T Saetre and a last modified date of 2007, so a quick search took me to www.satellite-calculations.com, where I found a real treasure chest of code snippets, applets and the like.

Jens' spreadsheet needs to know the diagonal size of your screen, its resolution in both horizontal and vertical dimensions, the distance from the center of the screen to your eye, and the arcminutes at which your eye saturates for detail. Plug in the values for a cheap 50in plasma TV running at 1,280 x 720 pixels and you'll find that your eye reaches vertical and horizontal detail saturation around 3m from the screen: in other words, if you're sitting farther than 3m from the screen you won't be able to perceive all the detail there.

Download a year of Jon Honeyball's Advanced Windows columns by heading to our Free Downloads site