Archive for April 18th, 2006

The QVGA factor

Sometimes it is a bit scary how myself and Russell Beattie think so compatible in terms of all things mobile. He yesterday blogged about the QVGA issue. He noted that while more and more phones are now using QVGA 240×320 screens, the text amount fit on the screen is the same as when using a 128×160 or a 176×220 phone. The reason for this is because handset manufacturers are using QVGA LCDs at no more than 2.2″ diagonial and so they are forced to use huge fonts. Now, compare a QVGA screen at 2.2″ and one at 2.7″ or 2.8″ PocketPC. Notice how much more text information the PPC can handle and display, without straining the eyes. Indeed, it seems that the 2.8″ is the “sweet spot” of how big a QVGA screen should be, averaging the dot pitch and quality in regards to the amount of text information fitting in one screen. In the past, I have made a similar blog post, discussing the same issue at Nokia’s brand new phones.

And while QVGA seems to be the best resolution for productivity and web usage, the first consumer VGA phone was announced in Japan yesterday: the Sharp 904SH. And Vodafone LIES about it. They keep showing their 2.4″ VGA LCD displaying 4-times the TEXT information than their previous 2.2 QVGA LCDs. This is a lie and I bet my head that this is just a mockup and not the real user interface. Their user interface can not justify making the fonts incredibly small just so they can fit 4x the amount of text — it’s impossible to see something like that even with a 20/20 vision. You don’t believe me? Have a look and tell me if you could read the calendar app shown on the second image on a 2.4″ diagonial. I didn’t think so. And so their user interface now will HAVE to use BIGGER fonts (check images here for my proof), essentially making the VGA LCDs almost the same as the QVGA ones, text-wise. Sure, the LCD is now brighter and with a more compact dot-pitch, and so their firmware guys can use a bit smaller fonts than normally, so they can fit a bit more of text information in one screen than they could with QVGAs. But that won’t be the expected 4x improvement (as 640×480 is over 320×240). It will be a 1.3x (or at best, a 1.5x) improvement or so.

The only reason why someone would want a VGA phone is for pictures. Pictures will just look better and more crisp, especially megapixel pictures captured via the onboard camera. Specially-compiled VGA games will look great too (but they will be slower as they will have to transfer 4x the information on the screen, plus they will be very few of these VGA games for the time being). But in terms of web, messaging and office productivity, the advantage is so small that doesn’t justify the cost of the VGA LCD on the handset. To justify that, the VGA LCD should have been about 3.6″ diagonial (and that ain’t gonna happen on consumer handsets). In other words, whoever buys a consumer phone just because it has a 2.4″ VGA (and doesn’t care about camera images), is a complete idiot.

UPDATE: Check this graphic representation of a VGA screen, and visualize it how small it would be on a 2.4″ placeholder. Think how popular phone home-videos shot at 120×96 or popular games developed for 128×128 would look at that screen. They will be impossible to watch them (especially because very few video playback apps on phones do resampling at higher resolutions or go fullscreen). Cellphones already have a legacy and all software around them has being built around that legacy (of small screens). In my opinion, going to VGA at 2.x” is a mistake. Going to VGA using big screens, is a better bet. And QVGA will continue to be the best resolution to do stuff without its effects getting on your way and still maintain a small-ish size for the handset.

Eugenia wins. Again.

A few days ago I emailed Bitstream’s Thunderhawk support to ask for a feature: to add the word “Thunderhawk” on their user agent so my scripts can recognize it as a mobile browser and serve it automatically the cHTML pages (they are using a standard desktop Mozilla user agent as default, even if their product runs on Windows Mobile and Symbian only). Unfortunately, as with most tech support people, I got a standard reply that helped neither myself or them. I replied in my usual agressive fashion, and I explained the situation further, showing them demos of the HTML, cHTML and WAP versions of OSNews and explained how it all works and why there are cases where they need to identify themselves as a mobile browser. Given the fact that OSNews will soon move to CSS, it is on their best interest to have access to the lighter, simpler pages. I… mandated for this issue to be formally discussed at Bitstream’s engineering meetings. The tech support person was kind enough to pass through my request and so the Bitstream engineers discussed the issue. Today they replied that they are seriously thinking of adding the word “Thunderhawk” on their next version’s user agent (at least as an option to the user, if not as a default). That’s good news as far as I am concerned not only for me and the users, but also for the browser.

Bitmap vs Vector

Jimmac is right. Bitmaps are not going away just because Vectors are “in fashion” lately. Bitmaps are faster to render and they look better on smaller sizes. This whole debate reminds me a lot of the Jpeg vs Png, or the CSS vs cHTML. There are practical uses for every of these technologies. That’s why they are still alive. Best tool for each specific job.