Bill Day's Blog

The Great Cellcam Debate

I'm not a big fan of the expression "digital convergence", but I do appreciate its results. One of the most important of these results: Cell phones with built in digital cameras (some people refer to them as camera phones, but I prefer cellcams) .

I first wrote about cellcams in my first java.net blog entry. In it, I discussed the Nokia 3650, my main device since mid-2003. The 3650 is a Series 60 based MIDP 1.0 smart phone with support for WMA and MMA. It even allows me to use MMA to capture still photos via the built-in VGA camera (though video capture is limited to a native Symbian app). This converged device has allowed me to finally shed the PDA-plus-separate-cell-phone-plus-separate-digital-camera lifestyle for a more liberating cellcam smartphone all-in-one persona.

Nonetheless, the 3650 has it's shortcomings. It's big and bulky, which proves a real pain when I'm out for a jog or short on pocket space. It's J2ME implementation isn't upgradable, so I'm not able to take advantage of all of the new MIDP 2.0 capabilities. And its camera, while taking serviceable 640x480 digipics (roughly 0.3 megapixels), isn't going to replace my Canon S400 with its four megapixel (Mpix) resolution for the shots that really matter. If Nokia wants to challenge the likes of Canon for the instant, in your pocket, "always ready" class of digital cameras as it seems to be indicating, they have to do better than the 3650.

Which brings me to the crux of this blog: Although the first generation of cellcams has sold me on the promise of converged devices, I'm looking for the next generation to truly deliver on that promise.

What I want from a cellcam in 2004:

• MIDP 2.0 including WMA via JTWI
• 1+ Mpix quality camera with flash and support for MMA
• Bluetooth including the Java APIs for Bluetooth (throw in WiFi, too, and I'd really appreciate it)
• Small and light (the Nokia 7210 and 7250 seem to have done well, so why not a next gen cellcam of roughly the same size and weight?)
• Tri- or quad-band GSM/GPRS/EDGE for worldwide roaming

I could come up with a longer wishlist, but I'd be happy with the one above if someone will just bring one of these devices to the market. Are my requests unreasonable? I don't think so. Several manufacturers are bringing out cell phones in 2004 that meet my J2ME requests, and multi-band GSM handsets are already widely available in most of the top cell phone makers' product lines. Kyocera has plans for a 1.2 Mpix cellcam with built in flash in the US in 2004, and NTT DoCoMo has already launched a 2.02 Mpix camera in Japan.

All that is missing is a device that integrates all of these capabilities together. The closest thing I've found so far is the Nokia 6230 (due Q1 2004). The 6230 will meet most of my requests save for Mpix image quality (no improvement over the 3650's 0.3Mpix and no flash, unfortunately) and WiFi support. Why not up the camera resolution, add a cheap flash and WiFi, and really take on Canon (and Kyocera, and PalmOne, and...)?

And while we're at it, why not enable not just still photo capture but also video capture via MMA? That could really shake up the applications of these gadgets. Heck, it's the holidays, maybe if I dream big Santa will take care of it, right? My fingers are crossed...

Comments

• My wishlist...
  Wouldn't it be just great if write once, run everywhere would become a reality? Sure, the devices are resource constrained, but that's no excuse for the number of bugs and oddities found in a large number of mobile phones.
  Having to work around firmware problems, not to mention having to find what causes a problem, is a major pain.

  Having easier access to support personel of some of the mobile phone manufacturer would be nice as well. Some of the MIDP developers do not have the funds to buy into expensive support programs just to get questions answered as to how a specific MIDP implementation handles certain details.

  Having a multi million pixel cammera in a mobile is very nice, but I would settle for being able to use the features that are supposed to be available right now.
  (MIDlet signing on Nokia 6600 for instance)

  Oh, and my global wishlist has world peace at #1 of course :-)

  Posted by: kersing on December 23, 2003 at 03:15 PM

• Cellphones aren't just for consumers anymore
  Wereas I appreciate Bill's comments, I'd like to express a my disappointment with the question at hand: compelling features for new cell phones. We need to add some economic backbone to this question.

  My premise is that it is our responsiblity as developers to buck the trend, resist the urge, boldly stand out, and instead of talking about the next consumer, techno-geek feature we'd love to have to impress our friends, we should be talking, making noise, standing up and shouting, about what features will help us create jobs, intersting jobs, good jobs, writing applications for said cellphones.

  It used to be fairly obvious that industry, typically the millitary or other government funded industry, drove most technology innovation. Take for example light-weight polymers and cell technology itself. Moreover, it was the governmental funding of this innovation that created the most intersting technology jobs. If you have ever worked in R&D you know what I mean. It is there where developers put to the test the first C++ compilers, the first OODBMS, the first CORBA implementations, the first Java versions, the first J2EE App Servers, the first XML processors, and yes, the first java-based cellphones and PDAs.

  The readers of this column/blog is not just your average cell phone user, it is your technology developer. So let's not get sucked into dreaming about what our favorite new techno-gadget would look like. Because having our favorite new techno-gadget won't create an intersting techno-related job for us. It may enable us to be a better techno-worker, but it won't create the demand to create the job. And after all, don't we really just want to have interesting techno-jobs in intersting techno-fields so that we can afford our techno-toys.

  So I pose a new question: What are the compelling features for new cell phones that will engage BUSINESS users and increase BUSINESS ROI, and (by-the-way) create great jobs for us developers?

  Here is my list, it consists on just one item:
  --- High-end phones for low-end workers ---
  Rugged, large application space, large data storage space, fast JVM, Bluetooth, accessories (barcode scanners, printers, touch pads, credit-card readers).

  Stop giving features to the consumer and the white-collar worker in let's-see-how-small-and-light-we-can-make-this-disco-styled-phone packaging. Put those same features in a ruggedized non-flip phone that can be sent out on the road with fat-fingered repair men like roofers, plumbers, appliance repair personel, sewer workers, a property managers.

  The white collar worker will get what they want. They will their get BlackBerry's integrated with their email systems and PDAs where they can finish their latest memo. And the applications that they use? Pocket Outlook, Pocket Excel and Pocket Word. Where does that leave you and me, the java-based wireless developer?

  But, for every one white-collar manager there are lots of low-end mobile dispatched workers - plumbers, electricians, applicance repair techs, property managers, truck drivers, etc.. And what business need is visibility to these workers, paperless communications, in-field invoicing, and paperless filing of field reports.

  Yet, there is only one phone available today (and looking into the next 3 quarters) for these workers - the Nextel/Motorola i58. Have you ever tried to build a compelling business application in 256K heap? These applications need two-way flow of information, geo reporting, invoice printing, credit card payment, time keeping, signature capture, inventory management, bar code scanning, text entry, and (yes Bill) digital images. All the business applications on the market today do one or two of these. Why? Because anything larger just won't fit.

  So we say: "But look, here is a really nice new phone from Motorola/Nextel an i730 with GPS, color screen, and a large application and data space". Price is really not an issue, but the reply is: "I'm not going to give that flimsy phone to my roofer, it would never last one day - how about this nice rugged i58?"

  Or they say: "But I need bar code scanning so that I can scan inventory parts." and there is no commercially available barcode scanner for the i730.

  Or: "Look at this great phone from Nokia with a flip out keypad... what you limited the jar size to 50K?"

  Well, you get my drift. I think the phone manufactures are well on the way of fixing many of the problems of the 1st generation of phones. All the new platforms will have MIDP 2.0, large application spaces, run multiple applications simultaneously, and maybe even have bluetooth. They will support a large subset of additional JSRs (MMS, WMA, ...). Accessories will come (though I'd like to some standards there). But we need these features in rugged packages for mobile-dispatched workers.

  I encourage you to talk to your contacts at the various phone manufacturers and bring up these issues. Our voices will be heard.

  Nancy Lehrer
  VP of Technology
  JumpStart Wireless
  www.jumpstartwireless.com
  

  Posted by: rovernl on December 24, 2003 at 12:10 PM

• Nokia 6230 URL
  For some reason the short URL for the 6230 (forum.nokia.com/6230) doesn't seem to be working right now. The long URL is:
  http://www.forum.nokia.com/main/1,6566,015_249,00.html

  Posted by: billday on December 24, 2003 at 03:17 PM

• The Nokia 6600 comes pretty close...
  I just bought a Nokia 6600 and it has everything you mentioned except for the 1+ Mpix camera.
  It's a tri-band phone that runs Symbian 7, has MIDP 2.0, MMA, WMA, and bluetooth API. It makes decent 640x480 pictures, the screen supports 16k colors and it is small and light.

  The battery doesn't seem to last that long... but maybe that's because I'm playing with it too much :)
  

  Posted by: amooy on December 24, 2003 at 03:19 PM

• Image quality
  While the average Joe may be most impressed by the megapixel count of a phone, there's much more to image quality than the number of pixels.

  To fit in a compact phone, manufactures need to compromise on three things: memory, sensor, and lens. Flash RAM will continue to increase in capacity, so technology will help us here. Sensors also continue to improve in quality and capacity. But as sensors shrink to fit in a phone, or as megapixel counts climb higher, the size of the individual photocell that captures light must decrease. As a result, the amount of light captured is also decreased, so it must be amplified. This increases the amount of noise, giving the photo a grainy or speckled look. Noise reduction technology may improve, but we're running into basic laws of physics here.

  Physics also works against the tiny lens included in cellcams. They, too, are limitted by the amount of light they can capture, and there optical quality cannot approach that of a quality digital camera.

  I don't believe cellcams will ever compete with digital cameras in the quality department, nor are they really meant to. They're a convenient way to take and share spur-of-the-moment snapshots, which is great for a lot of people.

  Posted by: jimothy on January 06, 2004 at 03:02 PM

This work is licensed under a Creative Commons License.