I’ve been following a disturbing trend over the last few years as the Android platform (and now WP7 as well) matures. Smartphone screen sizes just keep growing and growing, and they don’t seem to want to stop. I have a number of issues with smartphones that have overly-large screens. It pains me to see that, while Android is known for giving users many choices, it’s nearly impossible to get a reasonably-sized flagship phone. For me, for a smartphone to be a ‘smartphone’ at all, and not a tablet, it has to be easily usable with one hand. Of course then the definition of smartphone/tablet will change from person to person, because our hands are not all the same size, however, there is certainly a finite limit for everyone where a phone will become too big to be comfortably used with one hand.

I’m currently testing the Samsung Galaxy Nexus. So far it’s been a rather wonderful phone, and I recently wrote this on Google Plus:

I’ve been using the iPhone for 3 generations. Right now I’m testing a Galaxy Nexus. If they made the same exact phone in a size that’s actually comfortable for one-hand use, I might call myself an Android convert. Curse you 4″+ screens and the awful fad that you are!

For me, the 4.65″ screen on the Galaxy Nexus is just too big. I constantly have to shuffle the phone around in my hand because Android places the two most frequently used aspects of the interface (the menu buttons and the notification drawer) at opposite ends of the phone. The size of the phone and the required shuffling means that I’ve got a poor grip on it, and I’ve been rather worried about dropping it during use. Again, those with larger hands will not have the same issue at 4.65″, but at some point they will run into the same problem.

Android Handset Screen Size Over Time

To show the trends of Android smartphone screen sizes over time, I compiled screen size and release date data from 155 smartphones from five major manufacturers (Motorola, Samsung, HTC, Sony, LG). I’d like to thank PDADB.net for their comprehensive release date info. (click to enlarge graphs)

As you can see, since the introduction of the 3.2″ HTC Dream / G1, screen sizes have consistently increased. Today we’re seeing 4″, 4.5″, 4.7″, 5″, and even 5.3″ smartphones! A simple projection (seen on the main chart) suggests that before 2013 is out, many handsets will have 5″ screens, while the flagship phones of that time may have even larger screens (if this trend continues) of 5.5″ or perhaps 6″.

With a slope of 0.0016, LG is increasing its Android smartphone screen sizes the most rapidly of these five manufactures. Despite pioneering some of the largest phones on the market at certain points in the timeline, Motorola is actually showing the slowest rate of increase in Android smartphone screen size with a slope of 0.0009, but of course this isn’t very far off from the leader!

Why is This Happening?

A good question to ask is what’s prompting the growth in screen size. It seems natural for manufacturers to have experimented with screen sizes as the platform grew legs. Different screen sizes are a point of differentiation for an Android phone manufacturer — a way to stand out in a sea of similar options. Bigger screens were also an easy way for companies to try to beat out the iPhone on features, even if the ‘bigger is better’ argument doesn’t hold much water in this case. Now it seems to have turned into a snowball effect whereby manufacturers are trying to one-up each other to have the biggest screen in town (all the while, Apple has stuck with 3.5″ since the introduction of their handsets). You wouldn’t believe how many times I’ve heard the phrase “biggest and baddest” when marketers are referring to a new Android phone. They use this phrase as though bigger is always better, but I must say — when it comes to comfortable one-handed smartphone use — it is not.

Where Does It Stop?

My question is this: where do we draw the line? As I mentioned, despite variations in hand sizes, everyone reaches a limit of comfortable one-hand usability at some point. I don’t have the raw data to back it up, but I believe that Android smartphone screen sizes are rapidly surpassing the maximum size for comfortable one-handed use by the average Android customer. None of this is to say there aren’t advantages to having a larger screen (particularly when it comes to media viewing), but given that people much more frequently use their smartphones for apps rather than media viewing, the argument for surpassing a users one-handed comfort zone to provide a better media experience is a poor one.

It’s not so much that screen-sizes are increasing (the chart clearly shows that other sizes are still available), but the bothersome fact is that it’s near-impossible to get a flagship phone unless you’re willing to buy one of the massive phones on the market. If you want a phone that comes in a size that’s comfortable for one-handed use, you have to be willing to settle as a second-class Android citizen — the only options available to you will likely have slower processors, less RAM (and may be based on an older platform) than the newest and biggest flagship phone currently on the market.

Last year I wrote two report about tablet design. The first was written a year ago and discusses things to consider when designing or buying a tablet-style device. The second was a checkist for a quality handheld internet tablet.

The iPad and Galaxy Tab are currently the best examples of consumer tablets in the market today and for me the quality and mobility of the Galaxy Tab means it has been working well in my family of devices. I’ve had it for over 3 months now and during that time I’ve been compiling a list of issues and improvements I either think are possible now or in future versions of the Galaxy Tab. The list applies to other tablets too so if you’re designing or buying tablets for the market in 2011, this is the sort of list you should already have on your whiteboard. If not, take this one and make it better. (A donation is appreciated if this list does make it to your company whiteboard!)

The list is over 50 items long and is split into three sections. Changes that could be made to the current Galaxy Tab are listed first and this is a good place to scan through if you’re looking to buy a Galaxy Tab. The second section includes what I would call reasonable hardware expectations for second-gen or late 2011 high-quality tablet. I don’t expect all of these features to be included but it’s a list of possible improvements. The third section is a little way-out there and includes hardware changes that may not be possible this year or could be too niche. I did have a section including third party software but that list got very long indeed and is not really relevant here.

IMPORTANT: There’s more to be added to this list and a great discussion about tablets to be had here. Chip in with your thoughts too. You’ll have ideas that no-one has through of yet so don’t be afraid to dump them in the comments below. If it’s not too wild (lets keep it within a 2-year timeframe) i’ll add it to the list and credit you.

Software and experience changes

1. Web browser.  Faster, mouse/finger over. Higher quality. Plugins. Password sync. Common gestures.
2. Better protection of UI experience under load. (Run UI in separate core? Possible hardware change required for that.)
3. Improve UI physics. Response rate needs to be 20ms or less. (I believe from my audio work in the past that 20ms is where delays become noticeable. I could be wrong but you know what I’m getting at right?) Try a drum-machine programme, It’s impossible. (Android 2.3 bring in features that can improve this through the use of the NDK)
4. DLNA improvements. (Receivers and senders need to improve compatibility.)
5. Protect the audio playback from stutters in multitasking scenarios. Critical. (How about a ‘dedicated’ mode where the application is brought to ‘realtime’ status – in effect, like iOS does.
6. Vastly improved audio library features. playlist import/export, cross-fading, id3 tag editing, jukebox mode, cover retrieval, radio streams, integration with Last.fm amd similar services. Much quicker media scanning.
7. Easier way to auto-organise applications list. Sort-by: most used, alphabetical, recently added, categorise (based on market categorisation) Apps list is as important as an audio catalogue. Genres, personal ratings on client should feed back to Market.There are a ton of improvements that can be made here.
8. Samsung apps style improvements. Lose the wood-grain effect or allow theming! Some feature improvements are also possible.
9. More video content to buy / stream. Major issue outside the U.S.
10. Full BT 3.0 (wifi TX support, near-field pairing)
11. Longer battery life (of course!) Important – battery save mode that schedules network usage. Significant improvements will require hardware changes.
12. More tablet / pro apps to be included via the Samsung catalogue
13. Better gallery with sort, date search, rename, tagging, face detection etc. Stock Gallery is very limited.
14. Better printing support. Google Cloud Print service should fix this.
15. Apt-x codec support over BT A2DP. (Or some other HQ wireless audio transmission capability)
16. SMS remote kill feature
17. Compatibility with camera applications
18. Timed profiles, turn to silence.
19. Improve speed of re-scanning Wifi access points. Can be extremely slow in some situations.
20. Occasionally boot-up time can run into 2 minutes or more. This needs to be improved
21. Hot-swap SIM cards (where applicable)
22. Slippery when dry. The smooth, easy wipe finish needs a couple of high-friction areas for book-reading
23. AC3 and DTS soundtrack handling. (Down-sampling to 2-channel and pass-through for HDMI)
24. Lower the lowest screen brightness (for in-bed, next-to-partner usage.)

Essential Hardware changes:

1. Longer charger cable (and micro-usb port)
2. Camera Quality: Glass lens, better sensor, continuous AF when recording video. Lens cover.  Flash options.  HD recording.
3. Camera shutter release button
4. Fm receiver + transmitter
5. USB OTG support for external storage, keyboard, mouse, other peripherals (midi, dvb-t, external cam)
6. Better GPS.  Sensitivity, speed.
7. Analogue video out built-in. Is that old-school? E.G. VGA, S-Video, Composite
8. HDMI-out port.
9. Storage improvements. Speed of core storage needs improving. 32GB, 64GB option.
10. Indicator lamps – multiple for use with different applications. Programmable colour.
11. indicators and alerts need to feed to paired Android phones.
12. Replaceable battery
13. Kick-stand
14. SDXC card compatibility
15. Higher quality screen (at same or lower power utilisation.)

For the ultimate tablet:

1. Digital radio reception.  DAB (Update: and other standards)
2. Digital TV reception. DVB (Update: and other standards)
3. Digitiser for graphics and handwriting input.
4. HQ audio recording. External mic over BT? Array mics.
5. Near field payments support (Probably best on a phone tho)
6. Daylight readable screen (transflective)
7. Quick Fingerprint reader for security.
8. Midi support (be a midi sequencer/controller)
9. HQ Audio synthesiser and audio effect support in hardware. (for use as effects unit)
10. Built-in mini projector
11. Remote mini handset accessory for using tablet as a phone. (Over BT with address book, dialing, voice dialing, CLI
12. Screen extension (slideable, removable)
13. PVR facility and dock. (hardware video encoding – 720p)
14. Wi-Di and wireless audio over Wifi
15. Rear panel controls and gestures area.
16. Glasses-free 3D (plus content)

…and lastly. Lets have a penta-boot system with Windows, Android, MeeGo, Ubuntu and WebOS!

Update: Reader contributions

• Better device security including encryption of sensitive information (BryanB – via comment below)
• Good design and implementation of user interface frameworks and the design/implementation of the applications. Really poorly coded application user interfaces can use all the CPU available. [Chippy: I think that's really part of the Android framework though.] (Sam – via comment)
• Samsung Keyboard Editor – (Max – via comment)
• A hinged back cover with a kick-stand that would allow easy battery replacement, SSD upgrades, and safe storage of extra SDHC cards. (jjsjjsva via comment) [Chippy: I love that SDHC storage area idea. That's a winner!]

Continued from Twitter for  @alsutton @beantin @mkearley2008 and others that were maybe a little surprised at my tweet this morning…

“My current stance on iPad alternatives – There are none.”

I couldn’t answer the twitter responses in 140 chars so here’s a more detailed, and I hope, understandable explanation.

Consumer Internet device success continues to hinge on applications. It’s the reason that AppUp exists, that Nokia will invest 10m into developer incentives in the US, why Samsung is throwing money to Bada developers and why Chrome OS will have a web app store. It brings critical ‘value-add’ to a product for customers, incentivises (is that a word?)developers and provides revenue opportunities for operators. It really is a killer application and only one vendor has got it right in this space so far.

While the hardware and design for most consumer internet devices [as I write, this mostly means tablets] is the same and one could argue that there are, from a visual and usage-case perspective, many choices, only one device has the application ecosystem that gives it the ‘value-add’.’ There isn’t another stack of silicon, hardware, operating system and services that provides this and there won’t be until well into 2011.

Android is a fantastic alternative *opportunity* of course but having done more testing than most on large-format Android devices, (I cast a glance over to the Toshiba AC100) I can see that current applications are still focused on the small screen. There aren’t any compulsive large screen games, video creation, music creation or even productivity apps feeding through that consider the larger form factor, longer battery life and often, more powerful CPU and GPU of a consumer internet device. Why should there be? Android V2.x doesn’t provide the hooks for large-screen app development and Google limits the use of their Market to phones (and large phones.) I also think that AppUp is a good opportunity. There are now 1000 apps in the store that are written with a larger screen and CPU in mind. Most are monetised and there’s potential for much more to happen on MeeGo (not forgetting Ovi) during 2011 but right now, can anyone name me a ‘complete product’, from silicon, through design, operating system and applications ecosystem, that offers the same as the iPad?

We used to jokingly call the iPad a large iPhone but the application store has given it differentiation. In the Android world, that differentiation option hasn’t even been enabled yet. Android tablets with the application store really are large phones and until Android is enhanced and Google widens the doors to the market (and possibly creates a large-format application suite) the solution has a limited future and doesn’t offer an alternative to the iPad.

There’s one other point I want to make. If you’re looking for an alternative to the iPad, there isn’t one. If you’re looking for a different product that looks the same then there are some choices out there. Unfortunately, in this consumer internet device/tablet market, I don’t see many people defining their requirements before choosing a solution. I see the product desire growing through application desire (and style, of course) and not connectors and micro-sd card slots.

That’s just my opinion so feel free to ramble and rant below! We might give vendors something to think about in their next product planning meeting!

Update: I always encourage thought about personal requirements. This ‘chooser’ tool i’m working on (currently in Alpha) shows 4 leading tablet devices and allows you to set your requirements and see a ‘winner’ based on public ratings. It’s not a foolproof way to choose a device but it’s a good way to start thinking about requirements. Ipad, Galaxy Tab, Viewpad 7 and Dell Streak Chooser Tool.

This might seem obvious but we think that many of our readers are looking for a similar product –  a consumer handheld internet device, which kinda makes sense because that’s all we cover here at Carrypad!

We wanted to simplify the decision process and short-cut some of the fuss and hype that goes on around the tablet space so we’ve come up with a checklist for you. It’s focused at the consumer tablet but much of this applies to slider and clamshell designs too.

Bookmark this article and when you see a product you like, take a quick look at this list to see if it fits the mould for an enjoyable, usable, flexible, quality handheld internet product.

Read the full story

In a report from Morgan Stanley we find out that Intel have launched the Moorestown platform today.

Update: Information is obviously under embargo at this stage but there are a string of tweets that have just gone through that mention Moorestown and a Z6xx processor. e.g. this one. “Intel Z6xx smart-phone processor prototypes: Moorestown massacre: Intel says that fantasy phone is on its way, wit…” They all link to a Cnet UK article that has obviously been removed. The current Menlow platform uses Z5xx processors. This makes sense.

Update 2: Everything is official now. The Z6xx has been launched. Interesting news on Android too.

Morgan Stanley have studied Moorestown and believe that Intel is ‘well positioned with MeeGo’ and that video performance will compare with the latest smartphone platforms, including Tegra 2. They also think that Moorestown will ‘meet or exceed’ current smartphone performance. It’s a bright report that will definitely give the ARM ecosystem partners something to think about.

At its Moorestown launch on May 4, we expect
Intel to introduce and advocate multiple benchmarks to
measure and compare highly debated performance and
power consumption attributes of Smartphone
application processors. In this report, we present
several comparison frameworks, which we plan to
update after actual Moorestown data become available.
Our view is that with Moorestown, Intel will finally start to
meet the power budget for Smartphones, but will show
more favorably on processing power benchmarks.

See also: UMPCPortal Moorestown analysis here.

The financial report also states:

Our checks indicate that Intel and Nokia are also collaborating on a
Smartphone device, which we think is likely to become available in the market in
mid-to-late 2011.

Considering that this report is focusing purely on Intel’s smartphone processor and this statement appears in a section on MeeGo, the report implies that the Nokia phone will be based on Intel’s Moorestown platform and running MeeGo.

The Morgan Stanley report was published on May 3rd and has been promoted by Intel in its ‘Chip Shot’ blog. The full PDF, an interesting read if you’re comparing smartphone platforms, is available as a download here. (or via the Chip Shot blog linked above.)

Stand-by for official Moorestown launch info!

I’m somewhat frustrated by all the Apple tablet talk. Number 1, nobody has a clue what’s going on and 2, there are so many design issues with ‘tablets’ that people need to be very careful about what they get excited about. I wrote these notes a few months ago but it makes sense to post them now so that you can make assessments about WHAT YOU NEED before any wave of marketing hits you. (*1)

In addition to the notes, I’ve added a couple of diagrams that I’ve previously used in posts and presentations and I also want to point out that this article is mainly focused on hardware. Software is a critical part of the equation.

Finally, I’m not saying that this is a set of rules. This is just my perception of what’s going on. I’m excited to hear what you think and ready to learn more from you comments.

Usage Scenarios

Firstly, here’s a diagram of internet-connected usage scenarios that lie between Laptops and Smartphones. There might be a few other niche categories like digital photo frames and video conferencing but the ‘ring’ below covers the primary sectors.

Usage by screen size.

Secondly, here’s a diagram that highlights functions that sit comfortably in certain screen sizes. There are 4 screen size ranges and I’ve positioned tasks in the smallest possible category. Example: Try doing advanced photo editing on a 5” screen!

Click to enlarge.

Design/Usage Notes.

In no particular order, here’s a brain dump of design and usage considerations for tablet PCs.

• One handed use: weight needs to be under 400gm for one-handed (finger-touch) use.
• One-handed use: Width of device needs to rest comfortably in the hand in portrait mode (7” max)
• Two-handed (only) operation not possible unless frame, corner or rear controls are implemented.
• Current screen backlighting technology will add between 0.5W (for a 5” screen) to 2.5W (for a high-brightness 10” screen) Idle power drain tends to 200mw for well-designed screen-off, idle networking scenarios.
• Minimum power envelope for an operating, connected device is about 2W (5”) or 3W (10”) (Only latest RISC CPU’s on best silicon processes can achieve this – expensive.)
• Max drain can reach +2W over operating power envelope.
• Battery capacity required for a 5” smart device – 10Wh. For a 10” device = 15WhTo achieve 4hrs always-on, dynamic and multitasking environment such as web and internet apps.
• Battery weighs about 8gm per watt/hour (without control / feedback electronics – add 50gm for that)
• Min battery (removable) and power electronics weight for a 7” smart device = 100gm
• Screen backlights are ALWAYS needed for low-light operation.
• Backlight average in-use power ranges from 0.5w (3-5” screen) to 2W or more (10” screen)
• Capacitive touchscreens can’t be used with gloves or basic stylus.
• Capacitive screens can’t be used for natural handwriting input, annotations, graphic creation/painting.
• Keyboards need to split/position correctly when in portrait and landscape modes.
• Tethering keyboards / headphones / data modems via Bluetooth is a long-winded process and requires batteries in end devices.
• Pixel density should be between 180 and 250 pixels per inch for standard web pages unless intelligent reflowing /zooming is used. An 800-wide webpage would require about 4-inch width (about 6” diameter screen in portrait mode.)
• Resistive touchscreens are not rugged.
• Book reading does not require two-pages per view.
• Hard drives are not rugged, are noisy and generate heat and vibration. Not to be used in tablets.
• An integrated folding stand needs to be incorporated for PMP functions.
• Wifi/Bt/3G/FM/GPS antenna separation is needed.
• Mouse pointers allow selection with minimal hand movement.
• Removing buttons helps aesthetics and ruggedness, reduces shortcuts for two-handed operation.
• Moving buttons to ‘on screen’ reduces physical feedback, reduces usable screen area.
• Buttons need to be backlit for night=time use
• Screen brightness needs to be extremely low for night-time use.
• Indicator lights distract and annoy when reading.
• Good quality speakers allow for multi-person viewing.
• Wifi needs to remain connected at all times (standby disconnection results in awkward delays on start-up.)
• Web pages need to load in a 10-second average time. (with text and scrolling being available in 50-75% of that time)
• Access to 100% of the internet requires Adobe Flash support (and, to a lesser extent, Silverlight and other run-time apps.)
• Text selection for web applications is critical. (cut and paste.)
• Zooming a web-page must re-flow the text to avoid left-right panning.
• Auto screen rotate from landscape to portrait must not reset previous text selections
• Using 3G when moving in car / train can drastically increase power drain from 3G components. (3G can take up to 1.5W in these scenarios – huge % of power drain.)
• Docking stations help to keep a device charged and located.
• Removable battery gives customer confidence about life-span of product.
• Glossy screens can be filtered. Matt screens can not be made glossy with third party products.
• Bigger screens are expected to be faster by the customer (and therefore highlight slow performance)
• The concept of ‘idle’ is only for the lab.
• Slider keyboards increase cost and size. decrease style.
• Nothing over 4” screen is truly pocketable for most people.
• Flash running on multiple tabs can easily take 100% of CPU on a ARM-based device.
• A frameless device is impossible. 10mm frame is a tough design challenge.
• For productivity users, a new operating system is a new learning curve.
• Third party applications decrease stability and security.
• Multi-tasking decreases battery life.
• Fingerprint readers can drastically improve security while decreasing key presses.
• A 7” screen device can be too big for a car dashboard.
• Ebook readers are useless to the average consumer without commercial content being available through the device.
• Cloud-based usage model is currently a home-zone possibility. (Not mobile) i.e. local storage and sync is still needed.
• Larger designs permit higher pricing.
• Larger designs need to deliver a faster experience (to satisfy user expectations.)
• Notebook / clamshell designs are recognized as computers.
• Sub 5” designs can be mistaken for smartphones.
• New business models dictate that consumer tablets must be low-cost point of sale devices. (Marketplace for apps, content, accessories)
• There are undiscovered usage scenarios.
• There are huge numbers of new technologies and inventions that I don’t know about!

Example scenarios and solutions.

Ebook reading. 200 PPI screen with 300gm or less. Daylight (ambient light support) and backlight support needed. Content must be easy to access. Color screens important for education market (and advertising.)  See also this article.

PMP. 5” screen gives a comfortable 60-80cm experience. HD playback on small-screens is required because that’s what many users create and many websites deliver. You have no choice! 7” give HD (720p) experience and 1 meter. 5-7” screen in 300gm allows for only a 10-15wh battery. 5hrs online use. 7hrs video use.

Newspapers. Large screen format 9-12” format is impressive and permits newspaper layout ‘standards.’ Requires significantly more battery power. Advertisers want color and animated advertising. Difficult to design for one-handed use. 12” awkward for mobile situations. 9-12” also required for one-page per view A4/Letter sizing. (PDFs)

Web. Minimum resolution 800×480 although 1024×600 more comfortable and allows better portrait mode. 1024×600 at standard web (readable at 100% zoom) font sizes requires 7” screen. Multi-tab Flash enabled web pages can kill battery life and hog CPU if uncontrolled.

Mail, IM, social networking.Two handed thumb keyboard can only be built around max 5” (landscape)  or 7” (portrait) screen dimensions without separate keyboard. Requires always-on scenario. See also: Perfect Microblogging Device article.

An example multi-use tablet: 7” 300gm with 12wh battery, Cortex A8 CPU (high-end), 3D GPU, Video decoder hardware. (Intel Moorestown also a contender here.) Stand. Highly advanced on-screen keyboard with haptic feedback. 3mb HQ camera. 1.3mp LQ camera. Capacitive touchscreen. GPS, accelerometers, ambient light sensor, 3G, Wifi, docking port, 3.5mm headphone jack, array mic, stereo speakers.

The ideal tablet hardware is worth nothing without the correct software, services and content.

(*1) over the last 3.5  years I’ve learnt a lot about tablets and mobile computing devices including a lot about what doesn’t work for the masses through our pro-mobility-focused  sister website, UMPCPortal.