Supersonic Man

June 2, 2016

a small attempt to emulate the gadget press

Filed under: Uncategorized — Supersonic Man @ 9:08 am

Nowadays the popular media report on the latest gadgets almost as eagerly as they report on celebrity gossip.  Since my smartphone is now three models out of date, I’ve been reading more than my share of this stuff.  And this is inspiring me to try adding a little noise of my own to that topic.  So:

Five Things Premium Phones Will Need in Order to Stay Premium


There are more and more Android smartphones on the market that offer once-premium features at a midrange price.  It’s getting harder and harder for the top-end phones to justify their high prices, and about the only thing holding back a wave of middle-priced Chinese models taking over the market is the reluctance of the major carriers to embrace them.  (My carrier doesn’t seem to want to offer anything in the middle, premiumlike or not — almost everything is either under $250 or over $600.)  Now that it’s becoming possible for what used to be carrier-exclusive features, such as Wifi Calling, to be obtainable in unlocked phones, this effort to keep the old contract-subsidized business model going may not be able to hold back the flood for much longer.  And when that happens, it’s going to be more difficult than ever to justify paying $700 for a phone that used to be head and shoulders above the lesser choices, but is now just a mild bump in specs over its competitors.  Things like an aluminum unibody frame or a glass back just don’t cut it any more as differentiators.  Waterproofing still kind of does, but I bet it won’t in another year or two.

Given that these previously exclusive traits are now becoming so commonplace that the word “premium” is becoming seriously overused when describing all but the cheapest phones… what will it take in the future to make a high-end phone truly better than its midrange imitators?  Here are five things that I would hope to see as minimum qualifications for premium-ness a few years from now:

1. AMOLED display

If you want to impress people, a good IPS display doesn’t cut it anymore.  To be premium, you need the best, and that is AMOLED.  I think this differentiator is why Samsung’s Galaxy S phones are so popular, despite them having a lot of qualities that I’ve never liked very much.  Their version of AMOLED is the best display you can buy.  The shortcomings of any liquid-crystal technology such as IPS — both in the dark, and in daylight for sunglasses wearers — are just too big a compromise for me to feel like a phone using them is premium.

Of course, there are good AMOLEDs and not-so-good ones. There are a number of midrange phones already trying the latter: displays with blue-green color casts, poor outdoor brightness, questionable long-term durability, and of course not a hint of colorspace management. These, of course, will not fly in premium-land.

Even Samsung has quite a ways to go yet on the colorspace front.  This is not a novel problem — it has long since been solved for PCs and Macs, though in a way which is inconvenient for most users as it requires calibrating whatever monitor you happen to buy.  On a mobile device with no choice of display, calibration can be done at the factory, allowing one to have both full vividness and photographic accuracy at the same time, as appropriate to context.  That combination will not be so easy for cheaper phones to deliver in the near future — they’ll generally have a single gamut, which might be accurate or might be wide but is unlikely to manage both at once.  This is because achieving accuracy on a wide gamut panel will also require a substantial amount of software support, which is much more easily within the reach of major makers.  (But note that Google might, at any time, build in better color management and make it easier for everyone.)

To be premium, they’ll have to offer not only well-controlled color, but the widest possible range of brightness, from something like 600 nits in sunlight down to about 2 in darkness. A range from like 375 to 10, as is all too typical with IPS phones, is too compromised to be premium.

A good AMOLED display can also extend battery life, and since no new technology seems to be available yet which can give the batteries in an expensive phone any qualitative advantage over those in midprice ones, anything to reduce power drain is very valuable.

(Speaking of which, why isn’t wifi calling standardized across carriers yet? That can save plenty of power indoors, as well as allowing calls to get through clearly in situations where they might otherwise fail.)

2. Audiophile sound

By audiophile sound, I do not mean supporting 24 bit audio formats.  I am very skeptical that anyone can really hear the improvement from that.  What I mean is providing DACs and amplifiers which are capable of giving you all sixteen of the bits you’ve already got in CD-quality audio.  Most phones use generic DACs and amps which don’t sound much better than a Walkman-type player from the turn of the century.  Plus, they are often lacking in power.  I am definitely encouraged by companies like HTC and LG and Alcatel for investing in significant upgrades to either the headphone amp, the speaker amp, or both, and am only disappointed that LG made this an extra-cost addon instead of building it into the phone in the first place.  This is exactly the sort of feature that a premium phone should have automatically, while a midmarket phone can get away with omitting it.  And if a midmarket company like Alcatel wants to give it a try as well, more power to ’em — maybe they’ll shake up Samsung, which as yet hasn’t bothered to make any real effort in this area.  Apple is much more conscious of it, of course, since their phones evolved from music devices originally.

On this front, one definitely has to watch out for phony snake oil audio features, which generally come down to just strange EQ curves that overemphasize bass and such, or to fake stereo expansion.  HTC is certainly guilty of such moves at times, but thankfully has allowed users to opt out.  Advertise such crap if you must, as long as the option exists to turn it off.  If the underlying audio left behind is of high accuracy and linearity, with low noise and distortion and tone coloration, with a solid amount of power behind it, then that’s definitely a mark of a premium phone.

Strong audio could cost some battery life.  As long as the user is aware that lower volume = longer listening, that’s acceptable.

Another feature that’s of value to true audiophiles is additional storage space, so they have room for lossless audio files instead of compressed ones.  Bring on the gigabytes.

3. Multiple camera lenses

The capabilities of a camera small enough to fit in a smartphone chassis are pretty near maxed out.  Adding more megapixels or a more advanced sensor is no longer going to produce much of an improved picture — the capabilities of today’s sensors are already too close to the absolute limits imposed by the nature of light.  Better lenses aren’t very feasible either.  That leaves only one way for future phone cameras to be better than today’s, without creating a big bulge on the phone: they’ll have to use more than one of them.

There have already been several phones experimenting with putting two cameras side by side.  The particular usage made of them varies: some use the pair to gauge depth, which can be used for simulating the bokeh of a large lens, or to aid focusing.  Some use one for wide angle and the other for a tighter view.  In one current model, the pair includes a color camera and a black-and-white camera, which assists the color camera in producing sharp detail.

Each of these is an improvement that can’t be obtained with a traditional single camera, and furthermore, they don’t exclude each other: it’s perfectly possible to do all of these at once, even though no phone yet does so.  I think that for a phone to have a premium camera, it had better try to do at least one of these enhancements.

The long term goal is a technology called synthetic aperture.  That’s a technique by which a device can combine the imagery of many cameras — perhaps dozens — to simulate the effect of viewing the scene through a single lens of much greater size and acuity.  Prototypes have been built of “phones” which have lenses sprinkled all over the back — some large, some small, some wide-angle, and some telephoto — and combine them all to produce images that rival those from much larger cameras.  This combination not only allows you to exceed the sharpness of any one lens, and to simulate a zoom lens using fixed ones, but to also have the low-light capability of a larger sensor (one as big as all your small ones added together).

When fully developed, this technology might not only allow phones to take pictures of a quality previously only possible with cameras far heavier than any phone, it could also produce 3D images, in which every pixel not only has a brightness and a color, but a known distance from the viewpoint.

Even the selfie cameras should eventually be synthetic aperture: by ringing the bezel of the device with several tiny cameras, you can simulate a camera that looks out from a viewpoint in the middle, rather than from the edge.  This will make video phone calls much more pleasant, as the person on the other end can look straight at you, instead of seeming to be looking at your chin or forehead.

This is a ways off yet, but some of the initial benefits are available now, and it’s time for those who want to set their phones apart from the pack to start taking advantage of these possibilities instead of just including a me-too camera.  And of course, since each camera adds a significant cost to the phone, the sheer count of lenses would be a strong differentiator of high-end phones from cheaper ones.

4. A choice of sizes, including small

Everyone complains about the lack of smaller sized Android phones and nobody does anything about it. It’s to the point now that Samsung’s Galaxy S phones are practically the smallest available in their class — and note that Apple, the other biggest selling brand, is similarly compact. Most of the moderately less expensive competitors to these phones are considerably bulkier, and if they offer a mini version, they tend to cut the specs waaay back. Less speed, less RAM, less everything.

Could it be that this is not because of customer demand, or because of a naive misreading of market trends… but because making a powerful phone small is genuinely difficult?  It may be that those who can charge premium prices are the only ones who can pull off the feat of making a phone smaller without compromising performance. If so, that’s an area of the market where they can sharply separate themselves from the me-too brands. Especially if they make it clear that you really are just getting a choice of size, and in every other way it’s the same phone.

Besides Apple, only Sony seems to be making the attempt at making a small phone with premium features.  But for some reason they’ve practically disappeared from the US market — why, I don’t know, as their recent models sound quite desirable to me on paper.  Maybe their design is just not attractive to people.

Of course, if we start stuffing the insides of phones with a dozen separate cameras, the difficulty of making it small only increases.

5. Absence of lag

This is the most difficult, because the issues that cause lag and stuttering and unresponsiveness are, unfortunately, built right into Android, just as they are in Windows. Eliminating it may require a substantial investment in deep Android modifications.

Hardware can make a difference. One thing about the flash memory that current phones use for file storage is that its speed at writing new data can be much lower than its speed for reading old data. This can cause horrible drops in performance during operations which do a lot of writing to storage, such as Play Store automatic updates. This issue can be mitigated somewhat by spending more money on your flash chips: the good ones don’t slow down as much. Samsung has definitely spent extra money on higher grade flash storage, to this end. Extra RAM also helps, and they provide it.

Yet they still have a reputation for failing to provide smoothness — one which, in my limited experience of their current models, they still deserve. They’ve greatly increased the speed and the number of cores in their processors, and it hasn’t resolved the problem.

This is because it’s mainly a software issue.  Lag primarily comes from one elementary mistake which is repeated throughout many layers of both the applications and the OS components.  That mistake is a failure to do interactive stuff first.  Correct software design for a responsive interface has follow one rule: nothing takes precedence over responding to user input.  Obeying instructions from the user is the device’s purpose — anything else is secondary, and can wait.  No matter how busy a phone or computer is, no matter if it is running at the limit of its cooling capacity with every core maxed out, that is no justification for failing to respond instantaneously if the user gives a new instruction.  Note that this doesn’t mean that you have to carry out the work instantaneously.  If the work is going to take a while, what you need to do instantaneously is acknowledge that you are starting it, and if appropriate, make sure that the command to cancel it is ready to use.  What’s crucial is that the user must see your response, and not be left wondering if he has to hit the button again.

I haven’t yet learned to program Android, but I’m familiar with programming Windows, which is the worst when it comes to this lag problem.  The essential core of the problem is that Windows treats user input actions such as keystrokes and clicks as if they were messages in an inbox, to be worked through when the app has time to get to them.  This makes it very easy for app programmers to accidentally get a little too busy to keep up.  It’s possible, with care and professionalism, to avoid this, but getting behind is a state that’s just naturally easy to fall into.  And this lackadaisical approach (which essentially stems from the limitations of the early 16-bit environments that Windows was first sold for, and hasn’t been fundamentally updated since) extends to the OS internals as well: it’s all too easy for it to get busy with something so that even if your app is ready, the OS doesn’t get around to your app in a timely way.

Microsoft tried to address this fundamental issue when designing their ill-fated “Windows RT” system (which lives on today as Windows Phone 10).  This was an all-new OS, a Windows in name only, which was designed so that anything which might take a bunch of time would happen asynchronously — that is, instead of the application performing the action as a single step, and being unready until it’s complete, the application instead initiates the action, and then actively monitors it to see when it completes, remaining ready and responsive in the meantime.  This is a far better approach for maintaining responsiveness, but a more difficult one to program.  This is a technique which is already available on every modern system, but which developers do not use nearly as often as they should, so Microsoft decided to just force the issue.

Apple’s iOS generally does a better job at responsive smoothness than Android does, even though Android systems often have faster hardware.  The Android approach does not overtly share the shortcomings of the Windows one, but behind the scenes it’s apparently got the same sort of single-threaded input queue, which leads to similar results.  A total redesign of Android is not in the cards, but there are some measures which a premium phone maker, or other entity with the resources to do some hard work on the OS, could take while preserving its basic structure.  The most important would probably be to make some changes to the part of the system which switches the available cores between multiple tasks, to make sure that whichever one is actually in front of the user, being actively used, is able to far more aggressively preempt any other task that tries to work in the background.  At present, if something like a virus checker is working hard behind the scenes, any modern operating system is supposed to give that a lower priority than stuff the user is doing actively, so it stays out of the way and, in theory, should not noticeably affect responsiveness.  This is fine on paper but somehow keeps failing to work in practice — even if cores are available for everyone, there can be conflicts over things like memory and bandwidth.  What Android needs is an adjustment such that as soon as the active foreground application needs to respond, anything else which is in its way can get sharply pushed back, or even slammed to a halt, if that’s what it takes to avoid lag.  The background stuff is supposed to stay out of the way when asked nicely, but when this is not working, harsher measures need to be employed.

There are cases where two operations can be legitimately in the foreground — not just when there’s a split screen, but at times such as browsing the web while also listening to music.  The fixes to improve responsiveness will have to adroitly handle a variety of such situations, and the complexity will be no small challenge.  For instance, there’s a whole class of actions where precise timing is more important than anything else, particularly when dealing with video and audio.

That’s far from the only adjustment they’d have to make.  And the phone maker could also devise new extended APIs that allow apps to have fewer difficulties with OS lags, make use of them in their own proprietary apps, and suggest them as models for future expansion of the system by Google.  This is the sort of thing that isn’t going to come from phone companies building down to a price.

Avoiding lag within the web browser is a whole separate problem.  Fortunately, desktop browsers have advanced a lot in how they deal with situations such as long download delays or overly heavy Javascript.  Mobile browsers have not kept up, and need to get with the program.  There are also opportunities to take more advantage of using multiple cores on a single webpage.

 

Advertisements

5 Comments »

  1. Right now I’m looking at phones and very much wanting half of it to be an S7 and the other half to be an HTC 10. But hopefully in the coming months there’ll be a good choice for my needs that costs less. Perhaps the Alcatel Idol 4.

    Comment by Supersonic Man — June 3, 2016 @ 1:18 am | Reply

  2. I think the phone I want might be one that’s coming out in a couple of weeks: the ZTE Axon 7. It has amoled, it has stereo speakers, it has upgraded audio amps, and (most importantly for me), it has support for T-Mobile wifi calling even though it’s sold unlocked. It doesn’t have a dual camera, but I don’t care, since when I want to take pictures I bring a real camera. The one thing it conspicuously lacks is the option of getting it in a compact size, though it did eventually offer one for the preceding Axon model. If I buy this any time soon I’ll have to surrender my distaste for phablets.

    My one worry is whether, as a maker with great success at making cheap crappy phones, if maybe they lack the competencies necessary to make a dependable high end one.

    Comment by Supersonic Man — July 21, 2016 @ 12:45 pm | Reply

    • Nope, no ZTE Axon 7 for me. The promised Wifi Calling support has yet to materialize four months after release, and I kept hearing that there were shortcomings in the radio antenna sensitivity which might make it miss calls. Still, they had good sales success by making a front-speakered AMOLED audiophile phone, so I hope they and others do it again next year.

      Samsung ran into issues with bad batteries, so I eventually got an HTC 10. It has neither AMOLED nor front stereo, but I couldn’t find a viable alternative which did.

      Comment by Supersonic Man — October 8, 2016 @ 7:28 am | Reply

  3. I’m glad to see Google following at least two of my suggestions above. It may have been a disappointing shock when the transition from the Nexus series to the new Pixel brand was accompanied by a stiff price increase, but at least they did produce two equally high-spec AMOLED phones with one being compact.

    Comment by Supersonic Man — October 8, 2016 @ 7:21 am | Reply

    • This year’s dual camera iPhone, and next year’s rumored replacement of Retina displays with amoled ones, are further steps in the right direction.

      Comment by Supersonic Man — November 2, 2016 @ 12:52 am | Reply


RSS feed for comments on this post. TrackBack URI

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at WordPress.com.

%d bloggers like this: