I'm not sure how this keeps happening. The first year I waited at a mall for 5 hours to get the original iPhone. The following year my friend Mark Rein convinced me to see a midnight showing of Hellboy II and then wait outside of an AT&T store all night to get the iPhone 3G. You'd think I'd learn by the third year but once more I was in line at the mall hours before the Apple store opened to get the 3GS. This year I thought it would be different. Apple offered free overnight shipping to anyone who wanted to pre-order the iPhone 4. Figuring everyone would go that route I decided to beat the FedEx trucks and just show up at the mall at 6AM. I'd be in and out in a little over an hour, which would give me a head start on battery life testing on Apple's 4th generation iPhone.

I promise that not all of my decisions play out this poorly. Those who pre-ordered the 4 and requested overnight delivery got their phones early and my one hour wait turned into six hours at the mall, for the fourth year in a row.

Apple's iPhone 4 with Bumper Case

It's a self fulfilling prophecy. Steve gets up on stage, proclaims the iPhone 4 to be the biggest introduction since the original iPhone, and the public flocks to Apple stores to fork over $200 on day one and around $2500 over the course of two years for the privilege. But this isn't 2007. Apple has real competitors in the smartphone space. Android phones have grown in features, polish and popularity. Even Palm entered the race with a competant offering, and Microsoft isn't far behind. It's easy to start a revolution when everyone else is doing the wrong thing, but what about when more companies actually get it? Was Steve justified in his excitement over the 4? That's what we're here to find out today.

Straight on it looks like just another iPhone. You get the black face with a shiny trim. From the side it is the redesign that Apple has needed for a while now. It’s not revolutionary but it’s the type of improvement that makes its predecessor feel old. And that’s exactly what this does. Have a look for yourself:

iPhone 4 (left) vs. iPhone 3GS (right)

The straight lines, smaller dimensions and lack of unnecessary bulk make the 3GS feel like a car from the 90s, unnecessarily curvy. The styling is now so much more compact. Compared to the iPhone 3GS the 4 is around 5% narrower (but no more difficult to type on) and nearly 25% thinner. It even makes the Nexus One look dated:

The iPhone 4 is slightly heavier than the 3GS (4.8oz vs. 4.7oz). You feel the added weight but I wouldn't call it heavy. The front and the back of the iPhone 4 are both made out of glass, and they protrude beyond the stainless steel band that wraps around the phone (more on this controversial decision later). While this gives the 4 an amazing finish, it also makes carrying the phone nerve racking. Coupled with the smaller, more dense form factor I’m now deathly afraid of dropping and shattering this thing. Apple has done a lot to reinforce the glass, however there have been enough reports already of shattered iPhone 4s for me not to feel very safe. Only Apple would think to make the two surfaces most likely to hit something out of glass. It's like making mouse traps out of cheese, something bad is bound to happen.

iPhone 4 (left) vs. iPhone 3GS (right)

The physical buttons (but not their layout) have changed on the 4. The ringer switch has shorter travel and feels sturdier as a result. The volume rocker has been replaced by discrete volume up/down buttons, also very sturdy in feel. The power/lock button is also now made out of stainless steel. Only the home button remains unchanged, although it does seem to make a deeper click when you use it.

The speaker moved to behind the right grill at the bottom of the phone instead of the left. The dock connector thankfully remained unchanged. It looks like Apple is committed to maintaining this connector until it makes the jump to something wireless (or optical?).

The back of the phone is pretty. Apple broke with tradition and finally included a single LED flash on the phone. The flash comes on in low light conditions and is enough to take shots in total darkness.

The camera has been upgraded to a low noise 5MP sensor. It can shoot stills at up to 2592 x 1936 or video at 1280 x 720 @ 30 fps. We’ll go into greater detail on its quality in the camera section. The iPhone 4 also adds a front facing camera capable of shooting both photos and video at 640 x 480.

Apple quotes contrast ratio as 1000:1, in our measurements we got very close (952:1). A significant improvement over the 188:1 ratio of the 3GS. Apple achieved this by both dropping black levels and increasing the white levels on the display. Improving both is always fine by me.

Internally the iPhone 4 uses Apple's new A4 SoC, built around an ARM Cortex A8 CPU and a PowerVR SGX GPU. The new SoC is built on a 45nm process and features 512MB of memory on the package. Apple hasn't made CPU clock speed public, but I'm guessing around 800MHz compared to the iPad's 1GHz for reasons you'll see later. GPU clock speed is unknown as well. Having more memory on package is an interesting move by Apple as it makes the iPhone 4 better suited for multitasking compared to the iPad. Also implying that shortly after the iPad gets multitasking it'll be updated to a version with more memory as well.

The iPhone now has an gyroscope as well the rotation sensors of its predecessors. Developers are given full access to the gyroscope making the iPhone 4 capable of becoming a very expensive Wii-mote.

The iPhone 4's logic board shrinks in size thanks to further component integration, making room for a much larger battery. The 5.25Whr battery in the iPhone 4 is a 16% increase from what was in the 3GS, and 95% of what HTC put in the EVO 4G. While raw performance improved, it's clear that Apple's focus this time around was battery life. Again, we'll dive into specifics later in the review.

Moving back outside Apple surrounded the phone with a stainless steel band. This band doubles as the 3G, WiFi and Bluetooth antennas. And if you hadn't noticed, it also moonlights as a giant elephant. Let's talk about it.

The Real Story on iPhone 4's Antenna

"How are you holding it?" - Brian Klug

"Three fingers. Like a [redacted] ninja turtle." - Anand Shimpi

There's been a ton of discussion lately surrounding iPhone 4 cellular reception. Even before it was officially announced, the reason for the stainless steel band running along the outside of the phone seemed enigmatic; many called it un-apple and decidedly atypical of seamless apple design which eschews hard edges. The black strips were written off as aesthetic curiosities, possibly even markings which denoted a fake.

Then at the WWDC announcement, we learned the truth. The iPhone 4's antenna is the stainless steel band that runs around the edge of the phone. The antenna for WiFi, Bluetooth, and GPS is the smaller strip beginning in the bottom left and running to the top, and the cellular radio for voice and data is the much larger strip running around almost three quarters of the phone.

It's a design nod back to some of the earliest cellular phone designs which packed external whips that one could manually extend for improving reception. Since then, designs evolved, and until recently virtually all smartphones have packed internal antennas at the bottom of the phone. The iPhone 4's external antenna promises improved reception over the internal antenna in the iPhone 3GS.

Of course, the caveat is that as with all external antennas, the potential for both unintended attenuation and detuning is much, much greater. When I first saw the iPhone 4's design spelled out watching the keynote online, I immediately assumed that Apple was going to apply an insulative coating atop the stainless steel. Perhaps even use diamond vapor deposition (like they did with the glass screen atop the iPhone 3GS) to insulate the stainless steel from users. We now know rather definitively that this isn't the case. Of course, the result is that anything conductive which bridges the gap in the bottom left couples the antennas together, detuning the precisely engineered antennas. It's a problem of impedance matching with the body as an antenna, and the additional antenna that becomes part of the equation when you touch the bottom left.

The fact of the matter is that cupping the bottom left corner and making skin contact between the two antennas does result in a measurable difference in cellular reception. But as we'll show, RF is a strange beast.

Measuring Reception Without Bars

When I set out to characterize and understand the iPhone 4's antenna issue, I noticed that reports online varied wildly. Some claimed that they were always able to recreate a reception issue created by cupping the phone, yet others reported no change at all squeezing the phone tightly. After acquiring my iPhone 4, the first thing I did was try to fire up Field Test via the widely documented *3001#12345#* dialer code. Unfortunately, like iOS 4 running on the 3GS and 3G, Field Test is absent from the iPhone 4. It isn't a matter of the dialer code, it's that Field Test has been completely removed from the applications directory in the filesystem.

For those that don't know, Field Test variants exist on virtually every phone for purposes of debugging the air interface and baseband. Quality metrics like RSSI (raw signal strength) usually in dBm are reported alongside a wealth of other metrics like SNR and even what adjacent towers are visible to the phone for handing off. It's a tool usually buried deep in every phone because the amount of data would overwhelm normal mobile users, but is useful for engineers and curious but savvy users alike to find out what's going on with the cellular network. For whatever reason, Apple really doesn't want anyone running that tool anymore.

Just about everyone knows that although reporting signal strength in bars gets the job done, it's an absolutely worthless metric for comparison across devices and platforms due to lack of standardization. Further, iOS smoothes the quality metric with a moving average over as much as 10 seconds, masking how fast signal changes. There's also the matter of dynamic range, but more on that in a second. Without any numbers at all it would've been impossible to understand what's going on with iPhone 4. On my 3GS, I exclusively report signal numerically, and as a result have a very good feel for coverage in Tucson, AZ where I live.

But I found a way. Undeterred by the lack of field test on iOS 4, I was determined to enable numeric signal strength reporting in the top left where bars are normally displayed. If you've ever run a jailbroken iPhone and used SBSettings, or changed your carrier string, you've probably encountered the fact that iTunes will back up and restore the status bar configuration across OS restores. See where I'm going?

I took my iPhone 3GS, downgraded to 3.1.3, jailbroke, enabled numeric WiFi and GSM and backed up. I then took my iPhone 4 and restored with iOS 4, but pointed it to the backup of the jailbroken, numeric-GSM-reporting iPhone 3GS. You'll note that booting and activating the new phone required fitting the new iPhone 4 microSIM into a SIM carrier. I ordered one almost a month ago, but it still hasn't shown up.

My MicroSIM -> SIM adapter. The real one is still inexplicably in the mail 15 business days later.

Success ensued, and I had a numeric readout of signal strength on a non jailbroken iPhone 4. The results are interesting.

Before we dive in, let's talk about dynamic range for a second. For a while, I've talked about how iOS reports the quality metric with a compressed, optimistic dynamic range. On iOS, 4 bars begins at around -99 to -101 dBm. Three bars sits around -103 dBm, 2 bars extends down to -107 dBm, and 1 bar is -113 dBm. To give you perspective, for a UMTS "3G" plant, -51 dBm is the best reported signal you can get - it's quite literally standing next to, or under a block away from a tower. At the other extreme, -113 dBm is the worst possible signal you can have before disconnecting entirely. With a few exceptions, signal power as low as -107 dBm is actually perfectly fine for calls and data, and below that is where trouble usually starts. However, you can see just how little dynamic range iOS 4 has for reporting signal; over half of the range of possible signal levels in dBm (from -99 dBm to -51 dBm) is reported as 5 bars.

So, an entire day and more than a quarter tank of gas later, here are the results. Holding the iPhone 4 without a case, in your left hand, crossing the black strip can result in a worst case drop of 24 dB in signal. As we'll show in a second, how you hold the phone makes a huge difference across every smartphone - and we've tested thoroughly in 5 different positions.

Now, there are two vastly different possibilities for what happens to the bar visualization after you drop 24 dB. I happen to live less than one block from an AT&T UMTS tower (it's across the street, literally), and have exceptionally strong signal in all of my house - it's part of why I chose to live here, actually. Signal is above -65 dBm in every single room, in most cases it's at -51 dBm. When I incur that worst case drop of 24 dB from squeezing the phone, I fall down to -83 dBm, which is still visualized as 5 bars.

However, in locales that have less signal, but where iOS still displays 5 bars, the drop of 24 dB is visualized much differently. For example, at another test location, signal without holding the phone is -89 dBm, which is still displayed as 5 bars. Cup the phone, and you'll fall all the way to -113 dBm. All the bars dramatically disappear one after the other, people think they've dramatically lost all the signal, and you know the rest.

If you're at 4 bars already, (which puts you on the low end of possible signal strengths), cupping the phone even more delicately is enough to push you the remaining 10 or so dB to cutoff. It doesn't take much when you're at 4 bars, which is why the visualization is flawed. Complicating matters is that signal is completely fine until down around 2 bars at -107 dBm.

If you add a bumper case to the iPhone 4, the signal strength drop from holding the device is on par if not better than other phones. In the exact same location, in the exact same orientation, I carefully measured my iPhone 3GS and Nexus One with the same AT&T microSIM in my newly made SIM adapter. After lots of testing, I decided on 5 different positions for holding the phone, and tested signal repeatedly. 

1) Cupping tightly - This is the absolute worst case and involves squeezing the phone very tightly, like people are doing online in videos demonstrating all the bars going away. I squeeze the phone hard and make sure my palms are sweaty as well. You'd never hold the phone this way because it's physically painful.

2) Holding naturally, comfortably - This is just how one would hold the phone typically in a relaxed way. Not squeezing it to purposefully reduce signal, but making contact with the fingers and not an open palm.

3) Resting atop an open, flat palm.

4) Holding naturally, but inside a case - In this situation the Bumper for iPhone 4, an Otter Box for the 3GS, and a comparable generic case for the Nexus One.

5) Pinching the top and bottom - Our baseline, virtually no attenuation. Held only to keep the exact position constant. It's not reported since this is considered ideal.

It's difficult to be exact about the data, since signal is very sensitive to direction, ambient conditions, and cell breathing. To generate these numbers, I measured at least 6 times and took the average. The results are pretty self explanatory. Inside a case, the iPhone 4 performs slightly better than the Nexus One. However, attenuation gets measurably worse depending how you hold the phone. Squeezing it really tightly, you can drop as much as 24 dB. Holding it naturally, I measured an average drop of 20 dB. 

The drop in signal from cupping the device with a case on is purely a function of us being "ugly bags of mostly water." A material which happens to be pretty good at attenuating RF - thus increasing path loss between the handset and cellular base station. There's nothing Apple nor anyone else can do to get around physics, plain and simple. It's something which demonstrably affects every phone's cellular reception.

That's not all there is to the story, however.

The Antenna is Improved

From my day of testing, I've determined that the iPhone 4 performs much better than the 3GS in situations where signal is very low, at -113 dBm (1 bar). Previously, dropping this low all but guaranteed that calls would drop, fail to be placed, and data would no longer be transacted at all. I can honestly say that I've never held onto so many calls and data simultaneously on 1 bar at -113 dBm as I have with the iPhone 4, so it's readily apparent that the new baseband hardware is much more sensitive compared to what was in the 3GS. The difference is that reception is massively better on the iPhone 4 in actual use.

With my bumper case on, I made it further into dead zones than ever before, and into marginal areas that would always drop calls without any problems at all. It's amazing really to experience the difference in sensitivity the iPhone 4 brings compared to the 3GS, and issues from holding the phone aside, reception is absolutely definitely improved. I felt like I was going places no iPhone had ever gone before. There's no doubt in my mind this iPhone gets the best cellular reception yet, even though measured signal is lower than the 3GS.

Conference call with three calls going at the same time, and transacting data, all at minimum signal. Impressive.

That brings me to the way that signal quality should really be reported - Signal to Noise Ratio (SNR). SNR is essentially a measure of how much of the signal is compromised by noise or interference. It's readily apparent that because the iPhone 4 works almost perfectly fine at -113 dBm, it has much better sensitivity. The deciding factor for reporting the signal quality metric is then SNR, something Apple and other handset manufacturers will have to move to eventually instead of just power. In reality, reporting based on SNR makes a lot more sense, since I couldn't make calls drop driving around an entire day cupping the phone, despite being at -113 dBm (1 bar) most of the time.

The drop in signal from holding the phone with your left hand arguably remains a problem. Changing the bars visualization may indeed help mask it, and to be fair the phone works fine all the way down to -113 dBm, but it will persist - software updates can change physics as much as they can change hardware design. At the end of the day, Apple should add an insulative coating to the stainless steel band, or subsidize bumper cases. It's that simple.

WiFi/Bluetooth Reception

But what about WiFi? Surely since the UMTS/GSM antenna interferes with it, the WiFi signal has changed as well. It has, though not how one would expect. Holding the phone with no case actually improves WiFi signal strength by a measurable 5 to 10 dB. In the following plot, the dips are me releasing the phone from a tight grip and going to the two finger pinch. I verified the same ballpark level of performance increase on the phone as well. RF is truly an odd beast indeed. It just depends whether you're adding or subtracting length from the antenna, and thus moving away from or closer to an optimal solution.

The blue line is the iPhone 4. Look at the rate as well - more on that in a second.

The last lingering question is how GPS fix accuracy and acquire time changes depending on how you hold the iPhone 4. I spent a few hours testing and came to the conclusion that there's an insignificant difference gripped or not gripped, or compared to the 3GS. It's also difficult to repeat the same measurement since location services seems to keep the GPS going even after you stop using it, so subsequent API calls to it within a few minutes are very speedy. 

Getting an accurate location is still nearly instantaneous using WiFi through skyhook, and then AGPS takes it the rest of the way. It's impressive that we're talking on the order of seconds for a location within tens of meters of accuracy - considering that a cold fix on a standalone GPS used to take minutes. If you don't have line of sight to the sky, GPS fixes will take longer no matter what smartphone you're using. I have to wonder whether improved WiFi reception has an effect or not on skyhook (WiFi MAC address and signal strengh based) trilateration accuracy. Again, I couldn't be certain.

Network Improvements

The iPhone continues to be an AT&T exclusive in the US. With the iOS 4 upgrades the iPhone 4 supports tethering over Bluetooth or USB. The feature is an extra $20 per month on top of your existing iPhone data plan and it also deducts bandwidth from the 2GB you get with the data plan. It's not a coincidence that AT&T timed the release of its iPhone tethering option with the move away from unlimited data plans. You don't have to give up your existing unlimited data plan if you don't want to, however if you want to enable tethering you have to sign up for the new $25/2GB plan.

I'm not a huge fan of iPhone tethering right now because despite paying for the service and having full signal strength on AT&T, I'm getting horrible transfer rates while trying to upload this article. I had to sign up for airport WiFi to get it live, thanks AT&T. When it does work however, it works well. As you'll see later you can easily get multiple Mbps out of AT&T's network in areas with good coverage. That easily equals the lower end of what you'd see from WiMAX today. As I've mentioned before however, it's really hit or miss with AT&T. The network is either great or totally unusable, while its competitors are generally more consistent but never quite as fast. I'd say that there's a good chance Apple will bring the iPhone to Verizon, it's just a matter of when.

AT&T's plans are actually reasonably priced if you don't go over the data limits. A $15/mo data plan will get you 200MB of transfers per month and $25 will give you 2GB. For users like my parents the 200MB option is great. Even for me personally, 2GB is fine. I tend to peak at 700MB per month, but that's because at the office I'm almost always on my desktop or connected via WiFi. It's unclear how tethering is going to change this for me. Our own Brian Klug on the other hand easily pushes more than 2GB of transfers per month. So AT&T's rate switch is either going to save you a few bucks per month or make you really unhappy.

The iPhone 4 brings HSUPA class 6 for upload speeds of up to 5.76 megabits/s to the platform. This is a 15 fold improvement over the 384 kilobits/s maximum of the iPhone 3G and 3GS, which I routinely see. Not all AT&T markets are updated to HSUPA, and in practice I saw uploads of around 1.5 megabits/s, in line with class 2 or class 3 HSUPA.

Downstream HSDPA speeds remain unchanged from the 3GS, supporting up to 7.2 megabit/s HSDPA. I'm lucky since my market is HSDPA 7.2, as I routinely see speedtests of 5 megabits/s or above very late at night when there isn't very much plant load. I haven't seen any measurable increase in speeds over the 3GS, except in locations with very low signal as noted before.

My fastest iPhone 4 speedtest so far

Even Anand in one of the slowest 3G cities in the US saw significant improvements with the iPhone 4. AT&T's network in general seems to be improving. While the best he'd been able to achieve was around 1Mbps a year ago, these days he can break 2.5Mbps down during the evenings.

iPhone 4 speedtest in Raleigh, NC

WiFi speeds have also improved, as Apple has added 802.11n in the 2.4 GHz band. 5 GHz support remains absent, something which would have likely complicated antenna design even further. Interestingly enough, the BCM4329 WiFi and Bluetooth SoC does contain 802.11a 5 GHz support, the reason it's disabled is again likely due to antenna design constraints. The Broadcom SoC also includes an FM tuner and transmitter, though support for either remains and sadly (given Apple's historical lack of FM radio support) will remain such. Bluetooth 2.1 EDR is there as well.

The iPhone 4 seems to connect at 802.11n rates of 72 megabits/s in best case, far from the maximum without channel bonding of 150 megabits/s. This is still a welcome improvement from the 802.11b/g in the iPhone 3GS, which seemed to never connect above b rates in practice. As an aside, mobile devices using 802.11b rates (modulated using DSSS) are a huge contributing factor to WiFi congestion at conferences - I've even seen DSSS modulated rates (and thus 802.11b devices) explicitly disallowed from connecting to APs at conferences. It's a welcome improvement to see iPhone bringing N support.

A typical WiFi test result

However, even on my 25/4 DOCSIS 3 cable connection, I could only squeeze out a maximum of 8.5 megabits/s down and 8.0 megabits/s up while connected at 72 megabits/s best case. This was using the speedtest.net app to a local test location. I tried with an Airport Extreme (new generation), a WRT600N running DD-WRT, and a WRT54GL-TM running Tomato. All three showed similar results capping out around 8 megabits/s down when I could run tests in excess of 30 megabits down on my desktop. This is probably more of a CPU bottleneck appearing than anything else.

Update:

I thought there was something wrong with my WiFi performance, turns out the iPhone 4 is indeed faster than a palty 8 megabits/s. ;)

There were a number of comments by folks who were able to run speedtest.net and get throughput above 10 or 11 megabits/s. Testing earlier today on a much faster connection, I managed to get something in line with their numbers:

Early today, the folks at DigitalSociety also managed to get much faster WiFi speeds in the neighborhood of 20 megabits/s, way higher than my meager 8 megabits/s. To do so, they loaded an MP3 in safari stored on a local webserver and watched network utilization. I wish I had thought of this, because it's perfect. In the past, the speedtest.net app always used to saturate my connection over WiFi. My only explanation is that the application performs slower over WiFi in iOS 4 than it did in iOS 3, something Anand noted as well. Thanks for all the heads up messages, everyone!

I set up a similar test to DigitalSociety's. I opened an 85 MB PDF stored on my local web server in the browser of each device and watched network utilization using bwm-ng. I tested with an Airport Extreme connected over gigabit to my webserver with no other network utilization. I took the average of 5 runs on the iPad, iPhone 4, iPhone 3GS, and my AT&T Nexus One running Froyo 2.2 (which is 802.11n). The results are much, much more in line with earlier expectations.

The iPhone 4 comes close to but can't quite best the iPad, though the difference is minimal. I did notice that the iPad associates at the same 72 megabits/s connection speed as the iPhone 4. Thankfully, the iPhone 4 easily bests the iPhone 3GS. Finally, although the AT&T Nexus One associates at an 802.11n rate of 65 megabits/s, it's slower than the iPhone 3GS. I have a feeling the device is writing the PDF into flash, whereas the iOS devices are loading it into memory.

Screen - Retina Display

Right out of the box, the iPhone 4's new 326 PPI, 960x640, 3.5" display is arguably the single most striking change the new iPhone brings. In a word, it's dazzling. Text and high res images look amazingly sharp on the iPhone 4’s retina display. It’s an improvement over the 800 x 480 AMOLED screens that have been shipping on most Android phones. But if you’re comparing it to an iPhone 3GS the difference is huge.

iPhone 3GS	iPhone 4

Text on the Google Nexus One

Text on the iPhone 4

The dot pitch is truly remarkable, so much so that Apple makes the claim that their display outresolves the human eye; its advertised ability to do so has earned it a new Apple tradename, "retina display."

Text on the HTC EVO 4

Text on the iPhone 4

AnandTech Logo on the EVO 4G

AnandTech Logo on the iPhone 4

Immediately after hearing Apple's claim that the Retina Display outresolves the human eye, I snapped into optics mode and crunched the numbers, and tweeted that the results were valid.

In the days that followed, there was considerable debate about the validity of Apple's claims. However, nearly all of the debate really just hinged on a debate over angular resolution of the human eye, and a little more over viewing distance. They're both entirely conventions.

As you've probably discovered by now, the human eye resolution can really only be characterized in angular subtense. Hold something closer to your eye, and you can see smaller features better (in theory), move it further away, and you can't make out small spatial details. The minimum angle visible with the human eye is the angle at which features (for the most common definition, a black and white square wave) stop being visible, and are indistinguishable from each other.

Most measures of visual acuity test with this implicitly - the Snellen eye chart's use of the capital "E" is literally a perfect example, which has given rise to a "tumbling E" eye chart. At twenty feet, the capital E subtends 5 minutes of arc, and conveniently has five half cycles of white to black (from top to bottom). So 20/20 implicitly implies an angular resolution of 1 arcminute (1/60 degrees).

As an Optical Sciences and Engineering undergrad, I've had 1 arcminute drilled into my head more times than I can count as being the "normal" angular resolution of the human eye system. In practice, this is 20/20 vision, which is "normal," yet not perhaps the absolute maximum for human perfection. We can play games of course and argue that a small subset of the population has better than normal uncorrected vision, and thus an angular resolution of below 1 arcminute. I have above average uncorrected vision, which I've measured to be 20/15 on average, giving an angular resolution of approximately 0.75 arcminutes. Of course, the definitions stem from the spacing of cones in the fovea, the highest resolution part of the retina.

The other informational quantity needed to test the Retina Display claims is viewing distance. Again, there's a commonly agreed upon convention - standard viewing distance is considered to be 1 foot. This is another drilled into my brain number tossed around for comfortable viewing and reading. In practice, you can focus on objects much closer to your eye - this is called the near point and is often given as 10 inches, though as you get closer you increase strain aren't likely to keep it here.

Maybe not exactly the limit, but close enough.

Given the two most common standards tossed around, 1 arcminute and 12 inches, do the math out and you'll arrive at around 286 pixels per inch as the limit for eye resolving power, comfortably below the 326 on the Retina Display. Move to 0.75 arcminutes at 12 inches, and it's 382 pixels per inch, higher than the Retina Display. Honestly, I can't see the pixels at 12 inches.

Of course, the real story is even more complicated. Remember how the definition comes with the implicit assumption that we're dealing with a square wave pattern from white to black? That's a factor too - the contrast of the two pixels. Lower the contrast, and the eye's ability to pick out features decreases even more. So far, everything we've talked about has been first order, and without aberrations. Toss in spherical and astigmatism, two aberrations common to the eye system, and eye performance drops way more.

The human eye system is actually pretty poor, and shockingly easy to outresolve. In fact, if you saw the image your eye forms on your retina, you'd likely be appalled; it's your brain that makes the system usable. But at the end of the day, Apple's claims that the display outresolves the human eye are good enough for us.

More Display

It's obvious how Apple settled on 960x640; it's four times the resolution of the previous iPhones. However, instead of decreasing text size, iOS 4 scales appropriately, and the size of everything remains the same. The result is that there are small details everywhere that pop out. Apple's icons on the home screen are the first that really pop out, and new iOS 4 optimized applications will bring that increased detail as developers add higher resolution artwork.

The display panel itself uses a subset of IPS (In Plane Switching) display technology called Fringe Field Switching (FFS). Where IPS switches the crystal polarization in the plane of the display with two opposing electrical substrates composed of semi opaque metals (which decreases transmission and viewing angles), FFS uses considerably less metal by arranging the electrodes in a comb like structure.

See that - it almost looks like a comb. Or an impulse train. Or Dirac comb. So many combs.

The result is that there's considerably less metal in back and in front of the pixel, resulting in much higher transmission of light through the display, and higher brightness for a given backlight level. Using FFS to drive pixel switching is critical here because of the high dot pitch in the iPhone 4's display.

The other interesting difference between iPhone 4's retina display and previous displays is that the digitizer is in optical contact with the display itself. There's no longer an air gap, and as a result, no longer any opportunity for dust to gradually work its way inbetween. Over time, I've noticed a few dust specks creep in on my 3GS, it does happen. The digitizer and display panel are essentially laminated together. The added benefit is that fewer material interfaces results in fewer internal reflections - think the "super" in Super AMOLED but applied to TFT. That's what Apple has done here.

Apple is using Corning's Gorilla glass which touts hugely increased scratch resistance and robustness. Both the front and back of the iPhone 4 are that same type of glass. I've noticed a few superficial scratches (called sleeks) that have appeared on the back, but really the true test will be how the phone looks after 6 months in the pocket. It's interesting that the iPhone capacity markings have disappeared from the back of the phone - no doubt this was done so Apple could make one part and one part only for each color.

The rest of the details are in the specifications. Apple advertises increased brightness of 500 nits and a contrast ratio of 800:1. We measured, and our iPhone 4 exceeded specs at 571 nits and just under 1000:1 contrast ratio.

Note that the HTC EVO 4G is missing as Anand has it, but it's on its way to me. As soon as I get it, I'll measure display brightness, black point, and contrast and update these results. In addition, the HTC Droid Incredible (and thus Nexus One) contrast is effectively infinity by the way we calculate, due to pixels being completely turned off in the black state. In addition, I'm led to believe that the AMOLED's PenTile grid throws off our numbers when measuring brightness. I've run and rerun this test, it keeps coming up that way. 

Next to the iPhone 4, the 3GS display really shows its age. It leaks light when displaying black, with an absurdly high black point of nearly 3 nits. Just looking at the lock screen on the iPhone 4 next to the 3GS it's readily apparent how much better blacks are. iPhone 4 easily bests the 3GS but still isn't quite as contrasty as the Incredible or Nexus One AMOLED displays, or the IPS in the Motorola Droid. You do get higher resolution and brightness, however, but nothing is free.

Going from the iPhone 4's display back to the 3GS is pretty painful, but going back to even relatively high dot pitch displays on the desktop is painful as well. Even on the "high resolution" MacBook Pro with 1680x1050 display, displaying an iPhone 4 screenshot at native resolution uses up 91% of the height. If there's one thing I hope the iPhone 4 display does, it's generate demand for 300 PPI level desktop displays - the era of 110 PPI displays being the norm needs to end now.

The Display in the Sun

As we mentioned earlier, the new glass is more reflective than the 3GS' screen. In direct sunlight the 4 is somewhere in between the usability of the iPhone 3GS and an AMOLED Android phone. The photos below should help convey that.

All of the screens were set to max brightness with automatic adjustments turned off. The EVO 4G came in as the worst of the bunch, while the iPhone 3GS was the most legible.

The iPhone 4 in direct sunlight

The iPhone 3GS in direct sunlight

The Google Nexus One in direct sunlight

The EVO 4G in direct sunlight

Welcome to 2010, Apple Upgrades its Camera

The iPhone 4 is equipped with two cameras: a 5 megapixel camera with LED flash on the back of the phone and a VGA camera with no flash on the front. The LED flash works in both still and video modes. Like the EVO 4G, the iPhone 4‘s flash allows you to shoot in perfect darkness. If you’re filming a video in low light the LED will stay illuminated while you’re recording.

Taken with the iPhone 4 in total darkness

The same unfortunately can’t be said for the front facing camera on the 4. In anything but good lighting you’re going to get noise. It’s really only useful for FaceTime (or as an alternative to a mirror) and even then you need to be well lit for it to look decent.

Apple has opted for a 5 megapixel OmniVision sensor for the rear camera on the iPhone 4. What's interesting is that Apple has decided to bring backside illumination front and center with their marketing.

Backside illumination improves the sensitivity of CMOS and CCD detectors by reducing the amount of material in the path of incident light. In a frontside illuminated detector, a considerable amount of light is lost due to absorption that doesn't result in emission of an electron, in addition to reflection off pixel structures and electrical components near the frontside surface. Backside illumination greatly improves sensitivity by flipping the stack over. Instead of light having to pass through and possibly be reflected by metal structures, it is converted into electrons and read out by passing solely through silicon. Creating a backside illuminated part isn't as simple as flipping a sensor over, however, as manufacturers also generally thin the silicon light has to pass through before it can reach the photodiode. This further improves sensitivity and is generally accomplished through chemical etching in acid or by lapping (physically grinding) sensors at wafer scale.

OmniVision OV5650 - iPhone 4's rear camera SoC

Though backside illumination (BSI) improves quantum efficiency (how many photons are converted into electons), backside illumination is hugely important for another serious reason as well. Because the sensor is small at 4.6 mm by 3.4 mm, pixel size is also extremely small at just 1.75 microns square for the OV5650 in the iPhone 4 (state of the art sensors are 1.4 microns square, like those in the HTC Incredible's 8 MP sensor). Frontside illuminated parts generally have in the neighborhood of 10-15 microns of silicon before the active region of the photodiode where one wants photons to get converted to electrons. The result is that without backside illumination, pixels have a 10:1 ratio of height to length, you can visualize them as looking something like long square pillars. But that's a problem.

As photons are converted into electrons in that silicon, there's no guarantee that it will immediately travel down into the gate structure below to be read out by the camera. Electrons drift as they descend these columns, meaning that photons incident on one pixel don't necessarily map to the gate below. Because the smartphone camera sensors are so small, with a 10:1 ratio of height to size, the result is large amounts of so-called quantum blurring from electrons traveling into the gate structures of adjacent pixels. The result is a blurry image (and a decrease in MTF at the sensor level!), thus not representing the image that used to be incident on the sensor.

OmniVision and other smartphone CMOS sensor manufacturers thin that column down in an effort to come closer to having the pixel look more like a cube than a huge pillar. Ballpark numbers are between 3 and 6 microns, down from 10-15. The result is much more sensitive sensors that are higher resolution. While megapixels don't necessarily matter, neither does pixel size as much anymore; it's all about quantum efficiency, which is what engineers really care about.

OmniVision BSI - Courtesy OmniVision

The optical system of the iPhone 4 is difficult to characterize without disassembly, though the focal length is a bit shorter than previous iPhones. The result is that the photos are demonstrably wider angle. Backside illumination also allows for a bigger chief ray angle, higher numerical aperture (and thus lower f/#), but I won't bore you with the details.

Camera Usability

It still takes almost two seconds to activate the camera on the 4, which is enough time to miss whatever it is you’re trying to grab a photo of. Thanks to the iOS 4 update however, the shutter is almost instantaneous. The difference appears to be that the photo is committed to memory but not fully written to NAND, whereas before the photo would be written to the Flash before you could take another picture. Power loss in the middle of snapping photos seems pretty rare on a smartphone so the tradeoff, if I’m correct, makes sense.

Apple opted for a lower noise rather than higher resolution sensor in the iPhone 4 and it did pay off. The main camera shoots photos at 2592 x 1936 compared to the EVO 4G’s 3264 x 1952, but the resulting images are far less noisy - particularly in low light situations:

HTC EVO 4G - Low Light

Apple iPhone 4 - Low Light

The 4’s main camera, like the HTC Incredible and EVO 4G, is a decent replacement for a point and shoot if you’re primarily outdoors. You’re still going to get better image quality out of a good point and shoot, but the tradeoff is convenience. The limitations are significant.

Because you rely on the iPhone 4’s software controlled aperture and shutter speed you don’t have the ability to properly expose the image. You have to rely on Apple’s algorithms, which tend to either overexpose outdoors or miscalculate white balance with non-halogen light sources.

This is an example of a photo taken outdoors that’s more washed out than it needs to be:

And many of you picked up on the white balance issue I snuck into our EVO 4G review yesterday:

Regardless of where I tapped to focus, I could not get the iPhone 4 to set a proper white balance in our photo box.

While I was watching the screen, the iPhone 4 would alternate between yellow and white for the background color. It seemed to be trying to calculate the white point but was just being thrown off by the type of light. If I timed the shot right I could snag the photo while the iPhone was switching between white balance points:

I also had this problem in my office which uses LED can lights.

This is far more yellow than it should be

While Brian didn't have the same problems I did, Brandon Hill (DailyTech Editor in Chief) did. It seems to be very dependent on the type of lighting you have and even then it seems to vary based on the type of CFLs. And unlike the EVO 4G, there’s no way to manually set a white balance on the 4.

For overall image quality though I have to hand it to Apple, the iPhone 4 does do a better job than the EVO 4G or other phones I’ve used. Take a look at this shot inside my house:

The colors in the iPhone 4 shot are on point. The green is correct, the wooden floor is right and the black is, well, black. The EVO 4G didn’t do so well on this test by comparison:

The 4’s camera isn’t perfect, but it does appear to handle colors better than the EVO (with the exception of my white balance issue) and delivers lower noise photos.

Compared to other phones the 4 does similarly well, besting the 5 megapixel camera in the Motorola Droid easily in terms of color reproduction and sharpness. Though the HTC Incredible previously was a top performer alongside the N900, the iPhone 4 makes the Incredible look a bit oversharpened and artificial. Compared to the 3GS, the iPhone 4's improvement is obviously dramatic, as shown in the gallery below.

Video is recorded at 1280x720, in H.264 with AAC mono audio. We measured a bitrate of 1.35 MBps, outclassing all the other smartphones we've tested.

iPhone 4

What's interesting is that the iPhone 4 appears to crop the sensor down for video recording, taking the center most 1280x720 pixels instead of scaling down the entire image size. The result is that the focal length for video recording is notably longer than when taking photos.

You can see the difference is quite notable standing in the same place. Perhaps the A4 SoC lacks the compute power to apply a scale and encode at the same time, necessitating this crop. Whatever the case, video shot with the iPhone 4 still looks very good at the promised and delivered 30 FPS. Move the camera around enough, and there's still screen door effect from the rolling shutter like any CMOS sensor is going to give you - it's a fundamental problem no phones are going to get around soon. Its also right there in the specifications page for the camera SoC; rolling shutter.

Similarly, iPhone 4 does give you 5x digital zoom, though we still maintain you're better off taking photos at native resolution and messing with them later with better interpolation algorithms.

FaceTime

Meet Manveer. I’ve known him since I was in the 6th grade. Somewhere around the 8th grade we started a ritual of calling each other every day after school and talking about video games, computers and dumb things that happened at school. We talked on the phone for hours. Back in those days we would even play the same game on separate computers while talking on the phone. It was a precursor to voice chat in gaming networks like Steam or Xbox Live. It was ridiculous amounts of fun.

Manveer is up in the corner, the crazy guy is yours truly

Tonight Manveer and I FaceTimed. It’s the feature I wish we had when we were back in middle school. These days it’s a lot harder to explain why you’ve got some random dude’s face on your phone talking to you about completely pointless things at 12:54AM on a Tuesday night. I spent part of the chat covering one of the lights in my room with my head and calling it an Anand Head Eclipse. At the risk of further embarrassing myself publicly, it was fun.

For those of you who don’t know, FaceTime is Apple’s VoIP protocol built into the iPhone 4.

Granted we’ve been able to do this sort of thing for quite a while now over instant messenger networks. Integration into your smartphone is just the next logical step, and in the right conditions it works very well.

The first requirement is that both users need to have an iPhone 4 obviously. Both also need to be on WiFi. While the FaceTime icon will appear if you call another iPhone 4, if you try to activate the connection you’ll get this error unless you’re on WiFi:

With a bit of poor UI design Apple will actually display a FaceTime icon with a question mark in it if you call another iOS phone. Trying to activate FaceTime however gives you an error.

FaceTime requires roughly 100 - 150Kbps of bandwidth in both directions to work smoothly. The download should be fine but the upload is pretty high given that many broadband providers in the US are ridiculously stingy with their upload bandwidth. For FaceTime to work well you can’t be uploading or downloading anything large in the background, or if you are just throttle everything else to give you enough bandwidth to work with.

The next problem with FaceTime is the iPhone 4 lacks an absolutely necessary integrated stand. HTC got it perfect with the EVO 4G, unfortunately the 4 has no such thing. The closest you can get is Apple’s iPhone 4 dock, it’ll set you back $29 but it’s necessary to prevent you from getting tired holding your phone out in front of you. Laying the 4 down on your desk while you FaceTime just gives the person on the other end of the line a great view of your nostrils. Not very welcoming.

This is the on-table view of FaceTime, you need a stand

Even once you’ve met all of the requirements it’s still not a guaranteed thing. Even with ample bandwidth I had one FaceTime chat stall in the middle of the chat. My incoming feed froze and I had to wait a couple of seconds for it to resume. I also had a problem where FaceTime would fail to start on the first try. A subsequent try fixed it.

As you'll see in our camera investigation, the low light performance of the 4's front facing camera is horrid. Couple that with FaceTime and you will get bad image quality if you're not in a well lit room:

Ugh, what is this, 1998?

While in FaceTime you can easily switch between front and rear cameras by tapping the camera icon in the lower right corner. Unfortunately the compression on the video is enough to render text illegible while in FaceTime:

This was taken using the rear facing camera over FT

Obviously everyone knows where FaceTime is going. Apple is supposedly opening the protocol up to all developers, so you’d be able to theoretically build desktop and other smartphone FaceTime clients. And ultimately as mobile broadband speeds (hello WiMAX and LTE) improve the WiFi requirement will be dropped. But is it a good feature today?

It really depends on who you know with an iPhone 4. If you’ve got a Manveer, absolutely - the feature is worth it. If you have a bunch of casual acquaintances with iPhone 4s, probably not. It’s only useful if you know the person on the other side very well.

There is one other major benefit to FaceTime. Once enabled your call is routed over WiFi to the Internet, not AT&T’s 3G network. It doesn’t use any of your plan’s minutes and more importantly, voice quality is much improved over a regular 3G/EDGE phone call. It’s VoIP, not a crappy cell connection. Even if you just cover the camera it’s actually better to make calls over FaceTime than 3G based on the sound quality alone.

Sidebar: Luxa2 H1-Touch, a Great FaceTime Stand

Not too long ago I visited Thermaltake while I was in California. Thermaltake recently spun off a brand called Luxa2. This brand would focus on more lifestyle products, leaving Thermaltake to the typical PC stuff like cases and powersupplies.

One of Luxa2’s first products was the H1 Touch, an iPhone/iPod Touch cradle. The idea was to be able to hold your iPhone while you’re in the kitchen so you could read recipes or emails without potentially dumping a bunch of flour or oil on your phone. When the 4 was announced, Thermaltake was quite pleased - the H1 Touch would be perfect for FaceTime.

Indeed it is.

The iPhone 4 sits comfortably against a cushioned black pad that holds the back of the phone very well. There are six rubber feet that hold the outside of the phone, they move as one and allow you to maintain a snug fit. The 4 isn’t falling out of there. The only problem here is that one of the feet obstructs the volume up button, so you’ll have to take the phone out of the H1 to increase volume.

The cradle can tilt and rotate, making it far more flexible than the Apple dock for FaceTime. If you are planning on putting in some hours with FaceTime, I’d say the H1 (or perhaps the new H2?) is a must have. It sells for $29.99, the same price as Apple’s dock. It isn’t technically a dock but the connector is still exposed and the cable still works, it’s just not as neat as Apple’s solution. Way more functional for FT though.

Speakerphone Volume

In my N900/Motorola Droid piece, I alluded to an interesting note I made about the iPhone 3GS's speaker volume. For a long while, people have complained that the iPhone's speakerphone volume was too quiet, complaining that even at the maximum volume it was nowhere near loud enough. The same applied to noise level over the handset through the earphone speaker.

I've noticed a similar issue, but picked out another detail while testing the AT&T MicroCell - calls on 2G are louder than calls on 3G. After I got my soundmeter, I set out about measuring, and found that my suspicions were confirmed. Calls on 2G GSM are indeed significantly louder than calls on 3G UMTS on the iPhone, and curiously not so on the Nexus One. The reason? Dynamic range compression.

If you're familiar with the FM loudness war, than you'll instantly understand what's at play with 2G versus 3G call loudness. Calls placed over GSM have a smaller dynamic range, while 3G calls over UMTS have greater dynamic range (and fidelity). The dynamic range of GSM is roughly equivalent to CDMA 1xRTT (which is what all calls on Sprint and Verizon are placed over - not over 3G Ev-Do), though that's a whole other can of worms to discuss. Regardless, when users unfamiliar with the inherent difference in fidelity hear a 3G call after getting accustomed to 2G loudness from 1xRTT or GSM, it sounds notably quiet on the iPhone.

This is a problem that's existed since the iPhone 3G, and has gone unchecked in the 3GS, and 4. For whatever reason, Apple isn't increasing the gain on 3G to match the same loudness, possibly to preserve fidelity, and the result is that the speakerphone and earpiece are never as loud as they really could be compared to a 2G call. As I mentioned before, Android on the Nexus One appears to use an automatic gain algorithm to normalize loudness. If you place a 3G call and listen carefully, you can actually hear the gain ramp up to meet the same loudness as the 2G GSM call.

The difference between these phones might not look like much, but keep in mind the logarithmic nature of dB: -3 dB is half the loudness. 

On the iPhone 3GS, the difference is 7.36 dBA, while the iPhone 4 somewhat lowers it down to a still-audibly-different 3.31 dBA. The Nexus One shows no difference in loudness between 2G and 3G, correcting both to a (likely purposefully exact) 80 dBA. The iPhone 4 is indeed louder than the 3GS, by 4 dBA. It's a difference, but not an overwhelming one. I'd say the iPhone 4's speakerphone is still loud enough, though calls over 3G are still a bit too quiet. Until Apple increases the gain on 3G calls, iPhone 4 customers who are hard of hearing should invest in a bluetooth headset.

Performance

When the 4 was announced, my curiosity (and perhaps yours as well?) was about the flavor of Apple’s A4 SoC inside the phone. The rumor was that the A4 in the iPhone 4 had a full 512MB of memory, compared to 256MB in the iPad’s A4. iFixit took an iPhone 4 apart (with relative ease I might add) and confirmed the presence of 512MB. Look at the A4 in this picture, the Samsung part number on the right edge of the chip starts out K4X4G. The K4X denotes a Samsung mobile DRAM and the 4G refers to its density: 4Gbit or 512MB.

More memory should mean a smoother multitasking experience on the iPhone 4. Remember that iOS 4 keeps all fast task switching enabled apps resident in memory even after you’ve switched away from them. They are only kicked out of memory if you run low or if you force quit them. With more memory you should be able to keep more apps in memory without unwillingly forcing them out. The 512MB of RAM should also give the OS more breathing room in lighter multitasking scenarios, a problem many are already seeing on the iPhone 3GS running iOS 4. In practice the 4 is smoother when running iOS 4. There are still some hiccups but not as bad as on the 3GS, and definitely not as bad as on the 3G. The OS was clearly designed with the iPhone 4 as the performance target.

The big question that remains is what clock speed the iPhone 4’s A4 is running at. The assumption was 1GHz however Apple was very careful not to mention clock speed, unlike at the iPad launch. Saying nothing usually means there’s nothing good to say.

I ran a whole bunch of benchmarks on the iPhone 4 and the iPhone 3GS running iOS 4 to try and figure out what CPU speed we might be dealing with. The early reviews imply that the A4 in the iPhone 4 is running at a speed greater than the 3GS’s 600MHz, but slower than the iPad’s 1GHz. Let’s try and find out what it’s really running at.

We’ll start with the real world tests, first SunSpider. A javascript performance test this benchmark is completely network independent but it measures the performance of the browser as well as the underlying hardware. It is small enough to make memory size differences negligible so between the two iPhones we should be seeing a pure CPU comparison:

The iPhone 4 manages a time that’s 25% faster. Note that this test is just as much about the software stack as it is about CPU/platform performance. The Froyo update makes the Nexus One ridiculously fast in this benchmark. It just shows you how much room there is to improve performance on these Android handsets. The next leapfrog is going to be once the entire Android world moves to 2.2.

Next up is the Rightware BrowserMark. This test combines JavaScript and HTML rendering performance:

BrowserMark spits out an overall score but with no indication of what the score actually means. In this case we’re looking at 18.4% better performance than the iPhone 3GS. The iPad is 34% faster than the 4, which supports the theory of the A4 running at ~750MHz in the new iPhone.

There's also the possibility that the A4 CPU clock varies depending on load and other factors but the run to run consistency in all of our tests seems to indicate otherwise. We also can't ignore the fact that the iPad and iPhone now run vastly different OS revisions. In some cases iOS 4 actually takes a step back in performance compared to iOS 3.2. That undoubtedly makes the iPad vs. iPhone 4 comparison about much more than CPU performance.

Froyo's improved Javascript performance sends the Nexus One nearly to the top of the list here, only bested by the iPad. While Apple has definitely improved performance with the iPhone 4, it seems that it will only take a software update for Android phones to surpass it.

To measure web browsing performance I downloaded a bunch of different web pages and saved as much of them as possible locally on a server. I used WiFi on all of the devices to connect to my local server and timed average load time. I repeated the test at least 3 times and threw out any unusually high or low numbers. Performance was from a clean restart with no additional programs running in memory.

Note that these numbers aren't comparable to other reviews as we've updated software versions on two of the phones. The iPhone 3GS is now running iOS 4 which resulted in some numbers going up while others went down. And the Google Nexus One is running the officially released build of Android 2.2, codenamed Froyo.

What these tests should show is the overall performance of the platform when all network bottlenecks are removed. Obviously hiding in a tunnel under a lead umbrella will make any phone slow, but we’re looking at peak performance here.

The first test is the new AnandTech front page. Here we’ve got tons of images and HTML, meaning we’re stressing both bandwidth and code parsing speed.

The iPhone 4 is no faster than the 3GS (actually slightly slower, but we'll chalk that up to timing variance) here. The Froyo update to the Nexus One makes it lightning quick, almost as fast as the iPad in our first test.

Next up we have the first page of our recent Zotac XBOX HD-ID11 review. The balance shifts from tons of images to more HTML processing:

Here we see more of what I expected: the iPhone 4 is around 25% faster than the 3GS running iOS 4. Android 2.2 running on the Nexus One is basically as fast as the 4.

Using our Print this Review function, this next test loads our entire 2010 15-inch Macbook Pro review. While the other two tests had some flash ad content, this one is completely devoid of it so the HTC phones shouldn’t be penalized:

Here the iPhone 4 is 11% faster than the 3GS and about the speed of the iPad. There are other bottlenecks at play here so we don't get further performance scaling. The Froyo update helps the Nexus One a bit but the iOS devices are still quicker.

Our most intensive test is up next with a load of the Engadget front page:

The Nexus One is ahead of the 4 once more with its Froyo update. And the iPhone 4 is 34% faster than the 3GS.

Our most CPU bound test is up next. I put together a custom page with a ton of tabular content and a single page copy of our 15-inch MBP review to make the load take some time at least.

Surprisingly enough there's no difference between the 3GS and the 4 here, perhaps my test is less CPU bound than I thought. Froyo improves the Nexus One's performance a little bit.

Low Level Synthetic Tests

If we assume that we’re mostly CPU bound in all of these cases (a fair assumption given how fast Atom can run through all of these tests), then we’d be looking at a ~750MHz clock speed for the iPhone 4’s A4 assuming no other architectural changes. That’s actually a pretty big assumption. The A4 is widely believed to be a 45nm SoC using an ARM Cortex A8. At 45nm there should be room for a larger L2 cache than what was used in the iPhone 3GS’s SoC.

Perhaps some more synthetic tests will help us figure out what’s going on. I turned to Geekbench, now available in an iOS version.

Geekbench spit out a number of overall results that gave me a good enough summary of what’s going on to make an educated guess:

The CPU specific tests all indicate the iPhone 4 is around 25% faster than the iPhone 3GS. That would imply at least a 750MHz clock speed if all else is the same. Assuming we don’t get perfect CPU scaling with all of these tests, I’d venture a guess that 800MHz is more accurate. If the A4 does indeed have larger caches however, Apple could get away with a lower clock speed.

The memory results are particularly telling as they all scale very well going to the iPhone 4, better than the CPU results in fact. This could lend credibility to the theory of larger internal caches or perhaps to an improved (faster) memory subsystem.

Unfortunately until we get the iPad on iOS 4 we can't get a better idea of CPU scaling. I'm not even sure how reliable that will be at this point. If Apple was willing to change the amount of memory the A4 package housed between the iPad and iPhone 4, who is to say that it wouldn't have a slightly different design for the iPhone 4 (e.g. larger caches). The designs may not be physically different but we may instead be looking at binning. Given Apple's unwillingness to talk about the architecture here I think the safest bet is that we're looking at an 800MHz ARM Cortex A8 core in the iPhone 4 and a 1GHz core in the iPad.

Why the lower clock? It's all about battery life.

Incredible Battery Life

I’ve publicly praised Apple for its honesty in reporting battery life over the past couple of years, and the iPhone 4 gives me no reason to stop.

The 4 has an integrated 5.25Whr battery. That’s around 95% of the battery capacity of the HTC EVO 4G, despite having much lower power frequirements. It’s also a 16% increase over the 4.51Whr battery that was in the iPhone 3GS. This means at bare minimum, assuming the iPhone 4 doesn’t draw any more power than the 3GS, we should get 16% more battery life.

In reality, we get much more.

When Apple introduced the iPhone 3G it dropped battery life to a level that we’d been stuck at ever since. The 3GS improved battery life a bit through better CPU power efficiency but you still didn’t have enough juice to get through a day without charging.

The iPhone 4 changes all of that. The combination of a larger battery and a more power efficient SoC results in an incredible amount of battery life.

Our first test is a basic web browsing benchmark. We've scripted almost two dozen webpages to load, pause for 20 seconds, then forward on to the next page. None of the pages use any Flash. This process repeats until the battery is dead. Screen brightness on the iPhones was set to 50% and the screens remained on the whole time.

Battery life improved nearly 38% with the iPhone 4. It's clear that while the A4 improved performance, the real improvement was in battery life. This test has enough idle time where good power management and low idle power can really impact the results. There's simply no other similar smartphone that can touch the 4's battery life.

We then repeated the same test over WiFi instead of 3G:

Apple claims the iPhone 4 will last for up to 10 hours over WiFi, we measured just under that at 9.96 hours. The improvement here is only 12.8%, which tells me that we're nearing the limit of how efficiently Apple can manage power in WiFi mode. There's a wall that we're quickly approaching with this current architecture.

To measure talk time we play MP3s on repeat into the mic of a phone and use it to call the phone being tested. The process continues until the test phone dies. In this case the screen is allowed to go to sleep, as it normally would be if you were talking on the phone:

Apple promised up to 7 hours of 3G talk time with the iPhone 4. We measured 7.47 hours. That's an increase of 54.9% over the iPhone 3GS. While in a phone call the majority of the A4 SoC is powered down, so the efficiency improvements here have to do with how much less power the A4 consumes while off and the new Skyworks 3G modem (the iPhone 3GS used an Infineon modem).

In our iOS 4 review we looked at the impact multitasking had on the iPhone 3GS' web browsing battery life. I ran our 3G web browsing test while playing music through Pandora in the background. I repeated the test with the iPhone 4 for today's article:

We actually see our largest battery life improvement in this test. With a 57.7% increase in battery life over the 3GS, the iPhone 4 is not only more efficient at idle workloads but also when the SoC is constantly busy. The A4 SoC is rumored to be built on a 45nm process compared to the 65nm SoC used in the 3GS. With a moderate increase in clock speed we should be seeing a lot of the power savings that a full node shrink brings to the table.

The battery life offered by the iPhone 4 is spectacular. My iPhone 3GS could hardly get through a full day of work while traveling, I'd always need to hunt for an outlet before heading into my dinner meeting. I'm about to take my first trip with the iPhone 4 but I get the feeling that I might finally be able to make it through dinner.

Early reports of 20 and 30 hours of battery life are simply exaggerated. They're only possible if you let the phone idle in your pocket for the majority of that time. In other words, if you don't use the phone it lasts for a long time. While that's a testament to the platform's incredible idle power, the real world usage is good enough to stand on its own. It's better than any iPhone or Android phone I've tested thus far.

Apple's Bumper Case

The iPhone 4 is Apple's second product to launch with first party accessories. Like the iPad before it, the iPhone 4 is available with an Apple fabricated and sold case, dubbed the "bumper." At launch, this was the only case available. Most Apple stores (and online) only had black in inventory, though I'm told a few also had pink in stock. As of right now, the Apple store shows other colors shipping on July 16th - white, pink, orange, green, and blue eventually. I just went with black. 

 

Buttons are metal and brought through the case, except the vibrate/silent switch.

 

Inside the case, there's a felt-like mask which rests atop the buttons and aligns itself up. It works surprisingly well, and the buttons feel like those on the device itself. There's no gap inbetween - hey, the buttons just work.

 

Putting it on is a little bit of a struggle. I've found the best way is to insert the top, standby button side first (so you don't accidentally turn it on and off while inserting it), and then shove the bottom part in. After you've got it in, there's just a bit of massaging to get the rubber seats settled, and then you're good to go. It's very snug, there's no flexing.

 

 

At the top, there's a break in the hard plastic for the headset jack. Instead, this is entirely rubber. No doubt this is to allow jacks that come out and have a 90 degree bend to be inserted, something we're all too familiar with being a problem on previous iDevices. There's a tiny hole next to it you want to be absolutely certain isn't blocked for the noise canceling microphone.

 

 

The nice thing about the rubberized strips is that you get a grippy surface to hold the phone with, and you raise the front and back glass panes about a millimeter off the surface you're resting it on when you lay it down flat.

 

The idea of having a glass back is fine and dandy until you lay the iPhone on another glass surface directly. The result (if both surfaces are very clean) is that the two will come in optical contact and become very hard to pry apart. It's the same sort of reason coasters aren't glass - the two seal together, and become difficult to separate. Not to mention any particulates underneath will make those hairline scratches (sleeks) I talked about before. Even the strongest of glasses will get micro scratches when rubbed in optical contact with other glasses. It's obvious that raising both the front and back were design objectives here.

 

Look ma, I can rest the phone face down and not scratch it up!

 

Apple's bumper case is unique that it doesn't cover the back of the phone. At all. If you prefer a case that covers every square inch of the device and offers total protection, don't get the bumper and instead wait for something else. If you want to show off as much of the iPhone 4's design as possible, Apple has done it with the bumper. If you already purchased an iPhone 4, well, you know that outside of a plastic bag or a rubber band, the bumper is really the only option. 

 

Like we talked about before, the other hugely notable benefit is that the bumper insulates the stainless steel band from being detuned by your meatbag extremities (read: hands). Instead of a 24 dB drop, you'll incur a 10 dB drop completely in line with every other cellular phone ever made. 

 

The only problem with the bumper is the dock connector port. It's about a millimeter too small. 

 

That doesn't fit inside. Frustrating? Yes.

The unfortunate result is that all of my third party accessories don't fit. My Alpine iDA-X001 head unit in the car, three different vehicle accessory port car chargers with dock connectors, all my third party cables, and not to mention docks all don't fit.

If you've rolled with other cases, this is an all too familiar story that seems to repeat every single release, but it's frustrating in this case because the connectors don't match up by seriously under a millimeter. You can jam the dock connector in, but it doesn't go far enough inside to snap into place, and comes right back out. 

The result is that I drive around with the bumper case halfway off, like this:

It's not the end of the world, but I would have to file down that edge a half millimeter on both sides to make my accessories work again. Considering just how much the bumper improves signal and protects against accidental scratching, it's arguably a must have accessory. One that should maybe be given away with every iPhone 4 purchase. But you already know our thoughts on that. 

No doubt within a month or two the usual assortment of iPhone cases will pop up again, tweaked to meet the iPhone 4's slightly thinner dimensions. For now, however, the bumper is essential.

Living with the 4

Thus far the majority of this article has focused on the technical details of the new iPhone both in its advancements and shortcomings, but what is it like to live with?

As I implied in the introduction, the 4 makes the previous iPhone feel very old. But in day to day use it doesn’t feel much different. It uses the same iOS that iPhone 3GS users get access to so with the exception of FaceTime and camera improvements, the iPhone 4 is functionally the same as its predecessor. That’s an important thing to keep in mind because to a 3GS owner it makes the iPhone not feel any different.

We've already covered iOS 4 extensively so there's no point in rehashing that here. The only new experience we have to add is our time with tethering.

The option to enable tethering is unfortunately buried in iOS 4's settings. You need to go into general settings, then network and then Internet tethering. You also need to either have Bluetooth enabled or be connected via USB to the computer you wish to tether.

You of course need to have tethering enabled on your plan, which you can't do from the settings menu. You either have to call AT&T or add the plan via attwireless.com. From there just connect to the 4 and you can use it as a modem. The phone can even be locked while tethered. Once enabled it ranges from a lifesaver to a total waste of money depending on AT&Ts coverage.

Getting used to the size of the new iPhone isn't hard at all. The device is a lot thinner which makes two handed portrait operation more strenuous especially if you're typing a lot on it. The 4 is going to be one great gaming device because holding it in landscape mode feels awesome. Overall I'd say the size feels right.

The glass everywhere is nice. It does make it impossible to quickly tell which side is front when you're fumbling around your pocket looking for it though. And I still feel like I'm going to drop and break the thing in no time at all. While not as slick as the 3GS' plastic, the stainless steel and glass aren't exactly the most grippy surfaces.

The higher res screen makes even sending text messages clearer. Everything is just so much more crisp.

Browsing the web is more pleasant as well, but tiny text is still tiny text regardless of how sharp the display is. You'll still need to zoom in on regular sites.

You do get used to the screen though. It's only when going back to the 3GS that the 4 feels new again. It's sort of like an SSD, you miss the retina display when it's gone.

The seat of the pants feel has definitely improved thanks to the A4. Although, I'd say its subtle. Web pages load quicker, animations are smoother. It's not perfect however. There are still some choppy animations and pauses from the scheduler giving too much CPU time to a hungry process. In general though, the phone feels like it was designed to run iOS 4 while the 3GS can get a bit overwhelmed at times.

The OS improvements, although not specific to the 4, are still lacking in two key areas. Apple still hasn't implemented a good notification system ala Android or webOS. I expected that update with iOS 4 but I guess we'll get it next year instead.

Secondly the multitasking interface is nice but still not quite perfect. I am still holding out for either an improved webOS or for Apple to do it's own take on the card system.

How does the 4 compare to Android? The same differences are still there. Currently there is no Android phone quite as sleek as the iPhone 4, although I expect that could change in less than a year.

The latest Android phones have had better cameras and higher resolution screens, so the iPhone 4 doesn't have as huge of an advantage there.

Apple still has the UI polish down pat. Scrolling is far smoother and performance more consistent, although I believe Google isn't far away from fixing this. The Froyo update alone does wonders to performance. If Google could fix the scrolling issues I'd say a number of iPhone users would for sure be fine with switching.

There is this more open, more configurable, more capable feel you get with Android that you don't with the iPhone. That can be both a positive and a negative. Android phones feel more like computers while iPhones have more of that appliance feel. It still boils down to personal preference, the 4 won't change that.

Final Words

The iPhone 4 is a tremendous improvement over the previous phones from Apple. Battery life alone is enough to sell a brand new notebook, there's no reason the same shouldn't apply for a smartphone. Being able to deliver real world battery life use of between 5.5 and 10 hours on a single charge is quite impressive. And if you aren't using the phone nonstop? Expect even longer battery life.

On top of the battery life improvements Apple finally plays catch up and surpasses the technological advancements in its Android competitors. The 4's camera is as good as it gets today and the screen is a beauty. I don't believe this advantage will last for the lifetime of the iPhone 4. The phones that are in development today at least surpass the 4's specs in terms of raw CPU speed. Given that Apple's retina display is in high volume production already I'd expect other manufacturers to pick it up in due time.

And believe it or not, even if you upgrade to the iPhone 4 today in another 12 months it'll most likely be replaced by a dual core version that you'll want even more.

HTC EVO 4G (back) vs. iPhone 4 (front)

The main downside to the iPhone 4 is the obvious lapse in Apple's engineering judgment. The fact that Apple didn't have the foresight to coat the stainless steel antenna band with even a fraction of an ounce worth of non-conductive material either tells us that Apple doesn't care or that it simply doesn't test thoroughly enough. The latter is a message we've seen a few times before with OS X issues, the iPhone 4 simply reinforces it. At the bare minimum Apple should give away its bumper case with every iPhone 4 sold. The best scenario is for Apple to coat the antenna and replace all existing phones with a revised model.The ideal situation is very costly for Apple but it is the right thing to do. Plus it's not like Apple doesn't have the resources to take care of its customers.

As for the iPhone vs. Android debate, the 4 doesn't really change much. If you're not a fan of iOS 4 or Apple then the 4th generation iPhone isn't going to change your opinion. If you're an existing iPhone user you'll want to upgrade. It's worth it. The 4 simply makes the iPhone 3GS feel dated, which it is. It's a mild update to three year old phone vs. the significant redesign that is the iPhone 4. If you're married to Android, in the next 6 - 12 months we should see feature parity from the competition. And if you're a fan of Palm, let's just see what happens when the HP deal closes.

There's another category of users who are interested in the iPhone but simply put off by AT&T. While enabling HSUPA and the improved baseband make the iPhone 4 more attractive from a network standpoint, if you hate AT&T's coverage there's nothing Apple can do about it. I do get the feeling that the AT&T exclusivity will be over sooner rather than later. The iPhone and iOS are soon to be a mobile advertising platform, which means Apple needs as many users as possible. This is in direct contrast to the Mac strategy which purposefully didn't focus on volume to maintain high profit margins. Ultimately it means that AT&T either has to grow to be much larger than Verizon, or Apple has to embrace both carriers in order to fend off Android.