f13.net forums - iPhone 4 press conf

Welcome, Guest. Please login or register.
April 19, 2024, 10:21:31 PM

we're back, baby

I really love the design of most Apple products, but the excuses for this blunder get really retarded.

The patch that "fixes" the connectivity display. Means they manipulated the display to show more bars than were actually present. On purpose.

Assholes.

And that has absolutely fuckall to do with the topic.

Quote from: Engels on July 18, 2010, 08:42:35 AM

Oh Kitsune, its only the Phone part in iPhone that doesn't work right. The rest works great.

Well, yeah. Like I said, I'm looking forward to this year's ipod touch refresh. I skipped the third generation for its lack of any substantial update over the second gen, but am hopeful that the new stuff in the new iphone will make it into this year's ipod. The only flaws in the phone are the reception and the tremendously breakable glass construction. Sadly, those are two doozies.

I know nothing about the HTC Evo, but this video is brilliant.

http://www.youtube.com/watch?v=FL7yD-0pqZg&feature=related

I have some experience designing wireless systems, my company has its own high frequency and antenna design departments and we own our own antenna testing facility (which BTW should cost a lot less than 100 milion dollars). We design wireless systems of all purposes so I know a bit about the topic. (attn.: long)

Basically everyone involved from Apple, to the press, to competitors are talking extreme bullshit. It's been a PR disaster for Apple, yes, but everyone involved, including the people who first demonstrated the effects of the "death grip" are completely clueless. Alas Apple has done a bad job so far clearing things up.

1. "Bars" are a fucking bad unit of measurement because the mapping from bars to dBm (the real unit of measuring signal strength) is completely arbitrary. There is no standard, each company is free to choose a mapping that they like. It can even be even different with each device from the same manufacturer. I've seen everything from the "5 bars displayed although not even a magician could place calls" to "insanely good signal equals 3 bars". Although most devices tend to be "optimistic" or "very optimistic" (or as I prefer to call it "bullshitting the customer")

Two different devices from two different manufacturer might even receive with the exact same signal strength but show a different amount of bars.

The only site that compared actual dBm values was Anandtech and even they got some details wrong but at least they posted actual numbers that showed conclusively that signal attenuation on the iPhone 4 is significantly worse than on the 3GS and worse than on a nexus one.

Although if you look at the numbers you can see that the problem affects all handsets albeit not always in the same way.

2. The iPhone 4's noise threshold (the minimum signal strength at which reception is still possible) is -121 dBm compared to the -113 of the nexus or the 3 GS. Since dBm is a logarithmic scale this means that the iPhone 4 receives signals that are nearly an order of magnitude (power of ten) lower than the nexus or 3 GS.

-113 dBm = .015 pW (thats pico)
-121 dBm = .0018 pW

Even if the effect on the iPhone 4 is more pronounced it allows for a larger safety margin because of the better receiver. This is consistent with reviews that claim the iPhone 4 to perform better although signal is attenuated more. If you take that into account the iPhone 4 performs on par with the Nexus One

Since dBm is logarithmic you can just add and subtract dBm values so you could subtract 8 dBm (-121 to -113) from the measured values of iPhone 4 if you assume a -113 dBm baseline.

So if we take anandtech's numbers:

	Cupping Tightly	Holding Naturally	On an Open Palm	Holding Naturally Inside Case
iPhone 4	24.6	19.8	9.2	7.2
iPhone 3GS	14.3	1.9	0.2	3.2
HTC Nexus One	17.7	10.7	6.7	7.7

subtract 8 dBm from the iPhone 4's numbers to get a notion on how it would perform compared to the -113 dBm receivers (the 24.6 dBm attenuation would then be just 16.6 which is on par with the nexus).

So they actually designed the phone with a larger safety margin to counteract the effects.

3. If this sounds like being an Apple apologetic let's get to the things even Anandtech got wrong. Touching the antenna directly will always attenuate the signal, every electrical engineer worth his salt knows that. Apple must have clearly known that as do all of Apple's competitors.

Where Ananadtech (and nearly everybody else) is wrong is why a wireless signal get's attenuated. It's not the water content of your hand. On the contrary as far as microwaves between 800 MHz and 2100 MHz are concerned (the bands in which GSM and UMTS communication takes place) your body is nearly transparent. Signal attenuation because of water is only relevant as we approach 10 GHz.

A microwave oven uses 2.4 GHz not because waves at that frequency effectively heat water (which they don't) but because the band is available without obtaining a licence, at the time the microwave was first intruduced higher frequencies weren't possible, the microwave was primarily designed to dethaw not to cook and the process of dielectric heating doesn't rely on absorption (also you need a few hundred watts of power to effectively heat food)

A mobile phone can transmit at a maximum power of 1W (or +30 dBm) and usually sends at only .125 W (-5 dBm).

If microwaves were very well absorbed by water you wouldn't be able to heat the center of your food as the waves wouldn't even get that far.

Attenuation comes mostly from changes in the wave impedance level of the antenna. Not unlike speaker cables the connection between antenna and receiver requires a specific wave impedance level, if that changes reception suffers because the signal gets attenuated. To adapt antenna and transceiver a lot of coils and capacitors are placed in the signal path from antenna to transceiver to make sure the connection has the right impedance. This can be made to be adaptable to some extent. The human body is a very good capacitor so touching the antenna will change the capacitance of the whole system and mess with the adaptation. This is a well known RF issue.

Since antennas are susceptible to changes in capacitance you don't even have to touch the antenna to see the effect. You can see and measure an effect even if you are not directly touching the device but just get near it (You can even see changes in the reception depending on the surface the device resides on).

Anybody who has ever tried to tune an indoor antenna for radio or TV reception knows what I mean. You touch the antenna and reception is great until you let go of it (or vice versa). Thats because your body changes the adaptation of the antenna which can make the signal stronger (if the adaptation was bad to begin with) or weaker as you become part of the antenna system.

That's why antennas, transceivers and the signal path between them are always designed in conjunction with the device enclosure, because a good adaptation depends on all of these factors. Internal antennas tend to perform better than external antennas because you can control the distance from the antenna to the device enclosure (and therefore walls, surfaces etc.) and nobody can touch the antenna directly.

Still you will see some attenuation if somebody holds the device in his or her hands. That's why every handset manual I know of clearly states that you shouldn't hold the device in a way that covers the antenna by your hand. Touching the antenna itself has to be a lot worse. Short circuiting two antennas (that's what happens when you place your finger over the gap, you connect the cellular antenna and the WiFi one) even more so.

Yet it is something every mobile handset designer should be aware of and has to deal with. The mere existence of the "Bumper" is clear evidence of the fact that Apple was aware of these issues right from the start, it covers and insulates only the antenna portions of the device from getting in contact with your body.

TL;DR conclusion. Apple clearly knew that the issue existed, claiming to be unaware of the problem is bullshit (or they only employ clueless engineers), any other handset manufacturer that claims he doesn't have the issue is also talking bullshit. If you take the better transceiver of the iPhone 4 into account it doesn't perform worse than competing products which unfortunately is not enough to magically make the crappy AT&T service perform better. People living in areas with good coverage (which would be everyone not being a customer of AT&T) shouldn't even notice it.

Placing the antenna somewhere people are likely to touch them will make matters worse, the existence of the bumper clearly shows that Apple knew the problem existed.

Changing the mapping of bars to dBm values to make a device (or service) magically seem to perform better is standard practice of the entire industry but also a dick move. Since digital transmission is digital it is also largely a non issue. You can either place calls or you can't, it's irrelevant if you have the best signal strength possible or are below -105 dBm, performance should be exactly the same (although data transmission speeds can be lower).

Also there is no way to improve this situation by means of software only a complete redesign of the signal path including the antenna would help.

Dropped calls, bad coverage and crappy service aren't related to signal levels as much as you'd like to think (a topic I won't get into here) so as far as I'm concerned it's largely a tempest in a teacup kind of situation.

Apple got some well deserved backlash for being a dick (you are not holding it right, my ass) and for selling plastic that's worth $ .10 and should be free with your purchase for $30. A lot of tech journalists showed their total lack of technical knowledge either defending or condemning Apple and Apple's competitors try to bullshit the media and their customers into believing that the issue doesn't affect them. The rest is just one big media circle jerk.

Interesting, thanks. Needs more equations though...

BTW, how do you know what the noise threshold is?

See the follow up from Anandtech.

The new mapping for the bars is so that the limit is -121 dBm

They compare the mapping to the ones from the 3 GS and the Nexus One:
http://www.anandtech.com/show/3821/iphone-4-redux-analyzing-apples-ios-41-signal-fix

Also you can get the info from the data sheets if you know the transceiver model.

edit: provided link

So what do I have to do for my free plastic thingy?

Love the solution for the signal problem.

Quote from: Jeff Kelly on July 19, 2010, 03:19:00 AM

Dropped calls, bad coverage and crappy service aren't related to signal levels as much as you'd like to think (a topic I won't get into here) so as far as I'm concerned it's largely a tempest in a teacup kind of situation.

I was over at my friends house last night, who has the iPhone 4 and he let me DL Speed Test and mess around with the phones for like 30 minutes. First, the area where we live has notoriously low signal strength for AT&T, so that makes the problem really stand out, I never get more than 3 bars on the new 4.01 OS bars.

Using speed test I got results like this. (Going from memory, I dont have the numbers with me).

I did each test 5 or 6 times.

Phone flat on table: Download 500kb to 1000kb. Upload: 300kb to 800kb.

Phone held loosely in left hand (as I currently hold my iPhone 3g): Download: 0 to 100kb. Upload: 0 to 80kb.

Phone "death gripped": Download 0kb Upload: 0kb.

Next we called my phone and put it on speaker phone.

Phone on table: worked great both ways.

Phone in hand held up to ear: cut in and out. He could hear me, but often I got no sound from his phone. Ended up with dropped call after about 20 seconds. (5 times).

Phone death gripped (same results with one finger touching the "magic spot": dropped call in 5 seconds, very choppy.

I was honestly surprised how bad it was. The phone was basically unusable for phone or data if being held in the hand anyway except for the "3 finger grip". After doing these less than scientific tests, I am honestly amazed. It was REALLY bad.

Doing the call tests, never resulted in a dropped call on my end no matter how much I death gripped my iPhone 3g.

Two bars at best would be -98 dBm according to Anandtech. So if we take the numbers Anandtech measured we'd get a drop of 20 or 24 dBm (gripping it tightly vs holding it) which would push it well beyond the noise threshold either way.

I am not saying that it won't be an issue in places with notoriously bad coverage because it obviously is. If calls are dropped in places with average to good coverage however it's usually not a signal strength problem but rather related to congestion.

Don't get me wrong the antenna construction on the iPhone 4 is clearly not a very good design, especially if compared to the 3G or 3GS. Its not worse than the Nexus though. Two bars on the Nexus would be -98 dBm so holding it in exactly the same way as the iPhone 4 would lead to a drop of 11 to 18 dBm which would also push it over the limit (which is at -113 dBm). Crappy reception is crappy, the only phone that'd still be able to place calls would be the 3G or the 3GS (as far as phones measured by Anandtech go). All assuming that the numbers I quoted are correctly measured.

Download speed is another matter entirely because it's closely related to signal strength, as signal strength decreases the maximum speed of data transfers decreases as well. The coding schemes used to achieve HSPA speeds are more susceptible to interferences. Also speeds are related to congestion because calls take priority over data so if a cell is heavily congested data transfer speeds are reduced.

In your case the crappy reception alone is not the problem (it's near the noise threshold bot not under it) but the crappy reception combined with several interference effects (not that it's any consolation) most probably short term fading effects by multipath reception.

In congested cells the cell radius shrinks as more people place calls, this is caused by the channel access scheme employed (CDMA). Each active handset slightly raises the noise threshold which translates into shorter range of communication.

Quote from: Jeff Kelly on July 19, 2010, 03:05:39 PM

I am not saying that it won't be an issue in places with notoriously bad coverage because it obviously is. If calls are dropped in places with average to good coverage however it's usually not a signal strength problem but rather related to congestion.

I think Steve really wanted to bitch out AT&Ts part in this. I think this is clear when he casually mentioned having a Verizon cell on site.

Quote

If microwaves were very well absorbed by water you wouldn't be able to heat the center of your food as the waves wouldn't even get that far.

Not to go too far off topic, but I seem to remember calculating that at 2.4GHz, assuming a relatively high dielectric constant for the food, that most of the energy transfer occurs in the first several centimeters of the food, which is why high-water-content foods tend to get hottest on the outside while the center stays cold. The rest of heating occurs by convection (which is why most foods with high-water content indicate that the food should be left standing for several minutes to heat properly).

Perhaps I remember wrong, but if food was transparent to radio energy at this frequency why would this occur?

Looks like the iPhone 4 isn't going to be tied to a single provider here in Canada. Thats was a real dumbass move on Apple's part to have only one provider in the US. Its enabled the competition to gain strong market share.

Quote from: Ginaz on August 09, 2010, 01:07:53 PM

wasn't Rogers the only provider for a long time? I don't associate CDN telecom with any kind of free market.

Quote from: Ginaz on August 09, 2010, 01:07:53 PM

Verizon turned Apple down when Apple first went looking for a partner in the US. AT&T was the next logical choice given their size compared to Sprint and T-Mobile.

The only thing that doesn't make sense is the five year exclusive, apparently even carrying over to future models. Typically you give a carrier 90-180 days exclusive on a form factor.

Supporting the iPhone was a huge risk for whatever provider was going to do it first so it makes sense to me that they would want a long term exclusive contract with Apple. 5 years does seem too long, though, and the rumors are getting louder that a CDMA version for Verizon will be coming in January 2011.

You make the same mistake most US analysts (and the pundits like Gruber) make when analysing Apple's strategy and the question of why they didn't go with Verizon instead.

The iPhone allegedly cost 4 billion US$ in research and development (which is probably in the right ballpark considering that the project went on for nearly five years).

It never was a choice of where they'd get the better deal at least at first, the technology choice basically limited them to AT&T in the US. Verizon is CDMA2000/EV-DO while AT&T uses GSM/UMTS

If you discount the United States of America and its associated territories like Puerto Rico, all MVNOs (mobile virtual network operators, companies that lease existing networks for their service) and also all foreign networks that only cover small areas (e.g. a few cities) then there are only around 30 EV-DO capable networks worldwide. There are more CDMA2000 networks but you would not get the bandwidth necessary to really use those phones without it being EV-DO capable.

Most of those network operators cover countries that are not very attractive markets (Bolivia for example) or are in countries where the iPhone isn't even available today.

The only three big markets for CDMA2000/EV-DO are the USA, the east of China and Japan and even NTT DoCoMo (the Japanese operator of CDMA2000) doesn't cover the whole of Japan.

Compare that to more than 220 different 3G (UMTS) networks worldwide.

They'd basically have the choice to start with two seperate versions of iPhone of which one would only really work in the US (forget worldwide roaming, most of Europe and most of Asia don't use CDMA) to start with a only CDMA version that would only work on one provider in one country or to start with one GSM version that could probably be rolled out all over the world.

If you choose UMTS/GSM however AT&T is the only choice for the US. Verizon uses a different wireless standard and T-Mobile USA uses non-standard frequencies for it's 3G network.

Starting off with two seperate versions of iPhone or with just a CDMA version would have been a bigger risk. It's also not that simple to design a seperate CDMA version as many of the tech journalists would like you to believe. You'd have to design a completely new phone.

You'd need a different logic board (other baseband and transceiver chips with different pinouts and different adaptation networks), a different antenna (CDMA operates in 450 MHz and 900 MHz as compared to GSM/UMTS 900/1800/1900/2100 MHz) a different signal path to connect the antenna to the transceiver, the baseband processor would be totally different and you'd have to adapt your low-level and OS routines to the completely different call handling and connection management of CDMA even then you wouldn't be able to get calls while using your data connection (and vice versa). Verizon is currently changing that behavious but that would be non-standard.

Also CDMA2000/EV-DO is an end of the line technology, there were attempts to devise a successor standard but the only proponent (being the only one making chips for CDMA200) dropped it in favour of LTE (long term evolution the successor to GSM/UMTS) so beginning in 2010 most CDMA2000/EV-DO capable networks will start to upgrade to LTE (and automatically to 3G).

Verizon is already rolling out upgrades as we speak. So two or three years from now there will be no longer a need for a CDMA version of the iPhone because most CDMA operators will then be LTE capable (and therefore GSM/UMTS).

Sprint's EVO 4g already uses an GSM/UMTS/LTE capable chipset (that is unfortunately currently only running Sprint's own proprietary 4G protocol stack).

I don't really believe the rumors of an imminent CDMA iPhone, for one thing that particular rumor surfaces every other month, it would also only attract US customers that haven't already bought an iPhone because they value Verizon's network more and it would clearly be a bridging technology for maybe the next two years until Verizon completed its roll out of LTE.

So basically they'd need to attract enough US Verizon subscribers to break even on the additional costs of developing and manufacturing a CDMA iPhone.

Quote from: Jeff Kelly on August 10, 2010, 02:36:25 AM

There's no mistake. It's a fact that Apple went to Verizon first and Verizon rejected the idea. Yes GSM makes more sense for an international phone but there's no reason why you can't have CDMA and GSM versions or a single phone that does both, though Apple probably won't do that cause of space constraints.

http://www.usatoday.com/tech/news/2007-01-28-verizon-iphone_x.htm

I'm going to necro this one instead of starting new.

I have a 3GS and need to jailbreak it. Is there a current idiots guide out there for doing this? Searching proves confuzzling since I don't know which sites to trust (I'm not a super iPhone user but need to tether).

Quote from: Grimwell on August 30, 2010, 05:06:24 PM

No, really - I wasn't attempting humor. Anyone got some love for pointing to a useful link that teaches this?

I have no guide, but is there some reason the website that jailbreaks phones wouldn't work for you? Just don't trust it?

Exactly. Too many sites that want to show me how that look dubious, and then there are the ones who want to sell me it. Not sure w here I should trust since I'm not an iPhone warrior.

You misunderstand, I think. Simply going to the correct site jailbreaks your phone. www.jailbreakme.com

http://www.pmptoday.com/2010/08/28/iphone-4-jailbreak-jailbreakme-website-alternatives/

Ahhhhh yeah that helps a lot. Now to downgrade the phone so I can break it. Thanks :)

You own a 3GS so downgrading is not easily done.

Shortly after a new OS version ships Apple no longer allows restores to previous firmware versions. All firmware versions are signed with the unique key of your 3GS, to restore Apple needs to verify the signature or the firmware can not be installed. After a new version ships Apple won't let you restore to previous firmware versions the old signatures are simply marked as invalid.

There are tools that help you back up your signature files (they are called SHSH blobs) for your current firmware versions and requests by iTunes for verification of a certain blob can be redirected to them (they simply tell iTunes that your version is still valid for restore, your unique SHSH blobs are needed for that)

If you are already on 4.02 and haven't previously saved your SHSH files you are SOL, downgrading is not possible and you have to wait for an unlock for 4.02 or later firmware versions (since the dev team already announced that they won't do a jailbreak for 4.02).

If you are still on 4.01 now's the time to save your SHSH blobs. Download firmware umbrella and save them. Even if you are still on 4.02 it might be worthwhile to save them for future downgrades.

Firmware umbrella stores the blobs locally and can also upload them to Cydia. You can even check if there are already SHSH blobs on file (if you had previously jailbroken your phone they might be)

Jeff, I don't know what SHSH files are so maybe I misunderstand, but I can assure you that downgrading the firmware is pretty damn easy. I did it on my own iphone 4 and lost nothing. I did it on my wife's 3G and lost some pictures and game saves, and only then because we didn't have a proper backup on her iTunes (because the phone used to be mine, and the backup she needed was long gone). I'm a complete amateur with this stuff, but it was a snap.

You can only downgrade as long as Apple allows it. Usually you have a grace period of about 90 days after the release of a new OS version until restoring to old versions is no longer allowed.

If you click restore, iTunes verifies the OS version with Apple. It exchanges a certain amount of information that is uniquely connected to your individual iPhone with an Apple server and either allows or disallows the restore. After the grace period has ended and Apple has flagged your old SHSH as invalid a downgrade will fail and you can only restore to the latest firmware version.

You can only get around if your unique SHSH for that particular version is on file somewhere. Note that this is not the case on the 3G and the classic iPhone.

Yup. I'm hosed until 4.0.2 is jailbroken. Ah well I didn't like internet anyway.

I love Netflix for the iphone. I'm able to let the little man watch something and I watch something on the ipod.

So, I ordered the new iPod, because they were finally giving it all of the iPhone's features!

Well, the camera actually sucks and isn't the nice one from the iPhone. Ohhhhh, I see.

And the screen has the iPhone's resolution, just not its contrast, brightness, or viewing angle. ACK!

Oh, and it has half the RAM of the iPhone and can't run the Unreal game demoed at the event because it needs the 512 megs. swamp poop

And none of those details were actually revealed by Apple before people found it out for themselves. Mark of quality!

The Unreal thing runs on a 3G. It sure should run on the new thing. The rest is totally dickish though.