Database Users

very short gps test with My Tracks

http://maps.google.com/maps/ms?ie=U...372,0.018947&z=16&iwloc=00048f225d8da9a32da25
on a very very clear day with the Vibrant in my cycling jersey pocket on my back.
supl.google.com
sky hook on
MS Based
a few blips but its fairly close. but should be better. It's not unusable like some say.

It is unusable for some people, you happened to be able to use skyhook. If you were in the middle of nowhere that wouldnt help you.

Yea I used it also and it seems to be a bit off on the track
Sent from my SGH-T959 using XDA App

Never been unusable for me either. Actually fairly accurate ill have to do a my tracks later....

Lol you guys sound like some of the customers I deal with everyday. These are consumer grade devices. If you want something accurate go buy a Trimble GPS device and spend the thousands of dollars it takes to acquire one. I think even with the hiccups, your tracks are pretty decent. This is a cellphone, not a true GPS device.
BTW I work for Garmin so with my background of GPS and how it works this is probably why the inaccuracies in your track do not bother me. One thing you have to understand there are so many factors that can play into why you might of had these spikes....Heavy tree coverage, cloud over cast, power lines, tall buildings etc...Signal multi-path can also play into these types of spikes. Not bashing, but I think it does a pretty good job for what it is.

Steeltippin said:
Lol you guys sound like some of the customers I deal with everyday. These are consumer grade devices. If you want something accurate go buy a Trimble GPS device and spend the thousands of dollars it takes to acquire one. I think even with the hiccups, your tracks are pretty decent. This is a cellphone, not a true GPS device.
BTW I work for Garmin so with my background of GPS and how it works this is probably why the inaccuracies in your track do not bother me. One thing you have to understand there are so many factors that can play into why you might of had these spikes....Heavy tree coverage, cloud over cast, power lines, tall buildings etc...Signal multi-path can also play into these types of spikes. Not bashing, but I think it does a pretty good job for what it is.
Click to expand...
Click to collapse
Um I am sorry I disagree with you, with my first smart phone, the G1 took seconds to lock on to 4+ birds ... and my vibrant .. see's 10 and locks on to 1 .. and its not even the one with the strongest signal ... then there is an issue. I think that is what people are complaining about.
I agree .. if the true purpose of getting something is for GPS alone, then a Garmin etc .. is great.

aohmer said:
Um I am sorry I disagree with you, with my first smart phone, the G1 took seconds to lock on to 4+ birds ... and my vibrant .. see's 10 and locks on to 1 .. and its not even the one with the strongest signal ... then there is an issue. I think that is what people are complaining about.
I agree .. if the true purpose of getting something is for GPS alone, then a Garmin etc .. is great.
Click to expand...
Click to collapse
The G1 had almost the same exact problem with the GPS when it was first released in October of 08. In the beginning of December, it was updated OTA and it was fixed completely. The update for the Vibrant is coming this month, have faith.

scooterman said:
supl.google.com
sky hook on
MS Based
Click to expand...
Click to collapse
Skyhook is not GPS. Period.

scooterman said:
a few blips but its fairly close. ...ore it a 7/10, usable, but far from the best.
Click to expand...
Click to collapse

The release notes for the new Obsidian 2.2 ROM from TW says the GPS is "very accurate". I'm assuming it's got the real 2.2 drivers and SW.
It would be interesting to see a comparison of the JI6 and 2.2 GPS accuracy.
I've definitely had issues with significant inaccuracy using the Vibrant GPS for driving. It's usually fine, but sometimes it goes into rapid recalculations due to miscalculating my location on nearby streets. It's can be quite annoying if you're depending on it. I've used several standalone GPS units over the years and haven't seen this behavior before. In some ways the Vibrant GPS is better than my current standalone unit, but this inconsistent accuracy is a significant flaw.
I'm hoping 2.2 will include some improvements.

samnada said:
The release notes for the new Obsidian 2.2 ROM from TW says the GPS is "very accurate". I'm assuming it's got the real 2.2 drivers and SW.
It would be interesting to see a comparison of the JI6 and 2.2 GPS accuracy.
I've definitely had issues with significant inaccuracy using the Vibrant GPS for driving. It's usually fine, but sometimes it goes into rapid recalculations due to miscalculating my location on nearby streets. It's can be quite annoying if you're depending on it. I've used several standalone GPS units over the years and haven't seen this behavior before. In some ways the Vibrant GPS is better than my current standalone unit, but this inconsistent accuracy is a significant flaw.
I'm hoping 2.2 will include some improvements.
Click to expand...
Click to collapse
Although GPS test reports an accuracy of 6-10 yards roughly, once I start Maps, the accuracy varies on a much wider range. Driving in big city is not much fun when you could be anywhere within a 200 yard range.

No more GPS problems

I bought my Galaxy S few weeks ago and I could barely use the GPS. It took way too long to get it working and still it was inaccurate.
Few days ago I flashed JM2 firmware and installed app called TrackerBooster (can be found in the Android Market). Now the GPS works perfectly, takes only few seconds to get it connected (outdoors). I have used it now like 15 times and max. time I had to wait was 20 seconds.
I'm not sure does TrackerBooster app need to be installed, might work well without it, haven't tested.
I hope this helps those who have problems with GPS but this is just my experience, I would wait for other comments before flashing new a firmware.

It's called new firmware flashing syndrom for GPS. Basically, if you flash to any new firmware, your GPS will work for a few days.

As Foxbat said, a full flash/reset cycle improves results on the pathetic, schizophrenic and overall random SGS GPS implementation. But I am not going to be a smart ass and insinuate you don't know what you're talking about.
Instead I'll just ignore all the previous "the GPS issue is fixed" claims by many users who prematurely yelled "fixed" and just cowardly quote one of my favourite authors..."Extraordinary claims require extraordinary evidence." - Carl Sagan
Plus we shall also ignore the previous 308 users who have "cried Wolf"^-1. Please refer to your favorite fable author for proper reference.

Its occurred so many times, it actually has it's own syndrome!! FFS

Beowulf_pt said:
As Foxbat said, a full flash/reset cycle improves results on the pathetic, schizophrenic and overall random SGS GPS implementation. But I am not going to be a smart ass and insinuate you don't know what you're talking about.
Instead I'll just ignore all the previous "the GPS issue is fixed" claims by many users who prematurely yelled "fixed" and just cowardly quote one of my favourite authors..."Extraordinary claims require extraordinary evidence." - Carl Sagan
Plus we shall also ignore the previous 308 users who have "cried Wolf"^-1. Please refer to your favorite fable author for proper reference.
Click to expand...
Click to collapse
So you still trolling the forums about your GPS problem even after you sold your Samsung if you ever had one from the beginning. I have never had serious problem as many others and if only 308 of millions have had a problem it must certainly be a really good GPS. Stop trolling you only because look like an ignorant user who really do not want to make the phone better but want to promote NOKIA.

For sure, milsjg. You can add me too as ignorant, 309th user who's GPS sucks d*ck hard.

I wish my GPS sucked d**k... that would at least make it useful! ) 310th user here noticing GPS issues... that said while I was on JM7, I managed to record a very very close to perfect track using mytracks... for the walking section of my test it went perfectly even showing where I retraced my steps to go to poo bins! But when I got on to driving it went a bit mental and I apparently drove through the fire station, the wrong way round roundabouts and demolished a few streets worth of houses, but occationally it did show me going down the right side of the road!

You forget that before he even got it, he was complaining that he was reluctant to buy the phone because of reports about lag and GPS (and then, proceeded to buy it anyway).
The main problem is that people have different quality of acceptance for GPS. Some people expect a signal which compares with a fitness GPS (accurate within 2 or 3 meters generally), and some of us don't care if it drifts off by 20m.
Any technology which doesn't have a definitive state will get various opinions...

andrewluecke said:
You forget that before he even got it, he was complaining that he was reluctant to buy the phone because of reports about lag and GPS (and then, proceeded to buy it anyway).
The main problem is that people have different quality of acceptance for GPS. Some people expect a signal which compares with a fitness GPS (accurate within 2 or 3 meters generally), and some of us don't care if it drifts off by 20m.
Any technology which doesn't have a definitive state will get various opinions...
Click to expand...
Click to collapse
But then when it comes to navigation (arguably the principal purpose of having GPS in these devices) things go totally haywire! ) Personally I wonder whether GPS was installed purely for marketing purposes, to allow app developers (and of course google!) to snoop as to where their users are and how they are using their devices. For that there's no need for close tracking! )

not sure why you would fault beowulf for buying, if as you indicate he had some concerns - the damn phone is drop dead gorgous. If any product could stimulate optimism or optimistic hope, even in a skeptic, this phone would. And that's what makes it so frustrating.
I'm going to hate sending it back (i negotiated an extended buyer remorse return period) but if GPS isn't corrected by mid sept, this phone is history
samsung having released it w/o working gps is maybe forgiveable, but then even assuming they didn't know about the issue before release, it's been over 90 days since the asian release, and it's still not working.
Lay on top of that that samsung announced the GPS had been "validated" in the EPIC 4G, and it turns out it isn't. That did nothing but strengthen my doubts on samsung delivering a solution.

Tried this tracker booster app, makes no difference, gps still horrendous. If some people think that a 20m accuracy variation is acceptable for a gps lock they are a bloody fool.
Sent from my GT-I9000 using XDA App

Bynar010 said:
Tried this tracker booster app, makes no difference, gps still horrendous. If some people think that a 20m accuracy variation is acceptable for a gps lock they are a bloody fool.
Sent from my GT-I9000 using XDA App
Click to expand...
Click to collapse
Then it must be the firmware I flashed. There's a big difference with JM2 and the previous FW I had, if I remember right it was JF3.
The accuracy is good, for example Endomondo regonizes few meter changes. (like changing side of the road).

Sorry to offend you but Beowulf_pt has no SGS and still he is lurking around complaining. If you think about it ,there must be some reasons that I and several others have a perfectly normal (no correction) GPS on our SGS. Why is it working so different for different persons? Some logical alternatives would be error in a batch of phones (switch it), or operator programming has messed it up (get new operator), or just overload caused by program running in background, memory and task mangers etc ( educate yourself or wait for a proper kernel - see below) or in some cases just plain trolling of some non SGS owners (see my last reply).
The GPS chip is new and according to the web only recently the code for efficient low level drivers, for the linux kernel, has started to appear on internet (23 august). It will take some weeks before they are implemented so I would bet that they have fixed it by the end of September. You know, bad driver = slow, no calender etc. You will notice it specially when you use program that uses lots of resources (Maps and My tracker). It is much more efficient with Cardiotrainer for example or GPS test. (see my third conclusion above). I have noticed that (at least for me) it takes 5 - 10 sec to get a first lock in GPS test but half a minute at least in Maps or My Tracks.
Another tip is to run for example GPS test for one minute or so before My Track so the GPS gets some satellite data downloaded (calenders). The time also helps the GPS to calculate the exact position to use as a start point. You normally need a minute or so if the calenders are to old or you moved to get the right positions.
Iphone 4 has a good GPS (3 is crap according to some owners and it is not fixed. Do you hear any complaints?). The chip is one generation older than SGS and Apple do not need to wait for the kernel to be changed in there Iphone. They do it them self and by the way Iphone 4 is not really multitasking which means that GPS programs can easily be prioritized.
Well this is a short version. The net is filled with facts and if you start read a little you can start puzzle together what the cause is for your faulty GPS (nearly nobody believes its hardware any more).
By the way, I am not ignorant and I know rather much about GPS and how to evaluate tracking (I have written several scientific papers on tracking of wildlife where at least one was on the subject - evaluation of ARGOS tracking devices using GPS as reference). I have owned several mobiles and GPSes, programmed and built fish tracking devices, and did my thesis on wildlife tracking. I stating this not to try to end the GPS discussion but to make it a little bit more "scientific" and not so emotional like the comments on my last reply. There still nobody that knows how many that has a problem with the GPS. We only know that a few of millions have complained and a few of millions have said there satisfied with it.

Bynar010 said:
Tried this tracker booster app, makes no difference, gps still horrendous. If some people think that a 20m accuracy variation is acceptable for a gps lock they are a bloody fool.
Sent from my GT-I9000 using XDA App
Click to expand...
Click to collapse
On the contrar my friend, 5 - 15 meters is what you can expect and that means that the fix has 95 % certainty to be in that interval. 20 is a little bit above what you can expect but not much. In good dedicated trackers or a good phones you can get 5 meters (which I often get after 30 sec outside) and 2,5 for some fixes but not constant. A mobile is not a dedicated GPS and with the suspected driver problem 20 would not be bad. Try with GPS test, under open sky, and holding it at the bottom of the phone. Before you start you should turn off and on the GPS and close all task managers and started programs. You know that home means often that you leave the program running in the background but using back shuts it down. And wait for 30 - 60 seconds. If you get 10 be satisfied it,s a phone and not a dedicated GPS.

My touch hd would actually follow the f$$king road. Not too much to ask that a brand new smart phone would do the same is it? Its nothing short of pathetic. Even my old tytn2 was superb as a gps device for navigation, this thing cannot be trusted.
Sent from my GT-I9000 using XDA App

I for one am getting SICK of installing fw's hoping for a fix for GPS.
They put a good hardware-wise device out.. But its kinda shoddy the fw development isn't equally as good. I have flashed nearly ALL the fw's to no real avail..
I am started to get frustrated after all this patience..

FWIW
Running 2.2 with no lag, fix no root, no apps to help. Google Nav and Co-Pilot work flawlessly. Mytracks follows me walking down the pavement at the correct position and on the correct side of the road with very few deviations.
I really cannot see what all the fuss is ablout here.

Tehpriest said:
FWIW
Running 2.2 with no lag, fix no root, no apps to help. Google Nav and Co-Pilot work flawlessly. Mytracks follows me walking down the pavement at the correct position and on the correct side of the road with very few deviations.
I really cannot see what all the fuss is ablout here.
Click to expand...
Click to collapse
well try it in a car and get supprised...

praun said:
well try it in a car and get supprised...
Click to expand...
Click to collapse
That is where I use both Co-Pilot and Google Nav.....

Beowulf_pt said:
..y quote one of my favourite authors..."Extraordinary claims require extraordinary evidence." - Carl Sagan
.
Click to expand...
Click to collapse
Well I rather stick with Popper. Extraordinary evidence or support isn't that hard to get (look at the paranormal claims etc). What is more difficult and more fruitful to come by is some clever tests that eventually will or can refute a claim. If a theory will pass those tests it is actually more trustworthy then when it gets "extraordinary" evidence.
Again, no theory will stand till the end of time, how extra-ordinary our evidences may be.
Lol, back to the gps.

Geocaching is a sad experience

I tried geocaching over the weekend with the Epic. I love the app (c:geo I think?), but the phone's GPS is spotty at best. Obviously, when geocaching, you're probably in an area with no signal, so I had to rely on whatever info the phone already had downloaded and being able to fix.
About 9/10 times, the phone would see no satellites at all. Not even yellow non-fixed ones. Randomly, it would suddenly see a bunch of them and lock on within 10 seconds. I had no cell signal the entire time so no AGPS here (except for the previous night making sure it had downloaded satellite data). GPS Status was useful in figuring out what was going on.
The accuracy was horrible compared to an old hand-held standalone device. My phone was easily 20-30 feet less accurate. Note that the shown 98 feet accuracy level is BS - i'ts definitely more accurate than that. But whereas a standalone GPS might be within 12 feet, mine would be about 30 feet.
So overall it's mixed. The app for geocaching is really great, but the hardware to run it on is pretty lacking on the Epic. I'll retest sometime if we ever get a GPS fix.

Did you have the Cold Start workaround applied or is this with stock?

I switched to warm. Cold would have sucked without the network. Hot seemed the most problematic from what I've read, but maybe I should have left it on hot?
To be clear my gps works awesome as long as I have a network.
Sent from my SPH-D700 using XDA App

yes the only drawback to cold start, is if you are outside of network connection of some kind.
We are also not even absolutely confirmed that Epic has true standalone GPS. And if it does, it maybe pretty poor due to the low signal to noise, which may point to attenuation problems at a hardware level, or some other issue. (I am fooling around lately with the settings for SNR threshold rejection)
There is a lot to like about the Epic, and some GPS issues have been reduced by the cold start workaround, but it may not be the best choice of smartphones, notably compared to the EVo's excellent GPS, if you need or enjoy use of GPS.
On accuracy, you have to remember that the real world accuracy is negatively affected by not moving or moving slowly. It is counter-intuitive but this is the result of two factors affecting real and simulated accuracy respectively: Multipath is better rejected when moving as the algorithms to reject and compensate for multipath are facilitated by moment. Secondly, there are all those predictive forward and back axis compensators based on speed (and more recently argumented on smartphones by inertial data from the gsensors) , and road "snap" compensators for lateral movement that help in nav based programs.
Also keep in mind that elevation measurements are not as refined (with inbuilt compensation tables) as far as I know on Andorid as they are on WM applications. I believe there are some apps that will query online elevation tables -- present an data connection.
Thanks for your data on off connection results.

That's why I removed everything but the basics from my Hero and am using it solely as my geocaching GPSr!
Then I don't have to worry about ruining my Epic by getting it wet or dropping it or just getting it dirty!

jirafabo said:
That's why I removed everything but the basics from my Hero and am using it solely as my geocaching GPSr!
Click to expand...
Click to collapse
Yea I have a Palm treo pro (designed and made by HTC) that has outstanding GPS, standalone on it is excellent. also it runs both a WM navigon and WM tomtom perfectly with no data connection. Works from Topeka to Timbuktu

GPS

Do these things have a gps issue? I just came from an Evo, and gps reception on these sucks compared to Evo.
Sent from my SPH-D700 using XDA App

herzzreh said:
Do these things have a gps issue? I just came from an Evo, and gps reception on these sucks compared to Evo.
Sent from my SPH-D700 using XDA App
Click to expand...
Click to collapse
Bwahahahaha!! The answer to your question is that all the Galaxy S variants have GPS issues!

Some people claim they have trouble
On my epic google maps takes a long time to lock
But on google navigator and sprint navteq it locks on no more than 20 seconds it usually locks on in 5 to 10 seconds
What's weird ,is that to use google navigator. You have to use google maps to get to google navigator
Open google maps type place or address then hit drive directions then it switches on to google navigator then instantly gps locks on
If you just open google maps and turn the gps on and hit my location the gps takes a while to lock....but not when u start google nav.
I have no problem since when i use google maps, I never use my location.
I look up address and for driving directions via google navigator.

The problem with the GPS is that the cache doesn't update unless you reboot the phone. The longer the phone has been on the worse it gets. If you follow the instructions below it cause the GPS to update the cache over 3G whenever you fire up a GPS App. It only takes a few seconds to do this and then you get a fast lock after that. The only problem is if you are in an area with no signal and try to use the GPS. If that is going to be the case then follow the instructions again and change it to Warm Boot. Here are the instructions:
Enter GPS debug/settings application by typing *#1472365#
click "Setup" tab
click "position mode"
Click "starting mode
Enable "Cold Start"
Back out.

Ah... well, my problem wasn't the lock-on. That was normal. It's the accuracy and ability to hold signal. Again, comparing it to the Evo... this phone had that huge circle of inaccuracy while driving and Evo did not at the same time. Slightest obstruction and this would lose the signal while the other phone would not. Very frustrating.

herzzreh said:
Ah... well, my problem wasn't the lock-on. That was normal. It's the accuracy and ability to hold signal. Again, comparing it to the Evo... this phone had that huge circle of inaccuracy while driving and Evo did not at the same time. Slightest obstruction and this would lose the signal while the other phone would not. Very frustrating.
Click to expand...
Click to collapse
for lockon issues, which due to a bug in the cache system and will affect all users who don't cycle the phones cache, use the cold start as indicated above.
For accuracy issues it is important to remember to make sure to leave "use wireless networking" off.

aero1 said:
for lockon issues, which due to a bug in the cache system and will affect all users who don't cycle the phones cache, use the cold start as indicated above.
For accuracy issues it is important to remember to make sure to leave "use wireless networking" off.
Click to expand...
Click to collapse
I thought we were supposed to leave that on.
Sent from my SPH-D700 using Tapatalk

acegolfer said:
I thought we were supposed to leave that on.
Click to expand...
Click to collapse
That advice is disinformation from Sprint customer service. They are trying to cover up the fact that the GPS has problems, so they advise users to enable the non-GPS location by cell towers, which is not as accurate.

aero1 said:
For accuracy issues it is important to remember to make sure to leave "use wireless networking" off.
Click to expand...
Click to collapse
I tried that, still not too accurate compared to the other phone. What I noticed is that it IS accurate right after a hard reset, but then half an hour later it goes back to showing the circle.

herzzreh said:
I tried that, still not too accurate compared to the other phone. What I noticed is that it IS accurate right after a hard reset, but then half an hour later it goes back to showing the circle.
Click to expand...
Click to collapse
All the Galaxy S variants have a problem with SNR. Basically, the antenna has issues getting a strong GPS signal and some are worse then others; some can't even achieve a lock! Low signal = bad accuracy....
As for a fix, I'm not sure if Samsung can really fix it, if it's a design issue. On the Galaxy Forum, there was a guy that said it was possible to fix a low signal hardware flaw with software. Me, I'm highly doubtful it can be done where a Galaxy S GPS will ever be as good as the other Android phones currently on the market. I think they will eventually get it to the point where Nav will work most of the time, but not to where others are using there Android phones for Geo caching.

Didn't they fix it on fascinate?
sent from my epic 4g with no 4g

churro7 said:
Didn't they fix it on fascinate?
sent from my epic 4g with no 4g
Click to expand...
Click to collapse
Nope... Just do a search on the Fascinate forum and you'll find it's got GPS problems too.
Rogers in Canada said their version of the Galaxy S (Captivate) will have the GPS fixed, but we'll know when it hits the market in a few days. It might take a couple weeks before we start seeing the reports roll in.... If they truly have a Galaxy S phone with a working GPS, I'd like to see the mainboard. If it's the same as other versions of the Galaxy S, then maybe they were able to fix it with software. If it's a hardware design change .... Well, then the rest of the Galaxy S owners may be SOL.

GPS on Epic = Epic FAIL
Shame on Samsung for releasing the Epic with a half assed implementation of a GPS.
It is just one of those "what were they thinking" situations.

I mean didn t they release a patch
sent from my epic 4g with no 4g

crabjoe said:
Nope... Just do a search on the Fascinate forum and you'll find it's got GPS problems too.
Rogers in Canada said their version of the Galaxy S (Captivate) will have the GPS fixed, but we'll know when it hits the market in a few days. It might take a couple weeks before we start seeing the reports roll in.... If they truly have a Galaxy S phone with a working GPS, I'd like to see the mainboard. If it's the same as other versions of the Galaxy S, then maybe they were able to fix it with software. If it's a hardware design change .... Well, then the rest of the Galaxy S owners may be SOL.
Click to expand...
Click to collapse
The GSM versions of the Galaxy S, which includes the Captivate, Vibrant and International I9000, have one GPS chip -- the relatively new Broadcom BCM4751. The CDMA versions, the Epic and the Fascinate , use another -- the multipurpose Qualcomm QSC6085 radio that has been in use for three years. As far as the GPS is concerned, these are really two different devices: Different hardware. Different firmware. Different bugs. Different symptoms. (I have owned Vibrants and an Epic, and have tested the GPS rigorously on both. They behave very differently. The GPS bugs on the Epic are unique to the Epic as far as I know, although it is difficult to tell about the Fascinate from anecdotal reports and Verizon slapped a confusing user interface on the settings for location services and GPS.)
The only thing the two general designs have in common is general incompetence. Samsung managed to foul up two different GPS designs in two different ways.
But it is folly to try to make inferential comparisons between the two types.

churro7 said:
I mean didn t they release a patch
sent from my epic 4g with no 4g
Click to expand...
Click to collapse
Yes, but it didn't work.

boomerbubba said:
The GSM versions of the Galaxy S, which includes the Captivate, Vibrant and International I9000, have one GPS chip -- the relatively new Broadcom BCM4751. The CDMA versions, the Epic and the Fascinate , use another -- the multipurpose Qualcomm QSC6085 radio that has been in use for three years. As far as the GPS is concerned, these are really two different devices: Different hardware. Different firmware. Different bugs. Different symptoms. (I have owned Vibrants and an Epic, and have tested the GPS rigorously on both. They behave very differently. The GPS bugs on the Epic are unique to the Epic as far as I know, although it is difficult to tell about the Fascinate from anecdotal reports and Verizon slapped a confusing user interface on the settings for location services and GPS.)
The only thing the two general designs have in common is general incompetence. Samsung managed to foul up two different GPS designs in two different ways.
But it is folly to try to make inferential comparisons between the two types.
Click to expand...
Click to collapse
You're correct that the GPS hardware is different and personally, I don't think either GPS chips are the problem. I think the problem is hardware design, as in placement or size of the GPS antenna.
It's like in the old days before cable was popular and everyone had rabbit ears on tops of the TV. Regardless of how good or bad the TV tuner was, the type or placement of the antenna made a huge difference. I remember putting tin foil on the antennas to try and get a better signal. I also remember my dad telling me to stand in a funny position or place, with the antenna in my hand, so he could watch his football. LOL...
I really think Samsung designers goofed, just like Apple did, on the antenna. Bad antenna = bad signal.

crabjoe said:
You're correct that the GPS hardware is different and personally, I don't think either GPS chips are the problem. I think the problem is hardware design, as in placement or size of the GPS antenna.
It's like in the old days before cable was popular and everyone had rabbit ears on tops of the TV. Regardless of how good or bad the TV tuner was, the type or placement of the antenna made a huge difference. I remember putting tin foil on the antennas to try and get a better signal. I also remember my dad telling me to stand in a funny position or place, with the antenna in my hand, so he could watch his football. LOL...
I really think Samsung designers goofed, just like Apple did, on the antenna. Bad antenna = bad signal.
Click to expand...
Click to collapse
I must say that I have zero accuracy issues whatsoever. I *do* have stale ephemeris data problems like nobody's business - cold-start mode won't help either if I'm connected to WiFi or 4G (though 4G seems to have better luck.) The software-reported accuracy issues are a hard-coded datum, which in the Captivate has now been removed.
If I don't exclusively sit on 3G, then either Hot-start or Cold-start are unreliable. If I *am* on 3G, then Hot-start pinpoints me exactly, and almost instantly. With cold-start, it takes a bit to narrow in, though the *initial* lock is often faster. Other times, I'll see (in GPS Status) that it has found all 10 satellites, but not locked - the ephemeris bug all over again. Disappointing to say the least. BTW, this is not fixed on Captivate - it cannot get a lock while on WiFi. This has got to be a software problem. Neither my Epic nor Captivate have any issues actually sticking to the satellites. If this were software, then the GPS itself would be unreliable *during* operation; not just initial lock.

crabjoe said:
You're correct that the GPS hardware is different and personally, I don't think either GPS chips are the problem. I think the problem is hardware design, as in placement or size of the GPS antenna.
Click to expand...
Click to collapse
The SNR levels on the Epic do seem moderately lower than what other phones report. Theroretically this could be due to antenna design. The SNR level itself is a complex estimate by the GPS chip, not an objective reality being metered. So another possible explanation is that the SNR calculation is buggy. The estimated accuracy calculation is obviously buggy, always reported at 30.0 meters as if it is hard-coded. So there could also be a bug in the calculation of SNR, too. As for antenna design, I don't even assume that the Epic has the same antenna design as other Galaxy S phones, because its form factor is completely different.
In any case, signal sensitivity cannot explain the locking behavior, which has been isolated by user testing to be caused by bad handling of the cache of GPS almanac and ephemeris data. This is a bug.
Poor signal sensitivity could explain problems with actual accuracy. But like APOLAUF, I don't think there is a problem with actual actual accuracy on the Epic. The way to test that empirically is by plotting the recorded tracks in software such as My Tracks. If have done that with my Epic, for both driving and walking tracks, simultaneously with benchmark tracks captured by a known good device. I used my venerable G1, which has excellent GPS performance. The actual accuracy of the Epic is just as good.
The other GSM-based Galaxy S devices have had problems with actual accuracy, verified by many tests with My Tracks. But they are wholly different devices.
It is hard to tell from anecdotal reports in forums how accurate GPS performance is. Some reports of poor accuracy can be explained by poor locking. Some can be explained by pilot error. It takes controlled testing to get at the facts. Unfortunately, most commenters in forums, including this one, haven't a clue how to test the GPS rigorously.

IMO
First I have a co-worker who has been working with a developer and can get a GPS lock on his captivate super quick. To aide he is using the cell towers to get him a list of applicable satellites to his approximate location and gets a massive list of responding sats.
The next point is that I think that there is likely an issue with the Antenna . . but I do not think that is exclusively the issue. If this were strictly a HW issue then the issue would not be alleviated after a restart of the phone. There are multiple accounts that this is what is occurring.
Personally (and perhaps because this is my first true GPS phone) it is a minor inconvenience and doesnt bother me too much. . . . but is annoying and something that I would expect Sammy to go after. . . or at least acknowledge

gps precision

Hello,
Got a gs2 and still have my incredible s so I tested GPS precision: got 4 meters for the HTC and not stable and lower precision 18/20 meters on gs2.
Is something adjustable with new kernels?
Another two questions : about voice low quality ( someone said that it depends from noise reduction filet) and low volume during a call are these issues already addressed by some roms?
Low volume and quality with earphones is addressed too?
Thanks!

GPS is a disappointment. On my old pda asus a696 precision is much better. I wish gps was working better for me, I use it with sport tracking software. I changed ntp server, played with all kinds of gps apps, updated AGPS data, helped a bit but fix is still slow and precision is almost always more then 30m.
If anybody saw an improvement on some ROM or kernel, please let me know.
Attached my device info.

bzpwhx2 said:
Hello,
Got a gs2 and still have my incredible s so I tested GPS precision: got 4 meters for the HTC and not stable and lower precision 18/20 meters on gs2.
Is something adjustable with new kernels?
No as no reported flaw in GPS so nothing with GPS altered that i have seen.
jje
Click to expand...
Click to collapse

i am honestly very disappointed with the sgs2's gps precision. tested it side by side with the HTC Chacha...rebooted, cleared gps data and redownloaded it (Gps status app), then run test on both devices for abt 2-3 mins.
SGS2 7/7 satelites, error 40m!!!
HTC Chacha 10/10 satelites, error 6-10m!!!
frm both rooted 2.3.4 and stock firmware ...

I agree that the precision (error) is bad on the S2. However, in my experience, when I open maps and look at the location, it seems to be just about dead on. Meaning, even though there is often a blue circle, the center of the circle is extremely close to my actual position. So, I think the issue may be with how the phone calculates the "error" and not with any real inaccuracy. It feels like the reported error value is double what it should really be based on my position on google maps. This is just my experience, ymmv.

TheSopranos16 said:
I agree that the precision (error) is bad on the S2. However, in my experience, when I open maps and look at the location, it seems to be just about dead on. Meaning, even though there is often a blue circle, the center of the circle is extremely close to my actual position. So, I think the issue may be with how the phone calculates the "error" and not with any real inaccuracy. It feels like the reported error value is double what it should really be based on my position on google maps. This is just my experience, ymmv.
Click to expand...
Click to collapse
I agree with this. I was puzzled at first too when I looked at the accuracy reading, but then I realized that the actual positioning was much better than the reported 'error margin'.
The GPS itself is fine, but whatever algorithm they use to report accuracy is flawed..or should I say, could be better.

TheSopranos16 said:
However, in my experience, when I open maps and look at the location, it seems to be just about dead on.
Click to expand...
Click to collapse
Same here.

jimbiye said:
I agree with this. I was puzzled at first too when I looked at the accuracy reading, but then I realized that the actual positioning was much better than the reported 'error margin'.
The GPS itself is fine, but whatever algorithm they use to report accuracy is flawed..or should I say, could be better.
Click to expand...
Click to collapse
GPS accuracy is a statistical estimate of the maximum error and many assumptions must be made when calculating this, and I don't think different manufactureres does this in the same way.
So a HTC could very well report better accuracy than SGS2 but provide worse position (or vice versa).
Here is an interesting article about GPS accuracy, it mainly concerns static accuracy (reported accuracy in specifications), but the problems are basically the same for dynamic accuracy.
http://www.romdas.com/technical/gps/gps-acc.htm
When I've logged driving on a highway with SGS2 (with my tracks, i e no map matching or other filtering) the position is often within the correct lane but still SGS2 reports 10 meter accuracy (good GPS reception). But once in a while the position drifts a bit further away, mabe 5-10 meters. I would guess I get similar results with other modern smart phones with good GPS even if they report 4 meter accuracy. What phone is wrong in this case? The one reporting 4 meter accuracy that often is correct or the one reporting 10 meters that is even more often correct?
So please compare the phones by logging routes simultanously with "my tracks" instead. I've seen very few tests doing this, and SGS2 seems to do very well in the few tests I've seen

But not only the accuracy is an issue but initial fix is very slow compared to other phones. Even in hotstart mode when it had fix ten minutes ago it still most times takes over a minute to get a fix vs 7 to 10 secs on my Sgs 1
Sent from my GT-I9000 using Tapatalk

I had this problem and found that disabling wifi with data enabled led to much faster first fixes - I assume because the a-GPS data could be obtained more quickly direct from the carrier rather than over the wifi?
Also my UK phone shipped with a US NTP time server. I have since rooted and used GPS Aids to change this and refresh the other GPS data and now have no problems getting a first fix, usually in under 10 seconds.

Last week i traveled to Bosnia and used GPS on my SGS2 for the first time. I have only good words for it as with IGo it was very accurate (in a meter).But as i see with Google maps app GPS is wrong up to 20m.

I was also concerned about the gps first cause it seemed to show very inaccurate results and I have used gps with running a lot. However after using it to track my tracks I have to say that its 98% perfect. Also used it to navigate while driving and its also perfect. Samsung has just done something which makes it look like it would be more inaccurate but reality it is not.
Sent from my GT-I9100 using XDA Premium App

same here. I tested GPS performance a few times when i was walking (relatively static), the error was never under 10 meters, however it did not have a problem tracking my position.
In the other hand, I did have some long long GPS locks from time to time, once in a while the GPS just cannot seem to find my position, despite the fact that I am standing outdoors with no tall buildings around.
Overall experience, just acceptable.