All Things GPS

[Android_J]

Also, while your Vibrant can't precisely locate your position, mine will pinpoint my location to within a few feet. And I've been saying this from the beginning.

So are you saying that everybody with GPS issues has faulty hardware? If you're getting a lock to within a few feet using the same firmware as everybody else (since yours has been accurate from the beginning) then what else can it be? Should everybody with a faulty GPS across the entire Galaxy S line keep exchanging until they get a good one?

Can you provide proof showing your GPS locks to within a few feet? You're probably the only person I've seen on the forums claim that the GPS works perfectly and locks to a few feet. I've seen others claim their GPS works even though it only locks to 30 meters. That's not the definition of a working GPS, at least not in 2010 standards. I've yet to see anyone on the various forums claim a lock to within a few feet. Sorry if I'm skeptical.

How can you draw valid conclusions from a sample of ONE GPS review of ONE Epic 4G?

Because Samsung told Engadget:

"We have TESTED and VALIDATED both Network Assisted (indoor) and Autonomous (outdoor) GPS on the Epic 4G."

This means that *all* Epic 4G phones should be performing to specs since they've tested *and* validated. While I agree that we technically can't complain until the fix, circumstantial evidence and Samsung being coy with their wording ("optimize") does not inspire confidence. Guess we'll find out in a week when the Epic 4G gets released.

 

[wildblue]

I posted a pic myself in this thread (page 8) that shows GPS Test accuracy of 18.4 ft, or about 6 meters. Much better than the 30 meters you state. I would say I nearly always get accuracy better than 10 meters. I don't think we know enough to conclude whether the problem is HW or SW.

The problem I have with drawing conclusions from that one video is it's a very short duration. I've seen the blue circle gradually get smaller for a couple of minutes until the final accuracy level is achieved. The posted video is just one short sample.

As you say, we'll seen in a week or so.

 

[robo21]

So are you saying that everybody with GPS issues has faulty hardware? If you're getting a lock to within a few feet using the same firmware as everybody else (since yours has been accurate from the beginning) then what else can it be? Should everybody with a faulty GPS across the entire Galaxy S line keep exchanging until they get a good one?

Can you provide proof showing your GPS locks to within a few feet? You're probably the only person I've seen on the forums claim that the GPS works perfectly and locks to a few feet. I've seen others claim their GPS works even though it only locks to 30 meters. That's not the definition of a working GPS, at least not in 2010 standards. I've yet to see anyone on the various forums claim a lock to within a few feet. Sorry if I'm skeptical.

Because Samsung told Engadget:

"We have TESTED and VALIDATED both Network Assisted (indoor) and Autonomous (outdoor) GPS on the Epic 4G."

This means that *all* Epic 4G phones should be performing to specs since they've tested *and* validated. While I agree that we technically can't complain until the fix, circumstantial evidence and Samsung being coy with their wording ("optimize") does not inspire confidence. Guess we'll find out in a week when the Epic 4G gets released.

First off I'm not the only one who has stated that my GPS works. I have no more clue than you do as to what the cause of the GPS problems are and why there is so much performance variation from unit to unit.

If you look at the various forum polls you will find that there are numerous people who have working GPS on their Vibrants.

I don't think that "optimized" necessarily means "fixed." I think we will need to wait not just for the Epic 4G to be released but for the upcoming fix which is said to be coming in September to really know what's going on. I'm more concerned about what the Vibrant will do after Samsung releases the fix and we are on Froyo.

As far as your skepticism, I'm sorry, but I won't be taking the time to do any more pics of my GPS - the last time I did that it was a complete waste of time and didn't settle anything. If I show you a GPS indicator on a Google Map and the indicator says I am at point A you would still have to take my word that I am at point A wouldn't you? So I'm going to hear that it proves nothing after shooting pics, uploading same and posting on this thread.

I guess you will have to trust, I have no reason to lie or mislead you. We're all in the same boat. I never said my GPS was perfect. It is sometimes slow to lock up. I'm waiting with everyone else to have the perfect mobile device with perfect GPS.

 

[krichek]

So are you saying that everybody with GPS issues has faulty hardware? If you're getting a lock to within a few feet using the same firmware as everybody else (since yours has been accurate from the beginning) then what else can it be? Should everybody with a faulty GPS across the entire Galaxy S line keep exchanging until they get a good one?

Can you provide proof showing your GPS locks to within a few feet? You're probably the only person I've seen on the forums claim that the GPS works perfectly and locks to a few feet. I've seen others claim their GPS works even though it only locks to 30 meters. That's not the definition of a working GPS, at least not in 2010 standards. I've yet to see anyone on the various forums claim a lock to within a few feet. Sorry if I'm skeptical.

Because Samsung told Engadget:

"We have TESTED and VALIDATED both Network Assisted (indoor) and Autonomous (outdoor) GPS on the Epic 4G."

This means that *all* Epic 4G phones should be performing to specs since they've tested *and* validated. While I agree that we technically can't complain until the fix, circumstantial evidence and Samsung being coy with their wording ("optimize") does not inspire confidence. Guess we'll find out in a week when the Epic 4G gets released.

I'm pretty sure what he is saying is that you don't know it's a hardware problem yet you keep asserting just that. If you want proof of someone showing you theirs works to within a few feet, scroll back. I posted screen caps of mine and my wife's Vibrant. Both had more than 2 sats locked and both were accurate to under 15 feet.

That said, here is my take. My gps and my wife's gps both work very well 98% of the time. My wife's in fact seems even better than mine as I've seen her have 9 sats locked with an average of 7. Whereas I generally only average 4 with an occasional 5th locked. Both of us have experienced times where we have 0 locked and 0 in view tho which a reboot quickly corrects or if we wait awhile, fixes itself.

That seems more like a software issues then hardware one because digital devices tend to be "all or nothing". Very rarely will you have a digital component fail in a way where it actually works correctly for part of the time.

 

[robo21]

I'm pretty sure what he is saying is that you don't know it's a hardware problem yet you keep asserting just that. If you want proof of someone showing you theirs works to within a few feet, scroll back. I posted screen caps of mine and my wife's Vibrant. Both had more than 2 sats locked and both were accurate to under 15 feet.

That said, here is my take. My gps and my wife's gps both work very well 98% of the time. My wife's in fact seems even better than mine as I've seen her have 9 sats locked with an average of 7. Whereas I generally only average 4 with an occasional 5th locked. Both of us have experienced times where we have 0 locked and 0 in view tho which a reboot quickly corrects or if we wait awhile, fixes itself.

That seems more like a software issues then hardware one because digital devices tend to be "all or nothing". Very rarely will you have a digital component fail in a way where it actually works correctly for part of the time.

 

[robo21]

I posted a pic myself in this thread (page 8) that shows GPS Test accuracy of 18.4 ft, or about 6 meters. Much better than the 30 meters you state. I would say I nearly always get accuracy better than 10 meters. I don't think we know enough to conclude whether the problem is HW or SW.

The problem I have with drawing conclusions from that one video is it's a very short duration. I've seen the blue circle gradually get smaller for a couple of minutes until the final accuracy level is achieved. The posted video is just one short sample.

As you say, we'll seen in a week or so.

I guess I should clarify what I mean by a "few feet" since you are using GPS Test to measure the accuracy. I am referring to my position as indicated on Google Maps. The indicator will quite often nail my position to the room in my house and my location in that room which I find amazing. My accuracy numbers when using GPS Test are similar to yours.

I agree with you 100% regarding the short duration of the video review. Furthermore, I beleive we really should wait and see what the upcoming "fix" does rather than assume that the Epic 4G release will somehow be instructive as to what is happening with the Vibrant.

FYI wildblue, when referring to another specific post on the forum it would be better to use the actual post number in the upper right hand corner of that post rather than the page number the post is located on. There are display options on the forum which could change the page number your earlier post would be on for other folks depending on their individual settings.

So for example, you could refer to post #358 or use a link http://androidforums.com/1347373-post358.html By clicking on the #358 next to your original post on page 8 (for most of us), a new window will open with a link to that exact post.

 

[Android_J]

The problem I have with drawing conclusions from that one video is it's a very short duration. I've seen the blue circle gradually get smaller for a couple of minutes until the final accuracy level is achieved. The posted video is just one short sample.

While you have a valid point, you can see in the video that the blue circle starts out small and gets larger, back to small, back to larger. Even the reviewer says that it's not totally accurate. When I had my iPhone, I can turn on GPS and it's locked rock solid almost instantly. There's a huge gap between the performance of the GPS on the Galaxy S phones vs the iPhone. The ironic part is that the GS phones use a *newer* version of the Broadcom chip used in the iPhone which is supposed to be even more accurate.

First off I'm not the only one who has stated that my GPS works. I have no more clue than you do as to what the cause of the GPS problems are and why there is so much performance variation from unit to unit.

You claim accuracy to within a few feet. The *best* screen caps I've seen show 12+ feet. That's more than a few feet to me. 12+ feet might be good enough for navigating in a car but it's still not good enough for someone that wants to use a golf gps app.

Too many people are saying "good enough" when in reality it's not performing up to the standards of other highend smartphones with GPS.

Let me make this clear: 10+ feet is not accurate. Taking minutes to lock on is not normal. It might be good enough for you, but that doesn't mean it's good.

I'm pretty sure what he is saying is that you don't know it's a hardware problem yet you keep asserting just that.

Where in my post am I asserting that it's a hardware problem? If anything, it's you guys that claim your GPS works fine. I'm simply saying logic dictates that if a phone worked fine from the beginning, running the same firmware as everyone else, then people with faulty phones should keep exchanging them until they find a good one since good ones exist. Is there a flaw in this logic?

I guess I should clarify what I mean by a "few feet" since you are using GPS Test to measure the accuracy. I am referring to my position as indicated on Google Maps.

And what does it give you for accuracy when you click on the blue arrow?

 

[robo21]

You claim accuracy to within a few feet. The *best* screen caps I've seen show 12+ feet. That's more than a few feet to me. 12+ feet might be good enough for navigating in a car but it's still not good enough for someone that wants to use a golf gps app.

Too many people are saying "good enough" when in reality it's not performing up to the standards of other highend smartphones with GPS.

Let me make this clear: 10+ feet is not accurate. Taking minutes to lock on is not normal. It might be good enough for you, but that doesn't mean it's good.

I believe the screenshots you are referring to are GPS Test results or similar. On my Vibrant there is a problem with the GPS testing software in general and/or perhaps the accuracy calculation algorithm. I say this because the accuracy estimates range between 15 and 50 feet. Yet my real world results are right on.

Let me make this clear: This is the last time I'm going to write this. The accuracy I'm referring to is NOT the testing software's attempt to calculate accuracy. I'm referring to the blue arrowhead indicator on Google Maps in the satellite view. It generally nails my position to the room I'm in at home and more often than not my exact location within that room as an example. It has shown more consistency outdoors, like the parking space I'm parked in at the grocery store. To me, the way I use GPS, this kind of accuracy is more than just "good enough" and vastly more meaningful than testing software output.

Typically I will have my position locked in under 10 seconds occasionally it can take up to a minute.

 

[Android_J]

I believe the screenshots you are referring to are GPS Test results or similar. On my Vibrant there is a problem with the GPS testing software in general and/or perhaps the accuracy calculation algorithm. I say this because the accuracy estimates range between 15 and 50 feet. Yet my real world results are right on.

So now the GPS test software is bunk? Really? Do you think the software also has a problem counting the number of satellites that's locked on? Why does it seem to work well on other phones such as this one: http://androidforums.com/1327578-post303.html

 

[krichek]

You claim accuracy to within a few feet. The *best* screen caps I've seen show 12+ feet. That's more than a few feet to me. 12+ feet might be good enough for navigating in a car but it's still not good enough for someone that wants to use a golf gps app.

Too many people are saying "good enough" when in reality it's not performing up to the standards of other highend smartphones with GPS.

Let me make this clear: 10+ feet is not accurate. Taking minutes to lock on is not normal. It might be good enough for you, but that doesn't mean it's good.

So now the GPS test software is bunk? Really? Do you think the software also has a problem counting the number of satellites that's locked on? Why does it seem to work well on other phones such as this one: http://androidforums.com/1327578-post303.html

But that obviously isn't working either man. I mean look at the accuracy.. It's only 13.1 feet. Thats even crappier than the shitty 12 feet I got on my Vibrant.


Just took this pic of my Vibrants on the kitchen table, and before anyone asks. Yes, it's still inside the house.

 

[Android_J]

But that obviously isn't working either man. I mean look at the accuracy.. It's only 13.1 feet. Thats even crappier than the shitty 12 feet I got on my Vibrant.

Sure the HTC isn't great, but it's kicking the crap out of the Vibrant with its 0 locked sats.

Just took this pic of my Vibrants on the kitchen table, and before anyone asks. Yes, it's still inside the house.

That's not bad for indoors. How accurate are you outdoors? This is stock firmware? How quickly do you lock on from a cold boot?

 

[krichek]

Sure the HTC isn't great, but it's kicking the crap out of the Vibrant with its 0 locked sats.

That's not bad for indoors. How accurate are you outdoors? This is stock firmware? How quickly do you lock on from a cold boot?

Define cold boot? Just powered the phone on? I get a lock in less than 20 seconds that's usually around 100ft accuracy wise.. It then improves from there over the next 30-40 seconds..

This is pretty shitty quality sorry, I was trying to take it without flash since it was s dark outside the flash washed the screen out. In case you can't read it that's 11.6 feet, 7 out of 11 sats locked.

I saw my original Droid get to about 6 feet accuracy once or twice but it was almost always around the 10-15 foot mark so I'm satisfied with the Vibrant. It works fine for what I use gps for.

 

[robo21]

So now the GPS test software is bunk? Really? Do you think the software also has a problem counting the number of satellites that's locked on? Why does it seem to work well on other phones such as this one: http://androidforums.com/1327578-post303.html

First, your response was "So are you saying that everybody with GPS issues has faulty hardware?"

Now: "So now the GPS test software is bunk?" The trend in your responses is pretty clear, you apparently come here to inflame this discussion with extremist commentary not have a rational discussion. Your conclusions and confrontational questions are not based on logical thinking but rather seem to be intended to stir up an argument. That's unfortunate.

I will exercise my patience one last time. Did I say the software was "bunk?" No... What I said was: "On my Vibrant there is a problem with the GPS testing software in general and/or perhaps the accuracy calculation algorithm. I say this because the accuracy estimates range between 15 and 50 feet. Yet my real world results are right on."

How can anyone who isn't trying to start an argument say that that translates into "The GPS Test software is bunk?"

Then I went on to say that "real world results" are more important to me than testing software output.

Since my position is shown accurately (within a few feet at most - and often right on the mark) on Google Maps satellite view while simultaneously GPS Test is showing an accuracy of 50 feet then which software am I going to believe?

Which software is meaningful in my day to day use of the Vibrant? I don't dwell in testing software, I do however use Google Maps daily.

The answer is obvious to me, I will choose to believe the software that comports with reality. How can testing software "know" with any precision what the accuracy is? It is a mathematical computation based upon the number of locked satellites.

If the map shows my EXACT location within a few feet REPEATEDLY and GPS Test shows my GPS accuracy to between 15 and 50 feeet REPEATEDLY then clearly ON MY PHONE the Google Map with satellite view is the better "test" of accuracy.

Finally, it is irrelevant that the GPS software may be accurate on OTHER PHONES, remember, or maybe re-read my post, I said "ON MY VIBRANT" there is a problem with the accuracy reporting or algorithm.

Thank you for the link to the HTC phone info, if I ever get an HTC phone I will be sure to look that up.

You have a premise, or theory you want to prove. Which seems to be that the Vibrant's GPS is horrifically inaccurate, so much so that it is unusable. If this is what you are attempting to prove you have failed and continue to fail.

 

[Android_J]

The trend in your responses is pretty clear, you apparently come here to inflame this discussion with extremist commentary not have a rational discussion.

Are you for real? You're calling poster Samsung Vibrant an alarmist, etc. and you're telling me that I'm trying to inflame? Sounds like you're the only one flaming here. The thread is about GPS issues. Not the Samsung Fan Club.

You claim your phone works fine yet the only shred of evidence you provide is a statement of your real world results? Maybe that works in 5th grade but in the adult world people provide evidence to back up their claims.

Some of the other posters have provided evidence to back up their claim. You don't see me denying their proof do you? It didn't take 20 replies to get the proof out of them either. Why are you making things difficult?

Your posts simply obfuscate things and provide nothing.

 

[robo21]

Are you for real? You're calling poster Samsung Vibrant an alarmist, etc. and you're telling me that I'm trying to inflame? Sounds like you're the only one flaming here. The thread is about GPS issues. Not the Samsung Fan Club.

You claim your phone works fine yet the only shred of evidence you provide is a statement of your real world results? Maybe that works in 5th grade but in the adult world people provide evidence to back up their claims.

Some of the other posters have provided evidence to back up their claim. You don't see me denying their proof do you? It didn't take 20 replies to get the proof out of them either. Why are you making things difficult?

Your posts simply obfuscate things and provide nothing.

Here you go again: "Samsung Fan Club?" There's plenty of evidence already in this thread and elsewhere, but you obviously refuse to see it. You are in denial. What difference if I provide evidence or someone else does. Not everyone lies - but it does give some insight where you are coming from. Dude get a clue, I refuse to engage in a battle of wits with an unarmed man.

 

[smutek]

But that obviously isn't working either man. I mean look at the accuracy.. It's only 13.1 feet. Thats even crappier than the shitty 12 feet I got on my Vibrant.


Just took this pic of my Vibrants on the kitchen table, and before anyone asks. Yes, it's still inside the house.

Nice, I remember getting excited when mine briefly locked onto 3 sats one time.

 

[krichek]

I'm very hesitant to post this but...

I think I have figured out what the GPS problem is. I am going to refrain from posting what exactly I think I have discovered until I can get some more data to confirm my theory.

The reason for this post is that I need a couple other people that are willing to help by providing me with some data. If you are interested here is what I need from you.

You need to be STOCK GPS wise. If you have tried or are using any of the fixes I ask you not to volunteer at this point. You need to be willing to install the GPS Test app from Chartcross in the market (it's free) and you need to be able to give me some pretty frequent updates over the next several hours.

If you are interested please pm me and I will tell you what I need and provide you with my email address to send the data to.

Thanks!

 

[SamsungVibrant]

I'm very hesitant to post this but...
If you are interested please pm me and I will tell you what I need and provide you with my email address to send the data to.

Thanks!

Meet me at the corner payphone, 100,000 unmarked bills. J.K, Ok I'll do it, what kind of data are you collecting, and tell us your theory. We don't bite.

 

[krichek]

Meet me at the corner payphone, 100,000 unmarked bills. J.K, Ok I'll do it, what kind of data are you collecting, and tell us your theory. We don't bite.

Sorry, I should have clarified on the data. I don't need anything personal, just some info from the GPS Test app. At this point I'd like to keep my hunch private tho since it is only a guess. The data I have indicates it is most likely correct which is why I'd like more input.

I promise to share later tonight or tomorrow whether it pans out or not so don't worry.. You'll get to laugh if I'm wrong!



Edited to add.. I actually will need 1 piece of personal info. Your City/State. Makes it easier for me hehe.

 

[elgibby]

It's been quite an adventure with Vibrant GPS today. Decided to go to a late morning movie at a theater I've never been to. Knew generally where it was, so I started driving and waited for GPS to lock in. It never did. During the 20 minute drive I rebooted once, no help. Once I got near where I guessed the theater was, I pulled over and rebooted again. That began about 20 minutes of black screens, 5-10 second lags, locked screens that wouldn't unlock, a spontaneous reboot etc. Missed the movie. Got it to lock in for just a second, but lost it soon as I started driving. Never got it back. Right now, Navigation has me down in South St. Louis County, 20 minutes from home. At least that's dry land. Maps has me in middle of the Mississippi River.

 

[krichek]

Sorry for not being able to get back and update this thread before the weekend ended. Sadly I was only able to get one person to volunteer to provide me with some data so it took much longer to accomplish what I was hoping to using data from 3.5 Vibrants.

That said I have a quick question for the masses. Does anyone think it would be a step in the right direction if I am now able to pinpoint to the minute when my GPS is functional and when it is not?

By that I mean if I were to tell you that my GPS is not functional from 14:10 to 16:11 everyday +or- 30 seconds. This is true for my Vibrant, my wife's Vibrant as well as a friends Vibrant (he is about 125 miles from me). I have more info to share, just trying to consolidate it into something a bit more manageable and easier to read. I'm going to try and post it tonight tho.

 

[krichek]

Alright everyone, I am attempting to post the data I collected over the weekend again. I typed a long post out earlier but when I went to submit it there was a database error and of course the post was lost.

This was a pretty labor intensive effort so I want to give out the details of exactly how things were setup and the devices/software used.

I had access to 3 Vibrants, mine, my wife's (who was out of town) and my friends who relayed me info from his device (He lives in the same town as I do but about 10 miles away). Additionally I had an original Droid, Touch Pro and Garmin Etrex Vista for comparison purposes and utilized info from this website LIVE REAL TIME SATELLITE AND SPACE SHUTTLE TRACKING AND PREDICTIONS and software available from this site Trimble - Planning Software

My Vibrants were placed in different locations within my house, originally the other devices were located with my Vibrant but I later moved them outside when it became apparent that the Vibrant was considerably more sensitive then them.

Over the course of 3 days I collected the following information. Sats In View, Sats In Use, ID of Sats In Use and ID of Sats in View. What I found was incredibly consistent across all 3 Vibrants and at times down right strange.

Hopefully there is a GPS expert out there that can take this data and figure something out. Meanwhile, in no particular order here is what I Found:

1). The Vibrant is an incredibly sensitive GPS receiver. It was the only device that was able to show all of the satellites present in the sky at any given time as "In View".

2). According to the website and the software I should see sat 32, (I never did). And not see sat 25, (i did). It never locked onto 25 tho so this is more for completeness than anything else. I wonder if GPSTest is reporting sat 32 as 25 for some reason. I also never saw satellite 1.

3). The Vibrant NEVER locked onto the following satellites. 1,3,6,8,11,16,22,25,30.

4). There is a time component that I can't seem to work out. Sometimes there are satellites present that the Vibrant will lock onto but they are not in use. This is either because it is before or after the window in which they will lock on, this window has nothing to do with their position.

5). There are only 3 satelites that were able to get locks by themselves 7,26 and 29.

6). These results are incredibly consistent. I could predict within 1 minute which satelites would be "In View" and which would be "In Use". This was independant of which Vibrant I was looking at.

7). I have now mapped out when GPS will and won't work in my area, for me the absolute best time starts after midnight when I can progressively lock on to upto 7 satellites at it's peak. My dead time is from 1401 to 1713 when I have 0 satellites "In Use".

8). The chart looks kind of like a bio-rythym. Starting at 0 (1401) it will increase throught out the night till it hits it's peak of 7 at roughly 0200-0236. It then slowly swings the other way going down by 1 until at 1401 it's back to 0.


A few conclusions I've come to based on the very limited data that I've aquired. It's pretty apparent why some people have no issues with their GPS. If you are located geographically where the right satellites are in view you will have a much better experience then someone located in an area where the orbit is not ideal.

I believe this is a software issue, the results are just to consistent for this to be a hardware issue.. Does anyone know if it is possible that there is something in the driver preventing it from locking onto certain satellites? Like a preferred list or something? The other thought I had is based on an issue with time. It appears that the driver is applying the UTC offset (without regard for DST) twice. I think this could cause some serious problems with regards to what the software thinks should be in view versus what actually is.

I'm sure I am forgetting something that was in my original post, so if anyone has some questions or comments just let me know and I'll try to answer them.

 

[Big-Dog]

GPS satellites are not in geosynchronous orbit. They orbit the earth every 12 hours so they are constantly in motion. If the Vibrant is limited to what satellites it can lock on to, that may explain the fluctuation.

 

[robo21]

Wow, nicely done Krichek. You really put a lot of energy into this. Your conclusions sound reasonable and logical.

Thank you for the effort and posting your results!

 

[w_bovine]

GPS satellites are not in geosynchronous orbit. They orbit the earth every 12 hours so they are constantly in motion. If the Vibrant is limited to what satellites it can lock on to, that may explain the fluctuation.

Egad...if that's true it's a catastrophic blunder on the part of Samsung. Hard to fathom really. Nice work krichek.