i have done that, locked just the same. use wireless networks allows me to get somewhat of a location in the house and such though, but with wireless networks on or off, I lock just the same.
-
After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.
All Things GPS
- Thread starter SamsungVibrant
- Start date
-
- Tags
- samsung epic 4g
i have done that, locked just the same. use wireless networks allows me to get somewhat of a location in the house and such though, but with wireless networks on or off, I lock just the same.
I am new to all of this, so please excuse me if this is posted in the wrong place. This is the My Tracks from a 4 hour walk around the NY Renaissance Faire I had just GPS on, and no Sprint signal at all. From looking at it the GPS is pretty good for my purposes.
boomerbubba
Android Enthusiast
One thing I and others have criticized about the GPS performance on other Samsung Galaxy S phones is that in driving tracks captured with My Tracks, the track tends to veer from the road during turns and curves. This apparently is a function of smoothing and interpolating algorithms built into the GPS or its low-level drivers.
Today I did a side-by-side driving test of my Epic versus my G1 with My Tracks set for maximum granularity, over a course that included some turns and high-speed S curves.
The bad news is that the Epic still exhibits some of that behavior. But the good news is that the G1 does too, and the Epic's performance on the curves is noticeably more accurate than the G1.
On straight roads, the two simultaneous tracks are quite close indeed.
The combined track is too complex to upload to Google Maps without being mangled by Google into two pages on the My Maps web site. So instead I have attached the track into a zipped KML file anyone can view in Google Earth desktop. (Just rename the .zip extension to .kmz and Google Earth will load it directly.)
So between my walking and driving tracks, comparing the actual accuracy of the Epic with that of my venerable G1, I am ready to give the Epic a thumbs-up.
Today I did a side-by-side driving test of my Epic versus my G1 with My Tracks set for maximum granularity, over a course that included some turns and high-speed S curves.
The bad news is that the Epic still exhibits some of that behavior. But the good news is that the G1 does too, and the Epic's performance on the curves is noticeably more accurate than the G1.
On straight roads, the two simultaneous tracks are quite close indeed.
The combined track is too complex to upload to Google Maps without being mangled by Google into two pages on the My Maps web site. So instead I have attached the track into a zipped KML file anyone can view in Google Earth desktop. (Just rename the .zip extension to .kmz and Google Earth will load it directly.)
So between my walking and driving tracks, comparing the actual accuracy of the Epic with that of my venerable G1, I am ready to give the Epic a thumbs-up.
<=U=L=T=R=A=>
Member
Thats in your city....I tested the gps along my tom tom using google maps navigator cause thats what im use to using when i had the evo...
Anyway I used it for 12 straight hours on a cloudy day with a 60% chance of storms , And at work traveled through 13 different countys over 300 miles and performed flawlessly
Again this is in AUSTIN TEXAS so those in austin it should perform flawless these are the places i went to
DRIPPING SPRINGS,LAKEWAY,ELGIN,BASTROP,AUSTIN,ROUNDROCK,PFLUGERVILLE,BUDA,KYLE,
LOCKHART,UHDALE,WEBBERVILLE,SANMARCOS....IM a FEDEX FREIGHT driver ....if your wondering why im traveling everyware
whats most awesome is that the google maps streams and never once at all did it lose connectionSamsungVibrant
Android Expert
Thats in your city....I tested the gps along my tom tom using google maps navigator cause thats what im use to using when i had the evo...
Anyway I used it for 12 straight hours on a cloudy day with a 60% chance of storms , And at work traveled through 13 different countys over 300 miles and performed flawlessly
Again this is in AUSTIN TEXAS so those in austin it should perform flawless these are the places i went to
DRIPPING SPRINGS,LAKEWAY,ELGIN,BASTROP,AUSTIN,ROUNDROCK,PFLUGERVILLE,BUDA,KYLE,
LOCKHART,UHDALE,WEBBERVILLE,SANMARCOS....IM a FEDEX FREIGHT driver ....if your wondering why im traveling everyware
good to hear.
boomerbubba
Android Enthusiast
Unfortunately, another user tried this and said the GPS Status workaround failed to solve the problem. He still had to reboot.
I can confirm this won't work. Although I very much appreciate your thinking that through and identifying the utility and method that should have worked!
Do you have a few? I have three epics on my desk along with a number of other smartphoens. I am finding the ephemris in cache is stale and will damage ability to get a fix after about eight hours but not gps status manual clearing of cache appears not to be clearing for me while a reset does.
I think there are a number of things epic is misreporting. The bogus 98' accuracy nubmer is not coming from google maps calculus or any default inherent to GM, but is beign reported to multiple prgrams at some samsung GPS stack level. Then again the battery status number reported to third party programs also seems to b made up out of thin air!
If you can think of any test sequence I can attempt with the tree epics at same location that might be helpful please do pm me.
gps status has adjustable refresh rate. I am using that and gps test
the best program I use, and which is running on a tp2 and treo pro as controls is wm based efficasoft gps utlities. numbers and fully reports elevation, azimuth and snr as well as loging raw NMAE. Also calculates three postion confidence/dilution of precision values (HDOP, vDOP, PDOP)
I was planning this weekend to show lag on our epics as well. they are lagging in navigation by about 50' on google maps side by side, but I need to set up some kind of four unit mount system to show controls have same window to sky.
not sure if this is what you mean, but I notice when using google nav that I'm past the street where I'm supposed to turn and its still saying 50 feet, not a big deal as I just know now that when the amount of feet come on, I'm turning at the next street, but still its nothing like my car nav.
Anand at Anandtech also confirmed the Epic 4G having issues in his review - Samsung Epic 4G Review: The Fastest Android Phone - AnandTech :: Your Source for Hardware Analysis and News
Well your car nav almost certainly has and teh software uses accelerometers. So there has been for years a compensation for lag in speed changes for de/acceleration.
Nav software on dedicated nav units (in car or handheld) and smartphones has "snap to street" to keep your indicator accurate laterally and on the street.
That being said, I am noticing differences between epic and other smartphones, notably more lag on the epic, running the same software accross the same modes. for example driving on google maps snaps you to streets and in walking mode does not. but with either setting on different devices I am seeing more pronounced position lag on epic than other smartphones (testing against Evo, TP2 and TreoP.)
Thanks for posting this link. I've been thinking that possibly a weak antenna has been a part of the problem and why it's taking so long to get a proper "fix". Interesting to see Anand come to a similar conclusion.
that and sloppy firmware compensation for the week reception
boomerbubba
Android Enthusiast
The only test units I have at my disposal are an Epic and a G1. I used to own a Vibrant.
I would very much like to just log the raw NMEA data. I even have an Android app called GPS Nmea Monitor, developed by the same guy who wrote GPS Test. Unfortunately, it does not seem to work on the Epic, and did not work on the Vibrant, although it reportedly works on some other Galaxy S variants. On the Vibrant, I saw a system dump that showed NMEA output was expressly disabled in a configuration file. I assume it is also disabled by design on the Epic, at least for now. The GpsSetup utility menus show a provision for logging and for NMEA, but they don't seem to work. I suspect they are just unfinished stubs at this point. (Notably, the GpsSetup utility is not officially documented.)
whtat do you think of my thoughs there on why gpsstatus might not or not always be clearing cache?
also I wonder if we should try default to warm start.
GPS issues still persist in Epic 4G - Page 6 - xda-developers
also I wonder if we should try default to warm start.
GPS issues still persist in Epic 4G - Page 6 - xda-developers
Robert Schoenberger
Lurker
Go figure - suddenly today my Galaxy Captivate during several trials clicks onto 6 - 9 satilites within 20 - 30 seconds and is giving me an accuracy bouncing between 16' and 34'. Wireless is turned off. Why the sudden change for the good? Previously it took minutes to get about 4 - 6 satilites and accuracy was 100'+. What app will let me enter Lat/Long and guide me to a point? I need to rough stake some lots in my subdivision. 16' accuracy is ok for the time being. I know the Lat/Long of the points to be set.
boomerbubba
Android Enthusiast
As for XTRA, here is what I PMed you. Really no reason not to have this conversation in public:
As for your suggestion about trying a Warm Start, I wondered about trying that myself. Candidly, I'm still trying to comprehend the difference between Warm Start and Hot Start (the default). I just understand that Cold Start always forces a fresh initialization that can take a long time. In my spare time I have been plowing through this reference, but have not assimilated all that I need to.
the way I learned factory, cold, warm and hot standalone (non-a) gps is:
factory: possible incorrect almanac from change of hemisphere,up to an hour
cold: no almanac data , no ephemeral data. normal amanac broadcast from birds is a repeating cycle of every 12 minutes, download is a couple of minutes, so a 5 to to 15 mins ttff (if you are not moving and possible interrupting download from the one bird you are pulling almanac from.)
warm. almanac present, partial emphermis, sats or base station have moved. 90 to 20 seconds.
hot: almanac and ephemeris current, sats and MS (the device) in nearly the same position as last known fix. instant
But, most of these differences in time to fix are not important with a-GPS, since instead of finding one bird, and downloading ephemeris and maybe almanac, the MS (our Epics) is going to pull that from sprints assistance servers through a data connection.
So I am wondering if warm or cold on the settings may force a data call to sprint to get the assistance data. The only caveat is that in some MS devices a rough fix (through tower Triangulation) needs to be sent to servers for the appropriate local set of assistance data.
Also a) A heavier than normal reliance on the cache over fresh pull from satellites would be one way to compensate for nominal satellite reception capability. and b) If you consider handset makers who are also serving countries where data is charged or limited, the cache might also be set to higher priority than a data call for assistance data. Combine a and b not thought out in the software stack and you get a kind of very persistent cache even when incorrect which would explain situations we are seeing with 5 to 8 birds in view and locked, and yet no fix by the MS because none are able to be "in use" due to erroneous ephemeris
hence perhaps some combination of forced warm (instead of hot) and some enabling of some non default xtra settings may help with that one problem
Ideally here is what the phoen would be doing when LBA requests for location:
1) aGPS current: Good sat conditions with current current almanac/ephemeris, calculate from at least 4 birds.
Proper programmed result Result:Actual accurate GPS Fix in a coupel of seconds
2) aGPS stale data: good sats network, connection available, but Ephemeris if stale,
initiate data call to netowrk, pull ephemeris from network
proper programmed REsult: actual accurate GPS, time to First fix (TTFF) about 20 seconds
3) Standalone GPS: good sats, no net connections avaialbe, ephemeris or almanac in cache stale, (warm or cold start autonomous GPS only, non aGPS)
proper programmed Result, actual accurate GPS fix, TTFF 60 seconds to 12 minutes)
4) Hybrid: Only Two to three sats available
proper progerammed result: Hybrid (non triagulated sat and some tower for caulus) 50' to 100'. One minute or so. Fix will lag a bit more than GPS and be slightly less accurate.
5) no birds visible. (Indoors).
proper programmed result
a) Forward link triangulation aflt
b) simple tower triangulation and other methods (AOA, TDOA, etc)
c) single tower location/cell sector (you locatioon will be a cicle of the tower you are using radius)
TTF, 10 seconds. Accuracy poor to fair 100' to 3000'
I think the samsung
a) has slightly worse than average sat reception.EG it is often using four sats when other deivces in same location are using 8 meaning more lag, and fix loss as you pass under tree or alongside building for example)
b) is sometimes NOT determining if cache data is stale and attempting to calculate fix whit incorrect cached ephemeris (hence seeing 4 to 10 birds but failing to calculate ms fix
c) has reporting bugs/false statements in api or stack falsely reporting 30 meters (98 feet) to all LBAsd) is less than perfectly optimized to determine which of the #1 through #5 conditions it is under in order to determine best method of fix for conditions.
SamsungVibrant
Android Expert
useful info thanks
i can say my gps is not stellar, but it locks pretty quick, maybe 30 seconds to a minute and usually is locked on 1 to 2 less than what is visible at the time using gps status. I'm in south florida, maybe that helps, not sure. but with wireless networks on or off, it locks day after day. accuracy is not super, but good enough for when I might need it to navigate me instead of my car nav.
It appears that most of this testing is occurring as people are standing still outdoors/indoors or whatever. Today was the first time I had a chance to see how it would do picking up the sats while on the move in the car. Turned it on as I left the parking lot and drove in town roads and then got on the interstate. It would not lock no matter what I did...even after turning on the wireless networks. Got home and was able to get a lock inside the house with no problem as usual. Anyone else have a problem with getting a lock while on the move???
i'm the opposite, can't get a lock inside house or buildings for crap, but outside, it is fine, I did my work to my home with mytracks and it did the route perfectly, even movedthe little line when I went from one side of the highway to the other.
guys, can you do us a favor and when you hit a persistent no lock state, say of five minutes, do a reset. and report what happens? Thanks
I don't get a no lock state, in the house i do or in buildings but it can't see the sats, how is it going to lock?
I've reset, i've turned wireless networks off, it always locks in like 30 secs or less, usually in 5 or less.
On the way home from work today I checked the sat situations with GPS Status and GPS test up here is Massachusetts.
#birds was between 8 and 11 (inside car while moving) and typical use was either one or two less than seen or the same as seen. Tried with USE WIRELESS NETWORKS both on and off and saw absolutely no difference.
Just walked outside my house, rebooted the phone and had USE WIRELESS NETWORKS off (double checked) Activated GPS test and had 8 birds with 7 in use within 15 seconds. Max SNR was 23 and as low as 16.
Sure seems like it works for basic navigation for me.
Also...I always used to notice that with my BB Tour and Bold units, Sprint Nav was forever just behind when it came time to make turns. My EPIC seems to be dead Nuts at this BUT I have had 3 occasions in the past week when for no reason it wants to re-route me for no reason as if it hiccuped and then quickly corrected itself.
For driving navigation it is as good as my Garmin.
#birds was between 8 and 11 (inside car while moving) and typical use was either one or two less than seen or the same as seen. Tried with USE WIRELESS NETWORKS both on and off and saw absolutely no difference.
Just walked outside my house, rebooted the phone and had USE WIRELESS NETWORKS off (double checked) Activated GPS test and had 8 birds with 7 in use within 15 seconds. Max SNR was 23 and as low as 16.
Sure seems like it works for basic navigation for me.
Also...I always used to notice that with my BB Tour and Bold units, Sprint Nav was forever just behind when it came time to make turns. My EPIC seems to be dead Nuts at this BUT I have had 3 occasions in the past week when for no reason it wants to re-route me for no reason as if it hiccuped and then quickly corrected itself.
For driving navigation it is as good as my Garmin.
RoboMonkey
Android Expert
K i just checked my wifes Epic standing outside I only got 3 sats to lock on with 8 in view. On my Evo in the house I use 5 sats with 7 in view. I'd say the GPS on the Epic is a Fail.
Does anyone know if this is a software issue? I will say this at least the Epic only uses the GPS when you use an App that needs it.
Does anyone know if this is a software issue? I will say this at least the Epic only uses the GPS when you use an App that needs it.