• After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.

Pandora on the EVO

All evidence as well HTC and Sprint confirm this is an issue with Sprint's 3G network. It works fine on wifi. It works fine on 4G. And if I use roam control to put my Evo onto Verizon's 3G network, it works flawlessly. The moment you try to use Sprint's 3G network, it takes a crap.

There is really no sense in trying to screw around with it any further. Until sprint fixes their problem, we're basically screwed.
 
All evidence as well HTC and Sprint confirm this is an issue with Sprint's 3G network. It works fine on wifi. It works fine on 4G. And if I use roam control to put my Evo onto Verizon's 3G network, it works flawlessly. The moment you try to use Sprint's 3G network, it takes a crap.

There is really no sense in trying to screw around with it any further. Until sprint fixes their problem, we're basically screwed.

Just curious where you read this?
 
It was last month so I couldn't give you a link. I did a lot of searching on here and on XDA and that was what I found.
 
well..updated to 3.70 and running Pandora version 1.5.3
45 minute test...no song skips...and then it just started :( it skipped on 8 songs in a row.

noticed wifi and gps were on...turned those off. it stopped skipping for approximately 30 minutes, then skipped once....i turned it off.


have tested the fiancees htc aria on att twice for 45 minutes plus...no skips

i wish someone could find a fix :(
 
It was last month so I couldn't give you a link. I did a lot of searching on here and on XDA and that was what I found.

Now was this something that was quoted as Sprint having said? I am just curious because it seems like a search should bring up news like this pretty quickly seeing as how it should be official news. If it isn't official news, no offense, but the credibility is probably out the window.

I would still love an explanation of how a network is pushing information to Pandora to go ahead and skip the song. If it is having issues streaming the data, I undertand this, but the song SHOULD NOT SKIP. It should simply stop playing and resume once internet is back in swing. That is how it works on any other device I have seen, including computers.
 
Well, so far since the latest Pandora and Sprint updates (and rooting, running Fresh 3.5.0.1) I have not had a skip yet. I will keep reporting my results.
 
Well, so far since the latest Pandora and Sprint updates (and rooting, running Fresh 3.5.0.1) I have not had a skip yet. I will keep reporting my results.


Please list which version of pandora and length of test. As posted above....i have current sprint and pandora....skips occurred after 45 min

Interesting tid bit. Turn on pandora...play a song. Turn off ALL radios ie airplane mode.
 
All evidence as well HTC and Sprint confirm this is an issue with Sprint's 3G network. It works fine on wifi. It works fine on 4G. And if I use roam control to put my Evo onto Verizon's 3G network, it works flawlessly. The moment you try to use Sprint's 3G network, it takes a crap.

There is really no sense in trying to screw around with it any further. Until sprint fixes their problem, we're basically screwed.

well..updated to 3.70 and running Pandora version 1.5.3
45 minute test...no song skips...and then it just started :( it skipped on 8 songs in a row.

noticed wifi and gps were on...turned those off. it stopped skipping for approximately 30 minutes, then skipped once....i turned it off.


have tested the fiancees htc aria on att twice for 45 minutes plus...no skips

i wish someone could find a fix :(

If you had skipping while on wifi, then that sorta defeats the Sprint 3G statement above.
 
There is a difference between WiFi being on and WiFi being connected. That statement doesn't defeat anything without further details.
 
Per Mr. Ed's comment -

We were noodling around at the Shift forum and realized that 3G download speeds can tank when GPS is on - I verified this on my Evo, it dropped by half.

This maybe might make some sense - the GPS and CDMA radios are in the same chip on our phone.

Also - might check if other services are syncing, or tower location is on - that might be linked to some timed-based extra activity that could cause this to skip.

So far as I think I might know, GPS/location services have priority.
 
Kinda along the lines of what i was thinking. An interesting find (continuing on my last statement) Pandora's buffer on the phone is at least one full song. If you start Pandora, song playing and KILL ALL RADIOS....it will finish that entire song. Found this very interesting as my prior belief was that a combination of low buffer size and poor netemwork reception/translation by the Evo were culprit


In previous tests i experienced the skips while connected to wifi but i have not been home long enough to do a lengthy wifi test lately.

My mention of wifi above was under scenario of driving, no wifi connection but the wifi was on. My initial thought was that when in range of a wifi signal while driving it might be causing issues however once wifi was turned off it continued to skip.

I plan on repeating my test tomorrow...no wifi no gps, same route to see what precedence location may have on the issue
 
watching to see.. what you all find.. and can confirm..

I just tried.. it..
i am running Myn RLS5 with Netarchy-Toastmod..
using speedtest.net.

my speeds are very irratic lately.. sometimes fast.. sometimes slow..
3G....
with GPS off-- I got around 200-500kbs
with GPS on-- I get around 200-500kbs
**GPS On = radio switch is turned on.. not using any apps that call the radio to action.

so.. no conclusions.
 
^ same here...no difference in speed regardless of GPS state at house (note that speedtest does actually use the gps radio for a moment to determine location)

speeds here at the house are consistantly 1900 down ..upload very irratic anywhere from 150-500
 
I would love to find out.. why in the same bedroom at the same desk..
my speedtest.net on 3G... can be anywhere from 80kbs to 1.5mbs!

yes.. I know network load plays a huge determination to this...
but I have tried at all times of the day. Even at 2am! and the test can be fast o slow.

If I can find a problem; I would get a replacement.....

this bothers me to no end!!!!!!
 
"Looks like it's going to rain tomorrow. Probably can't paint the garage"

I WISH mine would skip....over that commercial. I sat in traffic for ninety minutes and heard that commercial three hundred and forty seven times.
 
ok so today's drive in...same exact route... skipped 8 songs after 45 minutes of driving/play time. I was in a completely different location on the route though. When I pick up the phone ...pandora says "loading" starts a song, then skips to another, then another, then another etc. went ahead and pulled a speedtest and dload speeds were less than 200 kbps upload speeds were 250-330 kbps

I think my theory on area/location interference is out the window.

I am back to it thinking it is related to some network issue, an issue with buffering, or both.
 
Mr. Ed - is it time rather than location dependent?

dan330 - try a different server. That's my gripe with Speedtest ratings. When I use that on my desktop, I don't care how St. Louis servers respond when I'm accessing San Francisco. Also the desktop / web version will show the server load (therefore, when it's pointless to try to use that server).

Remember, Speedtest doesn't show merely your 3G speed - it shows the speed to transact with a specific server on a specific path that includes 3G.

If you've checked servers and wifi is having no issue, then it's likely a case of the net bottlenecking over 3G in your neighborhood - either by being in a kinda fringe-y spot or having too much activity at the towers.

On my wifi, I can track just after 5:30 that everyone is home from work and checking things (response goes down), about 6:30 to 7:30 everything settles down (dinnertime), then a slight bump down while whomever is downloading Netflix or whatever.

In the ideal case, you want to know where the Pandora server feeding you is at, and Speedtest to the nearest location - that's an indication of your Pandora path.
 
so far it is more time than location Dependant. there is one spot that I lose signal on a bridge over the lake. I always drop a call or Pandora there but it picks right back up once out of the dead spot.


I am going to keep messing with it. I miss Pandora....I could even handle the occasional skip, but when it hits the 45 min mark and skips EVERY SINGLE SONG...it's crazy.


You make a great point about bottlenecking....I have always though there was just better quality from the cell tower at the house vs the one at work...

the area near work is much more populated than the rinky dink town I live in...so maybe it has more to do with that

what still confuses me is the airplane mode test...Pandora finishes the current song..every time..even though there is no internet connection.

So...in these cases that Pandora skips to the next , seconds into a song/halfway through it....when is the issue actually occurring (and what is the issue)
 
In the ideal case, you want to know where the Pandora server feeding you is at, and Speedtest to the nearest location - that's an indication of your Pandora path.

I've always assumed that SpeedTest looked at where you jumped off Sprint's network and hit the public intarwebz. So my phone goes 3G for less than a mile to the nearest tower, and in the cabinet or hut at the base of that tower, there is likely a Fiber Mux of some sort. Then, I travel through fiber (back in the day, it was T1) across many networks until I get to Wichita Kansas (I live in SC/NC) where Sprint's servers connect me to the world.

WHen I use 4G however, I usually come out in a different place. Though with either 4G or 3G, it is not always the same point. Sometimes I connect in Atlanta, and other times, I connect in NC. It's all over the place, so I am not sure what accounts for that.

That being said, it seems to me that using GPS in order to make SpeedTest pick a geographically closer server could throw off the results. The geographically closest server isn't always the the closest server with regards to networks.

If I am indeed leaving Sprint's network in Wichita Kansas, it's best to use a Kansas SpeedTest server rather than coming back across the country to a server in NC/SC isn't it?

Do I have an accurate take on things, or am I missing something?
 
That being said, it seems to me that using GPS in order to make SpeedTest pick a geographically closer server could throw off the results. The geographically closest server isn't always the the closest server with regards to networks.

If I am indeed leaving Sprint's network in Wichita Kansas, it's best to use a Kansas SpeedTest server rather than coming back across the country to a server in NC/SC isn't it?

Do I have an accurate take on things, or am I missing something?

True - routing is stochastic and need not take any direct paths. 100% correct.

In my case (and others) there are no speedtest servers anywhere near me. So, if I'm looking from services West, I don't speedtest servers East.

If you're lucky enough to be in close proximity to a server definitely use that, it's showing the shortest path from phone to tower to local intertubes - after that, the services are just the services. That shows your BEST CASE 3G response.

Still won't show the actual server path response.

When in doubt, using traceroute is the best bet - but I'm not remembering an easy way to do that directly from the phone - ConnectBot isn't helping.
 
I've always assumed that SpeedTest looked at where you jumped off Sprint's network and hit the public intarwebz. So my phone goes 3G for less than a mile to the nearest tower, and in the cabinet or hut at the base of that tower, there is likely a Fiber Mux of some sort. Then, I travel through fiber (back in the day, it was T1) across many networks until I get to Wichita Kansas (I live in SC/NC) where Sprint's servers connect me to the world.

WHen I use 4G however, I usually come out in a different place. Though with either 4G or 3G, it is not always the same point. Sometimes I connect in Atlanta, and other times, I connect in NC. It's all over the place, so I am not sure what accounts for that.

That being said, it seems to me that using GPS in order to make SpeedTest pick a geographically closer server could throw off the results. The geographically closest server isn't always the the closest server with regards to networks.

If I am indeed leaving Sprint's network in Wichita Kansas, it's best to use a Kansas SpeedTest server rather than coming back across the country to a server in NC/SC isn't it?

Do I have an accurate take on things, or am I missing something?

I'd bet it bases your "location" on your IP address. You'd surely get a different "location" between 3G (Sprint) and 4G (Clear), but I don't know if either has multiple POPs.

Either way, I trust the app almost as much as I trust task killers.
 
Back
Top Bottom