Closed GoogleCodeExporter closed 9 years ago
Original comment by pmerl...@googlemail.com
on 28 Jul 2010 at 8:20
I don't think this is due to the echo cancellation not working...
I think this is related more to the app playing audio through the rear
"external" speaker instead of the earpiece.
Original comment by jared.ya...@gmail.com
on 31 Jul 2010 at 2:44
Hoping this gets fixed soon, it's definitely due to audio playing through the
rear speaker instead of the proper earpiece.
Original comment by longmad
on 13 Aug 2010 at 2:15
same here - the sound comes out of the rear speaker on my i9000 (and therefore
gets recorded again?). Firmware JM5 (Android 2.1).
Original comment by markus.d...@gmail.com
on 17 Aug 2010 at 7:23
Same for me, sound come out from the rear speaker.
Samsung galaxy S i9000, firmware JG4, Singapore set
Hope there will be a fix soon.
Original comment by wiist...@gmail.com
on 29 Aug 2010 at 11:42
Here same problem. JM5+lag fix. Irritating echo problem. Also PCMU doesn't
work. Please fix those.
Also please add iLBC or/and G729 support!
Original comment by jar...@gmail.com
on 2 Sep 2010 at 11:11
no change with latest betas (1.57), samsung firmware JM7, and issue still
market as duplicate :-(
will there be any progress here?
Original comment by markus.d...@gmail.com
on 7 Sep 2010 at 7:46
I'm also with the same issue issue on the galaxy s i9000. now looking for
alternatives
Original comment by stichly...@gmail.com
on 17 Sep 2010 at 12:53
I have the same problem on the Samsung Galaxy S I9000.
Original comment by bjo...@gmail.com
on 19 Sep 2010 at 1:42
[deleted comment]
Lowering the gain on the mics makes the echo issue manageable for me on my
galaxy s i9000...
Original comment by setha...@gmail.com
on 24 Sep 2010 at 3:26
Yes, for lowering mic gain helped in my case as well. Please make it as a
default on Galaxy S!
Original comment by r...@dlux.hu
on 24 Sep 2010 at 5:59
[deleted comment]
... still no solution (my other post was right after some more tests). Sound
definitely comes out of the rear speaker even it should not be activated...
even with low mic gain, no good.
Original comment by markus.d...@gmail.com
on 24 Sep 2010 at 10:47
To sipdroid developers : something I want to share with you :
I've just found how to solve the issue on galaxy s i9000.
I'm the dev of another opensource sip application (CSipSimple) and I had the
same problem than you with galaxy S.
I've just found a solution today ! Samsung did something just awful with their
audio implementation :)... And, as I think that it can save time for you here
is the tip :
You should do
audioManager.setMode(AudioManager.MODE_IN_CALL);
audioManager.setMode(AudioManager.MODE_NORMAL);
before you start the stream/switch back to earpiece.
Setting mode to in_call is a good way to reset the audio routes. Else speaker
is always on keeping the last value. and if you ring previously... it routes
things to speaker.
If you need more details, my code is opensource and available on googlecode ;)
(relevant file is in src.com.csipsimple.service.MediaManager)
I hope that it will help you ;)
Original comment by r3gis...@gmail.com
on 29 Sep 2010 at 11:55
i can confirm using csipsimple works - ringtone speaker is disabled.
Original comment by markus.d...@gmail.com
on 30 Sep 2010 at 8:21
Based on r3gis' idea I have compiled an APK to test. Does it work better, maybe
on other Samsung phones, too?
Original comment by pmerl...@googlemail.com
on 1 Oct 2010 at 1:30
Tried pmerl's sipdroid.apk from Comment 17 on my SPH-D700. Got a few
force-closes before I could get it configured, and audio routing is unchanged.
Original comment by kean...@gmail.com
on 1 Oct 2010 at 4:49
With galaxy s it works better now. Volume control still not working... With low
settings I don't hear echo. Needs more testing but this goes right direction
now!
Original comment by jar...@gmail.com
on 1 Oct 2010 at 5:54
Unfortunately, it still seems to be initially routing to the rear speaker when
I answer the call.
Samsung Galaxy S I9000 (rooted with XWJM8 firmware + Voodoo Lagfix, Android 2.1)
NOTE: If I then do the "put call on hold, then take call off hold" trick... the
audio is only routed to the earpiece only. ????
However, there is still a problem with this... you can only go "on hold" once.
If you get a call, put it on hold, take it off hold, finish your call and then
receive a 2nd call and put it on hold, when you try to take it off hold, it
never picks the data stream back up (the box changes from orange to green,
message says call in progress but the display continues to show "No Data",
where it usually shows packet loss and latency etc.)
You have to exit out of SipDroid and restart, and then you can use hold again
(but again, only once)...
Is it worth logging this as a separate issue?
Original comment by jared.ya...@gmail.com
on 1 Oct 2010 at 8:16
This issue seems to be fixed on latest CSipSimple. However it is still present
on lates SipDroid and pmerl's fix, on my samsung galaxy s.
Original comment by furkansa...@gmail.com
on 3 Oct 2010 at 9:04
OK, please also test my second attempt as attached.
@jared: issue 294 has already been opened separately.
Original comment by pmerl...@googlemail.com
on 4 Oct 2010 at 2:05
No change on my Epic.
FWIW, I've noticed that when I plug headphones or a stereo cable into the A/V
jack, audio continues to come out of the speaker /and/ the connected device.
Original comment by kean...@gmail.com
on 4 Oct 2010 at 2:50
I tried it on my i9000 (jm8) and with headset it is much better (still a little
echo but it is usable now) without headset echo is also less but still quiet
obvious. So on the right track not quite there ;)
Original comment by wim.gard...@gmail.com
on 5 Oct 2010 at 2:55
@pmerl
no change on my galaxy s (vibrant). Still the external speaker.
CSipSimple and SipDroid are both open source projects. Why is it so hard for
one to solve this issue, while the other has already solved it?
Original comment by furkansa...@gmail.com
on 8 Oct 2010 at 12:01
That's difficult cause our two approach are different. I have a setup phase of
the in call mode called back from a native stack while it's done in the media
flow on sipdroid.
But really you should not blame pmerl for that. What I found is a hack that
fortunately works... who should be blamed are manufacturer's devs that
implemented their audio drivers without making it robust and coherent!
That's extremely difficult for devs to get something that works for all
platforms.
On my side I've still big issues with x10 (mic records mic +earpiece which
introduce a big echo). And even on galaxy s I will need to tweak default gains
to get something without echo... and I'm still not able to control in call
volume using android audio API on galaxy s. Once again...it's Samsung that did
crappy things!
@pmerl :
Something that could help you :
Whenever I switch from/to the earpiece I restart the audio track stream.
Probably hard to do with your implementation (was easy for me using pjsip
design). I found that it helps a lot for Bluetooth and as it was not so hard to
do I did it for each routing changes.
Also it's done after the ringer is stopped. If you need more details on
something on my code don't hesitate to mail me. If we could help each other on
this kind of issues could be profitable to users either if they choose
csipsimple or sipdroid.
Original comment by r3gis...@gmail.com
on 8 Oct 2010 at 6:21
Third attempt is uploaded as 1.6pre3. Is that one OK?
Original comment by pmerl...@googlemail.com
on 8 Oct 2010 at 2:32
No sound coming at all.
Original comment by furkansa...@gmail.com
on 8 Oct 2010 at 5:22
Sorry. I uploaded 1.6pre4.
Original comment by pmerl...@googlemail.com
on 8 Oct 2010 at 10:29
Still external speaker.
Original comment by furkansa...@gmail.com
on 9 Oct 2010 at 2:52
[deleted comment]
I tested it as well, audio still routed over the external speaker. As was
always the case with this issue, it happens on incoming calls every time.
Outgoing calls are routed over the correct speaker most of the time, that said
for some reason it still appears to be echoing on the other side on Outgoing
calls as well.
Original comment by schip...@gmail.com
on 11 Oct 2010 at 10:48
[deleted comment]
@pmerl :
I also tried out the provided patched apk you posted and can confirm schiphol's
observations.
Sipdroid.apk
844 KB
Using the I9000 when I make outgoing calls it seems to route correctly, but
when receiving calls it is still through the loud speaker. Thanks a bunch for
the attempt.
I will give csipsimple a try to see if it has the same problem.
Edit #1: Tried csipsimple it INDEED solves the incoming problem correctly...
HOWEVER the outgoing call routes to the loud speaker... hahaha... If I could
use sipdroid for outgoing and csipsimple for incoming I would be set ^_^
Original comment by gum...@gmail.com
on 12 Oct 2010 at 12:36
Just to complete my earlier post and in line with gumpwa's post #34 for the
Galaxy S i9000 with JM2 (2.1 Eclair). The version I tested was
Sipdroid-1.6pre4.apk.
Assuming the audio routing issue has to do with Samsung's coding on the driver,
I was wondering if someone has tested sipdroid on one of the leaked froyo
releases for the Galaxy S. If it's fixed on those releases that might came with
updated drivers I guess we can wait for the final release by Samsung (in 2
weeks hopefully)
@pmerl: Hope you can invest a little bit more time into this issue and make us
who bought a samsung phone happy. I really need pbxes to work when I'm away and
my company already gave away our nokia phones which worked perfect every call.
Original comment by schip...@gmail.com
on 12 Oct 2010 at 12:59
I have tried sipdroid using the JPK 2.2 Froyo on the I9000 and I still saw the
same issues. From what I understand the JPK is a pretty recent ROM so it does
not look like the beta froyo has addressed this problem yet.
Original comment by gum...@gmail.com
on 12 Oct 2010 at 1:03
Original comment by pmerl...@googlemail.com
on 14 Oct 2010 at 8:50
Thanks guys! This is awesome, completely changes the sipdroid experience on the
Galaxy S. I love how even the proximity sensor is fixed in this release. I'm
loving it, Ich liebe es!
Original comment by schip...@gmail.com
on 14 Oct 2010 at 9:30
Original issue reported on code.google.com by
albert.h...@gmail.com
on 28 Jul 2010 at 11:50