Problem with MRCPSynth on Hyper-V
114 views
Skip to first unread message
Mulai
May 30, 2012, 11:40:41 AM5/30/12
to UniMRCP
Hello Arsen,
I try to use Asterisk+UniMRCP on Hyper-V Guest OS (CentOS 5 and 6). I
installed Integration Tools and everything is working correct.
But I have problem with synthesis.
When I'm using MRCPSynth in dialplan, can be heard noises, cracks and
gnashing.
If I wrote sound on MRCP-server, it's OK. Problem is in Asterisk.
I didn't see errors in logs or Asterisk's console.
May you tell what could be the problem?
I try to use Asterisk+UniMRCP on Hyper-V Guest OS (CentOS 5 and 6). I
installed Integration Tools and everything is working correct.
But I have problem with synthesis.
When I'm using MRCPSynth in dialplan, can be heard noises, cracks and
gnashing.
If I wrote sound on MRCP-server, it's OK. Problem is in Asterisk.
I didn't see errors in logs or Asterisk's console.
May you tell what could be the problem?
Arsen Chaloyan
Jun 6, 2012, 8:38:30 PM6/6/12
to uni...@googlegroups.com
Hello Kirill,
First, I'd check RTP statistics logged at the end of each MRCP session
as follows
Close RTP Receiver [r:110 l:0 j:0 p:50 d:0 i:0],
where
r: received
l: lost
j: jitter
p: playout delay
d: dropped
i: ignored
Next, if you're using r1835 or later, then I'd suggest to try the
adaptive mode with the following parameters, for example,
<jitter-buffer>
<adaptive>1</adaptive>
<playout-delay>50</playout-delay>
<max-playout-delay>1000</max-playout-delay>
<time-skew-detection>1</time-skew-detection>
</jitter-buffer>
> --
> You received this message because you are subscribed to the Google Groups "UniMRCP" group.
> To post to this group, send email to uni...@googlegroups.com.
> To unsubscribe from this group, send email to unimrcp+u...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/unimrcp?hl=en.
>
--
Arsen Chaloyan
Author of UniMRCP
http://www.unimrcp.org
First, I'd check RTP statistics logged at the end of each MRCP session
as follows
Close RTP Receiver [r:110 l:0 j:0 p:50 d:0 i:0],
where
r: received
l: lost
j: jitter
p: playout delay
d: dropped
i: ignored
Next, if you're using r1835 or later, then I'd suggest to try the
adaptive mode with the following parameters, for example,
<jitter-buffer>
<adaptive>1</adaptive>
<playout-delay>50</playout-delay>
<max-playout-delay>1000</max-playout-delay>
<time-skew-detection>1</time-skew-detection>
</jitter-buffer>
> You received this message because you are subscribed to the Google Groups "UniMRCP" group.
> To post to this group, send email to uni...@googlegroups.com.
> To unsubscribe from this group, send email to unimrcp+u...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/unimrcp?hl=en.
>
--
Arsen Chaloyan
Author of UniMRCP
http://www.unimrcp.org
Mulai
Jun 7, 2012, 6:35:22 AM6/7/12
to UniMRCP
Arsen, thanks.
I use asterisk-unimrcp-0.3.2.
May you explain how can I check RTP statistics? I don't know what is
RTP Receiver and how I can use it.
Where is jitter-buffer parameters you are writing about?
I'm using mrcp.conf for all settings.
Default is:
; --- Jitter buffer settings ---
playout-delay = 50
; min-playout-delay = 20
max-playout-delay = 200
There are cracks always when synthesis is starting.
If I increase playout-delay cracks become louder and stronger. playout-
delay = 10 cracks are softer but still heard.
> > For more options, visit this group athttp://groups.google.com/group/unimrcp?hl=en.
I use asterisk-unimrcp-0.3.2.
May you explain how can I check RTP statistics? I don't know what is
RTP Receiver and how I can use it.
Where is jitter-buffer parameters you are writing about?
I'm using mrcp.conf for all settings.
Default is:
; --- Jitter buffer settings ---
playout-delay = 50
; min-playout-delay = 20
max-playout-delay = 200
There are cracks always when synthesis is starting.
If I increase playout-delay cracks become louder and stronger. playout-
delay = 10 cracks are softer but still heard.
Message has been deleted
Arsen Chaloyan
Jun 15, 2012, 5:54:02 PM6/15/12
to uni...@googlegroups.com
Hi Kirill,
On Thu, Jun 7, 2012 at 3:35 AM, Mulai <kir....@gmail.com> wrote:
> Arsen, thanks.
>
> I use asterisk-unimrcp-0.3.2.
> May you explain how can I check RTP statistics? I don't know what is
> RTP Receiver and how I can use it.
You can check the mentioned RTP stats from the log file.
Set log-level in /etc/asterisk/mrcp.conf. to debug
; UniMRCP logging level to appear in Asterisk logs. Options are:
; EMERGENCY|ALERT|CRITICAL|ERROR|WARNING|NOTICE|INFO|DEBUG -->
log-level = DEBUG
If you want them to see on console of Asterisk, set the following in
/etc/asterisk/logger.conf
console => notice,warning,error,debug
>
> Where is jitter-buffer parameters you are writing about?
> I'm using mrcp.conf for all settings.
> Default is:
> ; --- Jitter buffer settings ---
> playout-delay = 50
> ; min-playout-delay = 20
> max-playout-delay = 200
They are not available with the version you're using.
>
> There are cracks always when synthesis is starting.
> If I increase playout-delay cracks become louder and stronger. playout-
> delay = 10 cracks are softer but still heard.
I need to see the incoming RTP stream to make any comments. Wireshark
capture might be helpful.
> For more options, visit this group at http://groups.google.com/group/unimrcp?hl=en.
On Thu, Jun 7, 2012 at 3:35 AM, Mulai <kir....@gmail.com> wrote:
> Arsen, thanks.
>
> I use asterisk-unimrcp-0.3.2.
> May you explain how can I check RTP statistics? I don't know what is
> RTP Receiver and how I can use it.
Set log-level in /etc/asterisk/mrcp.conf. to debug
; UniMRCP logging level to appear in Asterisk logs. Options are:
; EMERGENCY|ALERT|CRITICAL|ERROR|WARNING|NOTICE|INFO|DEBUG -->
log-level = DEBUG
If you want them to see on console of Asterisk, set the following in
/etc/asterisk/logger.conf
console => notice,warning,error,debug
>
> Where is jitter-buffer parameters you are writing about?
> I'm using mrcp.conf for all settings.
> Default is:
> ; --- Jitter buffer settings ---
> playout-delay = 50
> ; min-playout-delay = 20
> max-playout-delay = 200
>
> There are cracks always when synthesis is starting.
> If I increase playout-delay cracks become louder and stronger. playout-
> delay = 10 cracks are softer but still heard.
capture might be helpful.
Mulai
Jun 19, 2012, 9:38:03 AM6/19/12
to uni...@googlegroups.com
Hi Arsen,
In attach is log Asterisk from console with enabled rtp debug and debug settings in logger.conf.
>> > To unsubscribe from this group, send email to unimrcp+unsubscribe@googlegroups.com.
>> > For more options, visit this group athttp://groups.google.com/group/unimrcp?hl=en.
>>
>> --
>> Arsen Chaloyan
>> Author of UniMRCPhttp://www.unimrcp.org
>
> --
> You received this message because you are subscribed to the Google Groups "UniMRCP" group.
> To post to this group, send email to uni...@googlegroups.com.
> To unsubscribe from this group, send email to unimrcp+unsubscribe@googlegroups.com.
Mulai
Jun 19, 2012, 10:32:37 AM6/19/12
to uni...@googlegroups.com
In attach dump by tcpdump.
On Saturday, June 16, 2012 1:54:02 AM UTC+4, Arsen Chaloyan wrote:
On Saturday, June 16, 2012 1:54:02 AM UTC+4, Arsen Chaloyan wrote:
>> > To unsubscribe from this group, send email to unimrcp+unsubscribe@googlegroups.com.
>> > For more options, visit this group athttp://groups.google.com/group/unimrcp?hl=en.
>>
>> --
>> Arsen Chaloyan
>> Author of UniMRCPhttp://www.unimrcp.org
>
> --
> You received this message because you are subscribed to the Google Groups "UniMRCP" group.
> To post to this group, send email to uni...@googlegroups.com.
> To unsubscribe from this group, send email to unimrcp+unsubscribe@googlegroups.com.
Arsen Chaloyan
Jul 27, 2012, 7:23:08 PM7/27/12
to uni...@googlegroups.com
Hi Kirill,
I'd like to bring your attention to the media path and codecs.
[Jun 19 23:22:35] DEBUG[2589]: app_unimrcp.c:4172 unimrcp_log: Media
Path 0x2dbf8b8 Source->[PCMU/8000/1]->Decoder->[LPCM/8000/1]->Bridge->[LPCM/8000/1]->Encoder->[PCMA/8000/1]->Sink
It looks like PCMU was used on the MRCP leg while the inbound call was
established with PCMA. This shouldn't cause any problem in theory, but
try to avoid the conversion by using the same codec on both legs. The
cracks you heard may come from the conversion, I guess. Otherwise, the
RTP stream wasn't that bad.
>> >> > unimrcp+u...@googlegroups.com.
> https://groups.google.com/d/msg/unimrcp/-/-lGbDbL8gpUJ.
I'd like to bring your attention to the media path and codecs.
[Jun 19 23:22:35] DEBUG[2589]: app_unimrcp.c:4172 unimrcp_log: Media
Path 0x2dbf8b8 Source->[PCMU/8000/1]->Decoder->[LPCM/8000/1]->Bridge->[LPCM/8000/1]->Encoder->[PCMA/8000/1]->Sink
It looks like PCMU was used on the MRCP leg while the inbound call was
established with PCMA. This shouldn't cause any problem in theory, but
try to avoid the conversion by using the same codec on both legs. The
cracks you heard may come from the conversion, I guess. Otherwise, the
RTP stream wasn't that bad.
>> >> > For more options, visit this group
>> >> > athttp://groups.google.com/group/unimrcp?hl=en.
>> >>
>> >> --
>> >> Arsen Chaloyan
>> >> Author of UniMRCPhttp://www.unimrcp.org
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups "UniMRCP" group.
>> > To post to this group, send email to uni...@googlegroups.com.
>> > To unsubscribe from this group, send email to
>> > unimrcp+u...@googlegroups.com.
>> >> > athttp://groups.google.com/group/unimrcp?hl=en.
>> >>
>> >> --
>> >> Arsen Chaloyan
>> >> Author of UniMRCPhttp://www.unimrcp.org
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups "UniMRCP" group.
>> > To post to this group, send email to uni...@googlegroups.com.
>> > To unsubscribe from this group, send email to
>> > For more options, visit this group at
>> > http://groups.google.com/group/unimrcp?hl=en.
>> >
>>
>>
>>
>> --
>> Arsen Chaloyan
>> Author of UniMRCP
>> http://www.unimrcp.org
>
>> > http://groups.google.com/group/unimrcp?hl=en.
>> >
>>
>>
>>
>> --
>> Arsen Chaloyan
>> Author of UniMRCP
>> http://www.unimrcp.org
>
> --
> You received this message because you are subscribed to the Google Groups
> "UniMRCP" group.
> To view this discussion on the web visit
> You received this message because you are subscribed to the Google Groups
> "UniMRCP" group.
> https://groups.google.com/d/msg/unimrcp/-/-lGbDbL8gpUJ.
>
> To post to this group, send email to uni...@googlegroups.com.
> To unsubscribe from this group, send email to
> unimrcp+u...@googlegroups.com.
> To post to this group, send email to uni...@googlegroups.com.
> To unsubscribe from this group, send email to
Mulai
Jul 28, 2012, 3:20:30 AM7/28/12
to uni...@googlegroups.com
Thanks Arsen.
>> >> > unimrcp+unsubscribe@googlegroups.com.
>> >> > For more options, visit this group
>> >> > athttp://groups.google.com/group/unimrcp?hl=en.
>> >>
>> >> --
>> >> Arsen Chaloyan
>> >> Author of UniMRCPhttp://www.unimrcp.org
>> >
>> > --
>> > You received this message because you are subscribed to the Google
>> > Groups "UniMRCP" group.
>> > To post to this group, send email to uni...@googlegroups.com.
>> > To unsubscribe from this group, send email to
>> > For more options, visit this group at
>> > http://groups.google.com/group/unimrcp?hl=en.
>> >
>>
>>
>>
>> --
>> Arsen Chaloyan
>> Author of UniMRCP
>> http://www.unimrcp.org
>
> --
> You received this message because you are subscribed to the Google Groups
> "UniMRCP" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/unimrcp/-/-lGbDbL8gpUJ.
>
> To post to this group, send email to uni...@googlegroups.com.
> To unsubscribe from this group, send email to
Reply all
Reply to author
Forward
0 new messages