[asterisk-dev] video-call recording

Discussion:

Kévin Bouchard

2007-11-13 18:46:46 UTC

Hi,

I'd like to know whether or not someone is already working on a video
call recording feature. I'm aware of app_mp4 which, to my knowledge,
only allow recording of an incoming stream and not a call like
chan_spy or MixMonitor.
Someone already suggested to use Record with h263 format but can't get
it working. I'm willing to start the development of the feature and if
someone already work on it, give a helping hand for completing the
work. However, I'd like to know which format would be preferred for
saving the files and start a little bit of discussion about licensing.
Is FFmpeg correct? Or MPEG4IP?

Thanks,

Kevin.

Klaus Darilion

2007-11-13 19:44:12 UTC

Permalink

There was a thread on the list (mayne 2 months ago?) about this someone
mentioned Matroska as container format.
e.g.
http://lists.digium.com/pipermail/asterisk-video/2007-August/001130.html

regards
klaus

Post by KÃ©vin Bouchard
Hi,
I'd like to know whether or not someone is already working on a video
call recording feature. I'm aware of app_mp4 which, to my knowledge,
only allow recording of an incoming stream and not a call like
chan_spy or MixMonitor.
Someone already suggested to use Record with h263 format but can't get
it working. I'm willing to start the development of the feature and if
someone already work on it, give a helping hand for completing the
work. However, I'd like to know which format would be preferred for
saving the files and start a little bit of discussion about licensing.
Is FFmpeg correct? Or MPEG4IP?
Thanks,
Kevin.
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
asterisk-dev mailing list
http://lists.digium.com/mailman/listinfo/asterisk-dev

Olle E Johansson

2007-11-13 20:05:37 UTC

Permalink

If you record the audio from a call with a video stream, the video
will be saved as well.
Just don't mention it in the record() or playback() statements. If
there's a video file
with the same name, it will be recorded or played back.

The actual format is very raw RTP data, that ffmpeg has a driver for
reading and writing.

/O

Matthew Yingling

2007-11-14 16:54:25 UTC

Permalink

Hi,

I was testing a PC set up with Asterisk 1.4.8 and Zaptel 1.4.4 and
discovered that hold music would play, but anything requiring a voice prompt
would hang indefinitely. I came across this issue, which is marked closed.
I am able to replicate this exact behavior. I am not using any zaptel
hardwar and if ztdummy is running, no voice prompts will play, but hold
music will. If I unload ztdummy, voice prompts and music both work, but
MeetMe() does not. It works whether Zaptel is loaded or not. What is
interesting is that I just tried this same installation in a different PC
and it works fine.

I am running a live installation (Slax 5.1.8.1 with kernel 2.6.16) of
Asterisk on a USB stick, so the complete OS and configuration is identical
on both systems, but one exhibits this issue while the other does not. Can
anyone suggest a way to determine whether this is a Zaptel or Linux kernel
issue?

http://bugs.digium.com/view.php?id=11153

Thanks,
Matthew

Tony Mountifield

2007-11-14 18:03:50 UTC

Permalink

Post by Matthew Yingling
Hi,
I was testing a PC set up with Asterisk 1.4.8 and Zaptel 1.4.4 and
discovered that hold music would play, but anything requiring a voice prompt
would hang indefinitely. I came across this issue, which is marked closed.
I am able to replicate this exact behavior. I am not using any zaptel
hardwar and if ztdummy is running, no voice prompts will play, but hold
music will. If I unload ztdummy, voice prompts and music both work, but
MeetMe() does not. It works whether Zaptel is loaded or not. What is
interesting is that I just tried this same installation in a different PC
and it works fine.
I am running a live installation (Slax 5.1.8.1 with kernel 2.6.16) of
Asterisk on a USB stick, so the complete OS and configuration is identical
on both systems, but one exhibits this issue while the other does not. Can
anyone suggest a way to determine whether this is a Zaptel or Linux kernel
issue?

Sounds like on the bad system ztdummy is not getting nay interrupts.

Try running zttest on each system and see what it shows. Also check
the output of "cat /proc/interrupts" on each system. You should have
an entry for "rtc" counting up at about 1024 per second.

ztdummy and zaptel should be loaded and running for the above tests.

What is the make and model of each PC?

Cheers
Tony

--
Tony Mountifield
Work: ***@softins.co.uk - http://www.softins.co.uk
Play: ***@mountifield.org - http://tony.mountifield.org

Eric "ManxPower" Wieling

2007-11-14 18:08:59 UTC

Permalink

There is a thread on -users this week about a no-audio issue as well as
the solution someone found.

Post by Tony Mountifield

Sounds like on the bad system ztdummy is not getting nay interrupts.
Try running zttest on each system and see what it shows. Also check
the output of "cat /proc/interrupts" on each system. You should have
an entry for "rtc" counting up at about 1024 per second.
ztdummy and zaptel should be loaded and running for the above tests.
What is the make and model of each PC?
Cheers
Tony

Matthew Yingling

2007-11-14 21:38:51 UTC

Permalink

Thanks for the replies. For anyone looking at -users, search for the
"ztdummy, zttest" thread. That does appear to match the issue I am
observing.

The working PC is a Dell Precision M65 laptop (Core Duo based), and the
non-working PC is an HP Compaq dx2250 MT (Sempron based).

I loaded ztdummy on both machines and then ran zttest. The working PC shows
a 99.96% accuracy and the rtc interrupt counting about 1024 per second. The
non-working PC shows no zttest output (it just waits after the start
message) and a constant zero rtc interrupt count.

Unfortunately I won't be able to test a new kernel anytime soon to see if
that fixes this issue. Does anyone have information on the kernel patch?

Matthew

-----Original Message-----
From: asterisk-dev-***@lists.digium.com
[mailto:asterisk-dev-***@lists.digium.com] On Behalf Of Eric "ManxPower"
Wieling
Sent: Wednesday, November 14, 2007 1:09 PM
To: Asterisk Developers Mailing List
Subject: Re: [asterisk-dev] Zaptel Issue 11153

There is a thread on -users this week about a no-audio issue as well as
the solution someone found.

Post by Tony Mountifield

Can

Post by Tony Mountifield

Post by Matthew Yingling
anyone suggest a way to determine whether this is a Zaptel or Linux kernel
issue?

Sounds like on the bad system ztdummy is not getting nay interrupts.
Try running zttest on each system and see what it shows. Also check
the output of "cat /proc/interrupts" on each system. You should have
an entry for "rtc" counting up at about 1024 per second.
ztdummy and zaptel should be loaded and running for the above tests.
What is the make and model of each PC?
Cheers
Tony

Matthew Yingling

2007-11-14 22:05:16 UTC

Permalink

I noticed that zttest reports "Best: 0.000000 -- Worst: 100.000000 --
Average: 100.000000" for zero runs. Is this a minor bug? It seems to me
that zero runs should report 0.000000%. On the other hand, this is an
unexpected edge case, so the output doesn't matter much.

-----Original Message-----
From: asterisk-dev-***@lists.digium.com
[mailto:asterisk-dev-***@lists.digium.com] On Behalf Of Matthew Yingling
Sent: Wednesday, November 14, 2007 4:39 PM
To: 'Asterisk Developers Mailing List'
Subject: Re: [asterisk-dev] Zaptel Issue 11153

Thanks for the replies. For anyone looking at -users, search for the
"ztdummy, zttest" thread. That does appear to match the issue I am
observing.

The working PC is a Dell Precision M65 laptop (Core Duo based), and the
non-working PC is an HP Compaq dx2250 MT (Sempron based).

I loaded ztdummy on both machines and then ran zttest. The working PC shows
a 99.96% accuracy and the rtc interrupt counting about 1024 per second. The
non-working PC shows no zttest output (it just waits after the start
message) and a constant zero rtc interrupt count.

Unfortunately I won't be able to test a new kernel anytime soon to see if
that fixes this issue. Does anyone have information on the kernel patch?

Matthew

-----Original Message-----
From: asterisk-dev-***@lists.digium.com
[mailto:asterisk-dev-***@lists.digium.com] On Behalf Of Eric "ManxPower"
Wieling
Sent: Wednesday, November 14, 2007 1:09 PM
To: Asterisk Developers Mailing List
Subject: Re: [asterisk-dev] Zaptel Issue 11153

There is a thread on -users this week about a no-audio issue as well as
the solution someone found.

Post by Tony Mountifield

Can

Post by Tony Mountifield

Post by Matthew Yingling
anyone suggest a way to determine whether this is a Zaptel or Linux kernel
issue?

Sounds like on the bad system ztdummy is not getting nay interrupts.
Try running zttest on each system and see what it shows. Also check
the output of "cat /proc/interrupts" on each system. You should have
an entry for "rtc" counting up at about 1024 per second.
ztdummy and zaptel should be loaded and running for the above tests.
What is the make and model of each PC?
Cheers
Tony

_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--

asterisk-dev mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-dev

_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--

asterisk-dev mailing list
To UNSUBSCRIBE or update options visit:
http://lists.digium.com/mailman/listinfo/asterisk-dev

Tony Plack

2007-11-14 18:57:31 UTC

Permalink

<html><head><meta name="Generator" content="PSI HTML/CSS Generator"/>
<style type="text/css"></style> </head><BODY BGCOLOR="#CED1E1" ><div><SPAN style="font-size:10pt;">The issue for me was a potential bug with tsc clock source on some motherboards in kernel version 2.6.18 this week.</SPAN></div> <div> </div> <div><SPAN style="font-size:10pt;">I have upgraded to 2.6.23.1 and this solved the problem.  Regardless of the below, my bet would be a kernel version issue.</SPAN></div> <div> </div> <div><SPAN style="font-size:10pt;">I was also able to make this work by using noacpi option on the kernel, however, this broke the rest of the system because IRQ's were not assigned correctly.</SPAN></div> <div> </div> <div><SPAN style="font-size:10pt;">There is some indication that the ACPI settings on the BIOS for power settings does impact this as well.<br /><br /></span></div> <div><SPAN style="font-size:10pt;color:navy;">> Hi,</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">></SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> I was testing a PC set up with Asterisk 1.4.8 and Zaptel 1.4.4 and</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> discovered that hold music would play, but anything requiring a</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> voice prompt would hang indefinitely.  I came across this issue,</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> which is marked closed. I am  able to replicate this exact</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> behavior.  I am not using any zaptel hardwar and if ztdummy is</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> running, no voice prompts will play, but hold music will.  If I</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> unload ztdummy, voice prompts and music both work, but MeetMe()</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> does not.  It works whether Zaptel is loaded or not.  What is</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> interesting is that I just tried this same installation in a</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> different PC and it works fine.</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">></SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> I am running a live installation (Slax 5.1.8.1 with kernel 2.6.16)</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> of Asterisk on a USB stick, so the complete OS and configuration is</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> identical on both systems, but one exhibits this issue while the</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> other does not.  Can anyone suggest a way to determine whether this</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> is a Zaptel or Linux kernel issue?</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">></SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> <a href="http://bugs.digium.com/view.php?id=11153">http://bugs.digium.com/view.php?id=11153</a></SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">></SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> Thanks,</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> Matthew</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">></SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">></SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> _______________________________________________</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> --Bandwidth and Colocation Provided by <a href="http://www.api-digital.com">http://www.api-digital.com</a>--</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">></SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> asterisk-dev mailing list</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> To UNSUBSCRIBE or update options visit:</SPAN></div> <div><SPAN style="font-size:10pt;color:navy;">> <a href="http://lists.digium.com/mailman/listinfo/asterisk-dev">http://lists.digium.com/mailman/listinfo/asterisk-dev</a></SPAN></div>
</body></html>

Tony Plack

2007-11-14 22:21:27 UTC

Permalink

Not directly. I kind of figured it out through a fuzzy logic after searching Google on TSC, HPET and these errors. I had a tack on the issue being in the kernel on Friday.

The article that was the lynch pin for me was https://bugs.launchpad.net/ubuntu/+source/kqemu/+bug/82149.

The conclusion of the article was that there was a problem in 2.6.20 in Ubuntu and they needed to recompile with different options that were not in 2.6.18 directly. That was Monday.

Tuesday, I woke up and realized it must be a kernel version issue. I tried to read through the change logs from .20 to .23.1 and saw there were many changes to the timer routines.

Seems to me to be more with the Core 2 products and multiple CPUs.

Tony Plack

Post by Matthew Yingling
Thanks for the replies. For anyone looking at -users, search for
the "ztdummy, zttest" thread. That does appear to match the issue
I am observing.
The working PC is a Dell Precision M65 laptop (Core Duo based), and
the non-working PC is an HP Compaq dx2250 MT (Sempron based).
I loaded ztdummy on both machines and then ran zttest. The working
PC shows a 99.96% accuracy and the rtc interrupt counting about
1024 per second. The non-working PC shows no zttest output (it
just waits after the start message) and a constant zero rtc
interrupt count.
Unfortunately I won't be able to test a new kernel anytime soon to
see if that fixes this issue. Does anyone have information on the
kernel patch?
Matthew
-----Original Message-----
Asterisk Developers Mailing List Subject: Re: [asterisk-dev] Zaptel
Issue 11153
There is a thread on -users this week about a no-audio issue as
well as the solution someone found.

Post by Tony Mountifield

Post by Matthew Yingling
Hi,
I was testing a PC set up with Asterisk 1.4.8 and Zaptel 1.4.4
and discovered that hold music would play, but anything
requiring a voice

prompt

Post by Tony Mountifield

Post by Matthew Yingling
would hang indefinitely. I came across this issue, which is
marked

closed.

Post by Tony Mountifield

Post by Matthew Yingling
I am able to replicate this exact behavior. I am not using
any zaptel hardwar and if ztdummy is running, no voice prompts
will play, but hold music will. If I unload ztdummy, voice
prompts and music both work, but MeetMe() does not. It works
whether Zaptel is loaded or not. What is interesting is that I
just tried this same installation in a different PC and it
works fine.
I am running a live installation (Slax 5.1.8.1 with kernel
2.6.16) of Asterisk on a USB stick, so the complete OS and
configuration is

identical

Post by Tony Mountifield

Post by Matthew Yingling
on both systems, but one exhibits this issue while the other
does not.

Can

Post by Tony Mountifield

Post by Matthew Yingling
anyone suggest a way to determine whether this is a Zaptel or
Linux

kernel

Post by Tony Mountifield

Post by Matthew Yingling
issue?

Sounds like on the bad system ztdummy is not getting nay
interrupts.
Try running zttest on each system and see what it shows. Also
check the output of "cat /proc/interrupts" on each system. You
should have an entry for "rtc" counting up at about 1024 per
second.
ztdummy and zaptel should be loaded and running for the above
tests.
What is the make and model of each PC?
Cheers
Tony

_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
asterisk-dev mailing list
http://lists.digium.com/mailman/listinfo/asterisk-dev
_______________________________________________
--Bandwidth and Colocation Provided by http://www.api-digital.com--
asterisk-dev mailing list
http://lists.digium.com/mailman/listinfo/asterisk-dev