Gstreamer echo canceller udpsrc and udpsink

[shrikantneotech]

Hi

I need to implement echo canceller in my audio application.

I am sending audio alsasrc to remote mobile android app from my hardware.

The command is as follows:

Sending and receive combined in single command:

"gst-launch-1.0 -v alsasrc ! audioresample ! audioconvert ! audio/x-raw, format=S16LE, layout=interleaved, rate=8000, channels=2, endianness=4321, width=16, depth=16, signed=true ! udpsink host=$1 port=5001 & gst-launch-1.0 -v --gst-debug-level=4 udpsrc port=5003 ! audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2, endianness=4321, width=16, depth=16, signed=true ! audioconvert ! autoaudiosink"

But right now I am doing push to talk to avoid echo and noise on the hardware side.

Please tell me how to implement AEC in GStreamer.....

Thanks,

Shrikant

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[Nicolas Dufresne-5]

Le mercredi 13 septembre 2017 à 22:36 +0530, Shrikant Lahase a écrit :
> Hi

Welcome to the mailings list.

>
> I need to implement echo canceller in my audio application.
> I am sending audio alsasrc to remote mobile android app from my hardware.
> The command is as follows:
>
> Sending and receive combined in single command:
>
> "gst-launch-1.0 -v alsasrc ! audioresample ! audioconvert ! audio/x-
> raw, format=S16LE, layout=interleaved, rate=8000, channels=2,
> endianness=4321, width=16, depth=16, signed=true ! udpsink host=$1
> port=5001 & gst-launch-1.0 -v --gst-debug-level=4  udpsrc port=5003 !
> audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2,
> endianness=4321, width=16, depth=16, signed=true ! audioconvert !
> autoaudiosink"
>

This was very recently asked on the list,

http://gstreamer-devel.966125.n4.nabble.com/Webrtc-plugin-for-AEC-support-td4684580.html

In this thread, I give a example pipeline for testing in loopback the
AEC. Note that you are using 0.10 syntax in your pipeline.

  audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2,endianness=4321, width=16, depth=16, signed=true

Is in fact:
  audio/x-raw,layout=interleaved,rate=8000,format=S16LE,channels=2

depth,width,signed,endianess have no meaning in 1.0 and may lead to
negotiation error. The in-gstreamer echo canceller works by placing
webrtcprobe close to your load speaker playback element, and placing
webrtcdsp filter close to your recording element.

Note that you'll always get better result due to timing precision with
using AEC in Pulse Audio rather then in Gst.

>
> But right now I am doing push to talk to avoid echo and noise on the hardware side.
>
> Please tell me how to implement AEC in GStreamer.....
>
>
> Thanks,
> Shrikant
> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[shrikantneotech]

Thanks for your reply.

For using "webrtcdsp" & "webrtcprobe" which plugin needs to install?

There is one more option I read on google "webrtcechoprobe" 

What is difference between  "webrtcprobe" & "webrtcechoprobe" 

What is attached signature.asc file?

On Wed, Sep 13, 2017 at 11:27 PM, Nicolas Dufresne <[hidden email]> wrote:

Le mercredi 13 septembre 2017 à 22:36 +0530, Shrikant Lahase a écrit :
> Hi

Welcome to the mailings list.

>
> I need to implement echo canceller in my audio application.
> I am sending audio alsasrc to remote mobile android app from my hardware.
> The command is as follows:
>
> Sending and receive combined in single command:
>
> "gst-launch-1.0 -v alsasrc ! audioresample ! audioconvert ! audio/x-
> raw, format=S16LE, layout=interleaved, rate=8000, channels=2,
> endianness=4321, width=16, depth=16, signed=true ! udpsink host=$1
> port=5001 & gst-launch-1.0 -v --gst-debug-level=4  udpsrc port=5003 !
> audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2,
> endianness=4321, width=16, depth=16, signed=true ! audioconvert !
> autoaudiosink"
>

This was very recently asked on the list,

http://gstreamer-devel.966125.n4.nabble.com/Webrtc-plugin-for-AEC-support-td4684580.html

In this thread, I give a example pipeline for testing in loopback the
AEC. Note that you are using 0.10 syntax in your pipeline.

  audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2,endianness=4321, width=16, depth=16, signed=true

Is in fact:
  audio/x-raw,layout=interleaved,rate=8000,format=S16LE,channels=2

depth,width,signed,endianess have no meaning in 1.0 and may lead to
negotiation error. The in-gstreamer echo canceller works by placing
webrtcprobe close to your load speaker playback element, and placing
webrtcdsp filter close to your recording element.

Note that you'll always get better result due to timing precision with
using AEC in Pulse Audio rather then in Gst.

>
> But right now I am doing push to talk to avoid echo and noise on the hardware side.
>
> Please tell me how to implement AEC in GStreamer.....
>
>
> Thanks,
> Shrikant
> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[shrikantneotech]

Is GStreamer  1.12 version is necessary for echo cancellation?

On Thu, Sep 14, 2017 at 9:49 AM, Shrikant Lahase <[hidden email]> wrote:

Thanks for your reply.

For using "webrtcdsp" & "webrtcprobe" which plugin needs to install?

There is one more option I read on google "webrtcechoprobe" 

What is difference between  "webrtcprobe" & "webrtcechoprobe" 

What is attached signature.asc file?

On Wed, Sep 13, 2017 at 11:27 PM, Nicolas Dufresne <[hidden email]> wrote:

Le mercredi 13 septembre 2017 à 22:36 +0530, Shrikant Lahase a écrit :
> Hi

Welcome to the mailings list.

>
> I need to implement echo canceller in my audio application.
> I am sending audio alsasrc to remote mobile android app from my hardware.
> The command is as follows:
>
> Sending and receive combined in single command:
>
> "gst-launch-1.0 -v alsasrc ! audioresample ! audioconvert ! audio/x-
> raw, format=S16LE, layout=interleaved, rate=8000, channels=2,
> endianness=4321, width=16, depth=16, signed=true ! udpsink host=$1
> port=5001 & gst-launch-1.0 -v --gst-debug-level=4  udpsrc port=5003 !
> audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2,
> endianness=4321, width=16, depth=16, signed=true ! audioconvert !
> autoaudiosink"
>

This was very recently asked on the list,

http://gstreamer-devel.966125.n4.nabble.com/Webrtc-plugin-for-AEC-support-td4684580.html

In this thread, I give a example pipeline for testing in loopback the
AEC. Note that you are using 0.10 syntax in your pipeline.

  audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2,endianness=4321, width=16, depth=16, signed=true

Is in fact:
  audio/x-raw,layout=interleaved,rate=8000,format=S16LE,channels=2

depth,width,signed,endianess have no meaning in 1.0 and may lead to
negotiation error. The in-gstreamer echo canceller works by placing
webrtcprobe close to your load speaker playback element, and placing
webrtcdsp filter close to your recording element.

Note that you'll always get better result due to timing precision with
using AEC in Pulse Audio rather then in Gst.

>
> But right now I am doing push to talk to avoid echo and noise on the hardware side.
>
> Please tell me how to implement AEC in GStreamer.....
>
>
> Thanks,
> Shrikant
> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[shrikantneotech]

I am getting confused where to keep element "webrtcprobe" & "webrtcechoprobe" . i tried but not getting result.

Can you please tell me exact location in my commands:

Recording & udpsinkcommand:

"gst-launch-1.0 -v alsasrc ! audioresample ! audioconvert ! audio/x-raw, format=S16LE, layout=interleaved, rate=8000, channels=2 !  udpsink host=192.168.1.139  port=5001"

 Receive & playback command:
"gst-launch-1.0 -v --gst-debug-level=4  udpsrc port=5003 !  audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2 ! audioconvert ! autoaudiosink"

Thanks,

Shrikant

On Thu, Sep 14, 2017 at 4:28 PM, Shrikant Lahase <[hidden email]> wrote:

Is GStreamer  1.12 version is necessary for echo cancellation?

On Thu, Sep 14, 2017 at 9:49 AM, Shrikant Lahase <[hidden email]> wrote:

Thanks for your reply.

For using "webrtcdsp" & "webrtcprobe" which plugin needs to install?

There is one more option I read on google "webrtcechoprobe" 

What is difference between  "webrtcprobe" & "webrtcechoprobe" 

What is attached signature.asc file?

On Wed, Sep 13, 2017 at 11:27 PM, Nicolas Dufresne <[hidden email]> wrote:

Le mercredi 13 septembre 2017 à 22:36 +0530, Shrikant Lahase a écrit :
> Hi

Welcome to the mailings list.

>
> I need to implement echo canceller in my audio application.
> I am sending audio alsasrc to remote mobile android app from my hardware.
> The command is as follows:
>
> Sending and receive combined in single command:
>
> "gst-launch-1.0 -v alsasrc ! audioresample ! audioconvert ! audio/x-
> raw, format=S16LE, layout=interleaved, rate=8000, channels=2,
> endianness=4321, width=16, depth=16, signed=true ! udpsink host=$1
> port=5001 & gst-launch-1.0 -v --gst-debug-level=4  udpsrc port=5003 !
> audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2,
> endianness=4321, width=16, depth=16, signed=true ! audioconvert !
> autoaudiosink"
>

This was very recently asked on the list,

http://gstreamer-devel.966125.n4.nabble.com/Webrtc-plugin-for-AEC-support-td4684580.html

In this thread, I give a example pipeline for testing in loopback the
AEC. Note that you are using 0.10 syntax in your pipeline.

  audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2,endianness=4321, width=16, depth=16, signed=true

Is in fact:
  audio/x-raw,layout=interleaved,rate=8000,format=S16LE,channels=2

depth,width,signed,endianess have no meaning in 1.0 and may lead to
negotiation error. The in-gstreamer echo canceller works by placing
webrtcprobe close to your load speaker playback element, and placing
webrtcdsp filter close to your recording element.

Note that you'll always get better result due to timing precision with
using AEC in Pulse Audio rather then in Gst.

>
> But right now I am doing push to talk to avoid echo and noise on the hardware side.
>
> Please tell me how to implement AEC in GStreamer.....
>
>
> Thanks,
> Shrikant
> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[Nicolas Dufresne-5]

Le jeudi 14 septembre 2017 à 16:34 +0530, Shrikant Lahase a écrit :

> I am getting confused where to keep element "webrtcprobe" &
> "webrtcechoprobe" . i tried but not getting result.
> Can you please tell me exact location in my commands:
>
> Recording & udpsinkcommand:
> "gst-launch-1.0 -v alsasrc ! audioresample ! audioconvert ! audio/x-
> raw, format=S16LE, layout=interleaved, rate=8000, channels=2 !
> udpsink host=192.168.1.139  port=5001"
>
>  Receive & playback command:
> "gst-launch-1.0 -v --gst-debug-level=4  udpsrc port=5003 !  audio/x-
> raw, layout=interleaved, rate=8000, format=S16LE, channels=2 !
> audioconvert ! autoaudiosink"

GStreamer AEC only works if the probe is in the same process and same
pipeline (sharing the same clock). Also, AEC rely on having proper
timestamp, which udpsrc alone won't give you. I suggest using RTP for
keeping track of time information. As mention in the other thread, a
corrected pipeline could be:

gst-launch-1.0 \
  alsasrc ! audioresample ! audioconvert ! audio/x-raw,rate=16000,channels=1 \
    ! webrtcdsp ! audioconvert ! rtpL16pay \
    ! udpsink host=127.0.0.1 port=5006 async=FALSE \
  udpsrc port=5006 caps="application/x-rtp,channels=1,clock-rate=16000" \
    ! rtpjitterbuffer latency=10 ! rtpL16depay ! audioconvert \
    ! webrtcechoprobe ! audioconvert ! audioresample ! alsasink

regards,
Nicolas

>
>
> Thanks,
> Shrikant
>
> On Thu, Sep 14, 2017 at 4:28 PM, Shrikant Lahase <shrikant@neotechind
> ia.com> wrote:
> > Is GStreamer  1.12 version is necessary for echo cancellation?
> >
> > On Thu, Sep 14, 2017 at 9:49 AM, Shrikant Lahase <shrikant@neotechi
> > ndia.com> wrote:
> > > Thanks for your reply.
> > >
> > > For using "webrtcdsp" & "webrtcprobe" which plugin needs to
> > > install?
> > > There is one more option I read on google "webrtcechoprobe"  
> > > What is difference between  "webrtcprobe" & "webrtcechoprobe"
> > >
> > > What is attached signature.asc file?
> > >
> > >
> > > On Wed, Sep 13, 2017 at 11:27 PM, Nicolas Dufresne <nicolas@ndufr
> > > esne.ca> wrote:
> > > > Le mercredi 13 septembre 2017 à 22:36 +0530, Shrikant Lahase a
> > > > écrit :
> > > > > Hi
> > > >
> > > > Welcome to the mailings list.
> > > >
> > > > >
> > > > > I need to implement echo canceller in my audio application.
> > > > > I am sending audio alsasrc to remote mobile android app from
> > > > my hardware.
> > > > > The command is as follows:
> > > > >
> > > > > Sending and receive combined in single command:
> > > > >
> > > > > "gst-launch-1.0 -v alsasrc ! audioresample ! audioconvert !
> > > > audio/x-
> > > > > raw, format=S16LE, layout=interleaved, rate=8000, channels=2,
> > > > > endianness=4321, width=16, depth=16, signed=true ! udpsink
> > > > host=$1
> > > > > port=5001 & gst-launch-1.0 -v --gst-debug-level=4  udpsrc
> > > > port=5003 !
> > > > > audio/x-raw, layout=interleaved, rate=8000, format=S16LE,
> > > > channels=2,
> > > > > endianness=4321, width=16, depth=16, signed=true !
> > > > audioconvert !
> > > > > autoaudiosink"
> > > > >
> > > >
> > > > This was very recently asked on the list,
> > > >
> > > > http://gstreamer-devel.966125.n4.nabble.com/Webrtc-plugin-for-A
> > > > EC-support-td4684580.html
> > > >
> > > > In this thread, I give a example pipeline for testing in
> > > > loopback the
> > > > AEC. Note that you are using 0.10 syntax in your pipeline.
> > > >
> > > >   audio/x-raw, layout=interleaved, rate=8000, format=S16LE,
> > > > channels=2,endianness=4321, width=16, depth=16, signed=true
> > > >
> > > > Is in fact:
> > > >   audio/x-
> > > > raw,layout=interleaved,rate=8000,format=S16LE,channels=2
> > > >
> > > > depth,width,signed,endianess have no meaning in 1.0 and may
> > > > lead to
> > > > negotiation error. The in-gstreamer echo canceller works by
> > > > placing
> > > > webrtcprobe close to your load speaker playback element, and
> > > > placing
> > > > webrtcdsp filter close to your recording element.
> > > >
> > > > Note that you'll always get better result due to timing
> > > > precision with
> > > > using AEC in Pulse Audio rather then in Gst.
> > > >
> > > > >
> > > > > But right now I am doing push to talk to avoid echo and noise
> > > > on the hardware side.
> > > > >
> > > > > Please tell me how to implement AEC in GStreamer.....
> > > > >
> > > > >
> > > > > Thanks,
> > > > > Shrikant
> > > > > _______________________________________________
> > > > > gstreamer-devel mailing list
> > > > > [hidden email]
> > > > > https://lists.freedesktop.org/mailman/listinfo/gstreamer-deve
> > > > l
>
>

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

[valer]

Hi Shrikant,

If the problem was not solved yet, you can always used a centralized AEC/NR like PBXMate.

---

From: Shrikant Lahase <[hidden email]>
To: [hidden email]
Sent: Wednesday, September 13, 2017 8:17 PM
Subject: Gstreamer echo canceller udpsrc and udpsink

Hi

I need to implement echo canceller in my audio application.

I am sending audio alsasrc to remote mobile android app from my hardware.

The command is as follows:

Sending and receive combined in single command:

"gst-launch-1.0 -v alsasrc ! audioresample ! audioconvert ! audio/x-raw, format=S16LE, layout=interleaved, rate=8000, channels=2, endianness=4321, width=16, depth=16, signed=true ! udpsink host=$1 port=5001 & gst-launch-1.0 -v --gst-debug-level=4 udpsrc port=5003 ! audio/x-raw, layout=interleaved, rate=8000, format=S16LE, channels=2, endianness=4321, width=16, depth=16, signed=true ! audioconvert ! autoaudiosink"

But right now I am doing push to talk to avoid echo and noise on the hardware side.

Please tell me how to implement AEC in GStreamer.....

Thanks,

Shrikant

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

_______________________________________________
gstreamer-devel mailing list
[hidden email]
https://lists.freedesktop.org/mailman/listinfo/gstreamer-devel