Re: gst-plugins-bad: aacparse: allow parsed frames on sink pad

[Tim-Philipp Müller-2]

On Fri, 2011-02-25 at 20:02 -0800, David Schleef wrote:

> Module: gst-plugins-bad
> Branch: master
> Commit: e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> URL:    http://cgit.freedesktop.org/gstreamer/gst-plugins-bad/commit/?id=e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
>
> Author: David Schleef <[hidden email]>
> Date:   Wed Feb 23 17:25:03 2011 -0800
>
> aacparse: allow parsed frames on sink pad

Unfortunately we can't do that (yet) for parsers with a rank > NONE,
because it might lead to decodebin/decodebin2 autoplugging the same
parser over and over again in some cases..

 -Tim

> ---
>
>  gst/audioparsers/gstaacparse.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/gst/audioparsers/gstaacparse.c b/gst/audioparsers/gstaacparse.c
> index 09e3e71..fbdc165 100644
> --- a/gst/audioparsers/gstaacparse.c
> +++ b/gst/audioparsers/gstaacparse.c
> @@ -58,7 +58,7 @@ static GstStaticPadTemplate sink_template = GST_STATIC_PAD_TEMPLATE ("sink",
>      GST_PAD_SINK,
>      GST_PAD_ALWAYS,
>      GST_STATIC_CAPS ("audio/mpeg, "
> -        "framed = (boolean) false, " "mpegversion = (int) { 2, 4 };"));
> +        "framed = (boolean) { true, false}, " "mpegversion = (int) { 2, 4 };"));

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

[Stefan Sauer]

Am 26.02.2011 15:43, schrieb Tim-Philipp Müller:

> On Fri, 2011-02-25 at 20:02 -0800, David Schleef wrote:
>
>> Module: gst-plugins-bad
>> Branch: master
>> Commit: e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
>> URL:    http://cgit.freedesktop.org/gstreamer/gst-plugins-bad/commit/?id=e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
>>
>> Author: David Schleef <[hidden email]>
>> Date:   Wed Feb 23 17:25:03 2011 -0800
>>
>> aacparse: allow parsed frames on sink pad
>
> Unfortunately we can't do that (yet) for parsers with a rank > NONE,
> because it might lead to decodebin/decodebin2 autoplugging the same
> parser over and over again in some cases..

I did that mistake once localy and I was wondering if we should extend the logic
in decodebin to not replug an element in the same group?

Stefan

>
>  -Tim
>
>
>> ---
>>
>>  gst/audioparsers/gstaacparse.c |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/gst/audioparsers/gstaacparse.c b/gst/audioparsers/gstaacparse.c
>> index 09e3e71..fbdc165 100644
>> --- a/gst/audioparsers/gstaacparse.c
>> +++ b/gst/audioparsers/gstaacparse.c
>> @@ -58,7 +58,7 @@ static GstStaticPadTemplate sink_template = GST_STATIC_PAD_TEMPLATE ("sink",
>>      GST_PAD_SINK,
>>      GST_PAD_ALWAYS,
>>      GST_STATIC_CAPS ("audio/mpeg, "
>> -        "framed = (boolean) false, " "mpegversion = (int) { 2, 4 };"));
>> +        "framed = (boolean) { true, false}, " "mpegversion = (int) { 2, 4 };"));
>
>
> _______________________________________________
> gstreamer-devel mailing list
> [hidden email]
> http://lists.freedesktop.org/mailman/listinfo/gstreamer-devel

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

[Sebastian Dröge-7]

On Sat, 2011-02-26 at 17:05 +0200, Stefan Kost wrote:

> Am 26.02.2011 15:43, schrieb Tim-Philipp Müller:
> > On Fri, 2011-02-25 at 20:02 -0800, David Schleef wrote:
> >
> >> Module: gst-plugins-bad
> >> Branch: master
> >> Commit: e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> >> URL:    http://cgit.freedesktop.org/gstreamer/gst-plugins-bad/commit/?id=e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> >>
> >> Author: David Schleef <[hidden email]>
> >> Date:   Wed Feb 23 17:25:03 2011 -0800
> >>
> >> aacparse: allow parsed frames on sink pad
> >
> > Unfortunately we can't do that (yet) for parsers with a rank > NONE,
> > because it might lead to decodebin/decodebin2 autoplugging the same
> > parser over and over again in some cases..
>
> I did that mistake once localy and I was wondering if we should extend the logic
> in decodebin to not replug an element in the same group?

Yes, something like that should be done.

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

[David Schleef-2]

On Sat, Feb 26, 2011 at 01:43:09PM +0000, Tim-Philipp Müller wrote:

> On Fri, 2011-02-25 at 20:02 -0800, David Schleef wrote:
>
> > Module: gst-plugins-bad
> > Branch: master
> > Commit: e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> > URL:    http://cgit.freedesktop.org/gstreamer/gst-plugins-bad/commit/?id=e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> >
> > Author: David Schleef <[hidden email]>
> > Date:   Wed Feb 23 17:25:03 2011 -0800
> >
> > aacparse: allow parsed frames on sink pad
>
> Unfortunately we can't do that (yet) for parsers with a rank > NONE,
> because it might lead to decodebin/decodebin2 autoplugging the same
> parser over and over again in some cases..

Oh duh.  I knew that.



David

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

[David Schleef-2]

On Sat, Feb 26, 2011 at 06:55:48PM +0100, Sebastian Dröge wrote:

> On Sat, 2011-02-26 at 17:05 +0200, Stefan Kost wrote:
> > Am 26.02.2011 15:43, schrieb Tim-Philipp Müller:
> > > On Fri, 2011-02-25 at 20:02 -0800, David Schleef wrote:
> > >
> > >> Module: gst-plugins-bad
> > >> Branch: master
> > >> Commit: e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> > >> URL:    http://cgit.freedesktop.org/gstreamer/gst-plugins-bad/commit/?id=e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> > >>
> > >> Author: David Schleef <[hidden email]>
> > >> Date:   Wed Feb 23 17:25:03 2011 -0800
> > >>
> > >> aacparse: allow parsed frames on sink pad
> > >
> > > Unfortunately we can't do that (yet) for parsers with a rank > NONE,
> > > because it might lead to decodebin/decodebin2 autoplugging the same
> > > parser over and over again in some cases..
> >
> > I did that mistake once localy and I was wondering if we should extend the logic
> > in decodebin to not replug an element in the same group?
>
> Yes, something like that should be done.

Can this be done easily/quickly?  Otherwise, I'm likely to duplicate
the element with a zero rank, so it can be used for transcoding.



David

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

[Sebastian Dröge-7]

On Sat, 2011-02-26 at 17:00 -0500, David Schleef wrote:

> On Sat, Feb 26, 2011 at 06:55:48PM +0100, Sebastian Dröge wrote:
> > On Sat, 2011-02-26 at 17:05 +0200, Stefan Kost wrote:
> > > Am 26.02.2011 15:43, schrieb Tim-Philipp Müller:
> > > > On Fri, 2011-02-25 at 20:02 -0800, David Schleef wrote:
> > > >
> > > >> Module: gst-plugins-bad
> > > >> Branch: master
> > > >> Commit: e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> > > >> URL:    http://cgit.freedesktop.org/gstreamer/gst-plugins-bad/commit/?id=e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> > > >>
> > > >> Author: David Schleef <[hidden email]>
> > > >> Date:   Wed Feb 23 17:25:03 2011 -0800
> > > >>
> > > >> aacparse: allow parsed frames on sink pad
> > > >
> > > > Unfortunately we can't do that (yet) for parsers with a rank > NONE,
> > > > because it might lead to decodebin/decodebin2 autoplugging the same
> > > > parser over and over again in some cases..
> > >
> > > I did that mistake once localy and I was wondering if we should extend the logic
> > > in decodebin to not replug an element in the same group?
> >
> > Yes, something like that should be done.
>
> Can this be done easily/quickly?  Otherwise, I'm likely to duplicate
> the element with a zero rank, so it can be used for transcoding.

Shouldn't be that much work, all the information is inside decodebin2
already. You only need to look at all elements in the current
GstDecodeChain and if the element you're going to plug next has the same
type as one of the elements in there you don't use it.

I'll work on it now and will CC you in the bugreport with the patch :)

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

[Sebastian Dröge-7]

On Sat, 2011-02-26 at 23:05 +0100, Sebastian Dröge wrote:

> On Sat, 2011-02-26 at 17:00 -0500, David Schleef wrote:
> > On Sat, Feb 26, 2011 at 06:55:48PM +0100, Sebastian Dröge wrote:
> > > On Sat, 2011-02-26 at 17:05 +0200, Stefan Kost wrote:
> > > > Am 26.02.2011 15:43, schrieb Tim-Philipp Müller:
> > > > > On Fri, 2011-02-25 at 20:02 -0800, David Schleef wrote:
> > > > >
> > > > >> Module: gst-plugins-bad
> > > > >> Branch: master
> > > > >> Commit: e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> > > > >> URL:    http://cgit.freedesktop.org/gstreamer/gst-plugins-bad/commit/?id=e49b89d5c5a1244fa0dcb8bb4996e38fb9bff9e5
> > > > >>
> > > > >> Author: David Schleef <[hidden email]>
> > > > >> Date:   Wed Feb 23 17:25:03 2011 -0800
> > > > >>
> > > > >> aacparse: allow parsed frames on sink pad
> > > > >
> > > > > Unfortunately we can't do that (yet) for parsers with a rank > NONE,
> > > > > because it might lead to decodebin/decodebin2 autoplugging the same
> > > > > parser over and over again in some cases..
> > > >
> > > > I did that mistake once localy and I was wondering if we should extend the logic
> > > > in decodebin to not replug an element in the same group?
> > >
> > > Yes, something like that should be done.
> >
> > Can this be done easily/quickly?  Otherwise, I'm likely to duplicate
> > the element with a zero rank, so it can be used for transcoding.
>
> Shouldn't be that much work, all the information is inside decodebin2
> already. You only need to look at all elements in the current
> GstDecodeChain and if the element you're going to plug next has the same
> type as one of the elements in there you don't use it.
>
> I'll work on it now and will CC you in the bugreport with the patch :)

Oh well, I've simply pushed it now, it was even easier than expected:
5058f792263e4214f4655c5ed5956ebfbeb17c99

Feel free to change all parsers now to accept their own output again ;)

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

[Tim-Philipp Müller-2]

On Sat, 2011-02-26 at 23:42 +0100, Sebastian Dröge wrote:

> > Shouldn't be that much work, all the information is inside decodebin2
> > already. You only need to look at all elements in the current
> > GstDecodeChain and if the element you're going to plug next has the same
> > type as one of the elements in there you don't use it.
> >
> > I'll work on it now and will CC you in the bugreport with the patch :)
>
> Oh well, I've simply pushed it now, it was even easier than expected:
> 5058f792263e4214f4655c5ed5956ebfbeb17c99
>
> Feel free to change all parsers now to accept their own output again ;)

We probably need something a bit more sophisticated though, or limit
this to parser factories. It's perfectly fine (and common enough) to
e.g. have two id3demux in a row.

Cheers
 -Tim


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

[Sebastian Dröge-7]

On Sun, 2011-02-27 at 00:01 +0000, Tim-Philipp Müller wrote:

> On Sat, 2011-02-26 at 23:42 +0100, Sebastian Dröge wrote:
>
> > > Shouldn't be that much work, all the information is inside decodebin2
> > > already. You only need to look at all elements in the current
> > > GstDecodeChain and if the element you're going to plug next has the same
> > > type as one of the elements in there you don't use it.
> > >
> > > I'll work on it now and will CC you in the bugreport with the patch :)
> >
> > Oh well, I've simply pushed it now, it was even easier than expected:
> > 5058f792263e4214f4655c5ed5956ebfbeb17c99
> >
> > Feel free to change all parsers now to accept their own output again ;)
>
> We probably need something a bit more sophisticated though, or limit
> this to parser factories. It's perfectly fine (and common enough) to
> e.g. have two id3demux in a row.

Hmm, I can add a check for parsers there. Good point, I forgot the
demuxers that are no real demuxers and don't get their own
GstDecodeGroup.

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