Commit graph

55 commits

Author SHA1 Message Date
Jan Schmidt
3b03db5e40 deinterlace: Don't pointlessly hold object lock over caps operations
Avoids a deadlock when getcaps is recursive due to the getcaps being
reflected upstream/downstream. The lock isn't actually protecting
anything here.
2011-10-28 00:41:45 +11:00
René Stadler
dcd493279d deinterlace: remove avoidable call to gst_object_set_name 2011-10-21 22:32:38 +02:00
David Schleef
0446787e65 deinterlace: change field handling through methods
This likely breaks stuff.  The good: all of the methods now create
field images aligned with input frames, without timestamp mangling.
The bad: this touches a lot of code, much of which is hairy and in
need of cleanup.  However, at this point we can reasonably create a
PSNR-based test.
2011-08-21 15:15:14 -07:00
Mart Raudsepp
62cd1215c7 deinterlace: Fix Since tags for fieldanalysis related new properties
commit c1b100cf9c is after 0.10.29 and 0.10.30 was a branched release.
So fix Since tags from 0.10.29 to 0.10.31 for the new properties.
2011-08-02 23:38:13 +01:00
Miguel Angel Cabrera Moya
16d71c7d96 deinterlace: fix parameter type in trace
https://bugzilla.gnome.org/show_bug.cgi?id=650937
2011-05-24 09:45:31 +02:00
Robert Swain
c1b100cf9c deinterlace: Add support for deinterlacing using buffer caps/flags
When not using the fieldanalysis element immediately upstream of deinterlace,
behaviour should remain unchanged. fieldanalysis will set the caps and flags on
the buffers such that they can be interpreted and acted upon to produce
progressive output.

There are two main modes of operation:

- Passive pattern locking
  Passive pattern locking is a non-blocking, low-latency mode of operation that
  is suitable for close-to-live usage. Initially a telecine stream will be
  output as variable framerate with naïve timestamp adjustment. With each
  incoming buffer, an attempt is made to lock onto a pattern. When a lock is
  obtained, the src pad and output buffer caps will reflect the pattern and
  timestamps will be accurately interpolated between pattern repeats. This
  means that initially and at pattern transitions there will be short periods
  of inaccurate timestamping.

- Active pattern locking
  Active pattern locking is a blocking, high-latency mode of operation that is
  targeted at use-cases where timestamp accuracy is paramount. Buffers will be
  queued until enough are present to make a lock. When locked, timestamps will
  be accurately interpolated between pattern repeats. Orphan fields can be
  dropped or deinterlaced. If no lock can be obtained, a single field might be
  pushed through to be deinterlaced.

Locking can also be disabled or 'auto' chooses between passive and active
locking modes depending on whether upstream is live.
2011-05-19 05:45:57 +02:00
Sebastian Dröge
88cbcf5aa4 deinterlace: Add support for NV21 colorspace 2011-02-24 14:08:25 +01:00
Carsten Kroll
ce0e34a0d5 deinterlace: Add support for NV12 colorspace
Fixes bug #642961.
2011-02-24 14:00:37 +01:00
Robert Swain
6607cdcc08 deinterlace: Simplify setcaps
The current code never uses upstream negotiation so the code can be
significantly simplified.
2011-02-23 13:09:19 +01:00
Robert Swain
6b26017ae7 deinterlace: small clean-ups
Improve debug output by printing the buffer pointer when
popping a buffer and simplify code to use scanlines.bottom_field
as appropriate.

https://bugzilla.gnome.org/show_bug.cgi?id=642691
2011-02-21 20:14:19 +00:00
Robert Swain
32f757786b deinterlace: fix assigned method_id when using fallback
https://bugzilla.gnome.org/show_bug.cgi?id=642691
2011-02-21 20:14:09 +00:00
Thiago Santos
cb527623a3 deinterlace: Handle image caps without asserting
Images might have framerate=0/1 in the caps, which caused an
assertion on deinterlace. I don't know of interlaced image formats
but deinterlace might be hardcoded on some generic pipelines and
it shouldn't assert.

The fix was to set field_duration to 0 if the input has a framerate
with a 0 numerator.

This patch also adds checks for this situation on the unit tests.

https://bugzilla.gnome.org/show_bug.cgi?id=641400
2011-02-04 09:38:35 -03:00
Tim-Philipp Müller
ede0635d45 deinterlace: simplify template caps
We can merge all the YUV variants into one single structure.
2011-01-31 13:44:45 +00:00
David Schleef
97d74f293c deinterlace: Change the default to linear
The previous default, greedyh, takes 4 times as long as MPEG-2
video decoding, and is unlikely fast enough on any current CPU
to play 1080i video in real-time.  greedyl isn't much faster.
linear was chosen over vfir, since the quality advantage of vfir
is minimal compared to the occasional visual artifacts and slower
processing.
2011-01-05 17:50:09 -08:00
Sebastian Dröge
76ad0ee09b deinterlace: Change classification to Filter/Effect/Video/Deinterlace 2010-12-16 10:04:19 +01:00
Jan Schmidt
9941d3f0a2 deinterlace: Avoid infinite loop draining frames
When the pipeline is flushed just as we're draining history,
don't loop infinitely, just discard the history and abort.
2010-12-12 10:15:13 +11:00
David Schleef
b27d93a84a deinterlace: analyse RFF fields in correct order
Code was repeating the second field, not the first.
Fixes: #636179.
2010-11-30 17:28:00 -08:00
Tim-Philipp Müller
baacc9952a deinterlace: fix reference leak 2010-11-14 00:18:16 +00:00
Jan Schmidt
89a24bd36f deinterlace: Flush QoS and history before applying segment
When handling newsegment, flush out the buffer history in the
existing segment, not the new one. Fixes playback in some DVD
cases.

Partially fixes #633294
2010-11-12 13:21:09 +00:00
Tim-Philipp Müller
5adbc1e4c0 deinterlace: improve event logging 2010-11-12 12:20:16 +00:00
Robert Swain
9be159b32c deinterlace: Implement field history flushing
In a number of cases it is necessary to flush the field history by
performing 'degraded' deinterlacing - that is, using the user-chosen
method for as many fields as possible, then using vfir for as long as
there are >= 2 fields remaining in the history, then using linear for
the last field.

This should avoid losing fields being kept for history for example at
EOS.

This may address part of #633294
2010-11-12 10:40:51 +00:00
Robert Swain
5a56274cba deinterlace: Refactor chain function
This is needed to be able to output a frame from outside the chain
function, i.e. in the following commit that adds flushing of the field
history.
2010-11-12 10:40:48 +00:00
Robert Swain
6a6f90e745 deinterlace: Fix required fields logic
Both history_count and fields_required count from 1. As per the while loop
condition that follows this code, to perform the deinterlacing method, we need
history_count >= fields_required fields in the history. Therefore if we have
history_count < fields_required (not fields_required + 1), we need more fields.
2010-10-06 15:05:36 +02:00
David Schleef
738e092d4e deinterlace: Document methods with bad quality 2010-09-05 22:23:58 -07:00
David Schleef
4eee45debb deinterlace: use separate buffer metadata for fields
Call gst_buffer_make_metadata_writable() on buffers that are
duplicated into fields.  Fixes #627689.
2010-08-23 02:51:24 -07:00
Sebastian Dröge
a11b271bf4 deinterlace: Call orc_init() before trying to get target flags 2010-06-14 15:38:02 +02:00
David Schleef
f14972bf20 deinterlace: convert from liboil to orc 2010-06-08 17:25:34 -07:00
Sebastian Dröge
d09ff4124e deinterlace: Don't reconfigure the caps when changing properties
Fixes bug #619848.
2010-06-01 11:21:30 +02:00
Sebastian Dröge
a9ed56b1ad deinterlace: Add support for UYVY 2010-05-07 20:41:31 +02:00
Sebastian Dröge
9d6e4a7ac8 deinterlace: Add support for all common RGB formats 2010-05-05 17:39:32 +02:00
Sebastian Dröge
848f071ef4 deinterlace: Add support for AYUV 2010-05-05 16:06:51 +02:00
Sebastian Dröge
2ac1f1c7ee deinterlace: Make automatic detection of interlacing the default
Previously "force deinterlacing" was the default, which is a not very
sensible default for the normal use case where deinterlace should act
in passthrough mode unless interlaced content is present.
2010-04-30 22:17:12 +02:00
Sebastian Dröge
61217b521c deinterlace: Improve segment handling a bit 2010-04-29 19:28:24 +02:00
Sebastian Dröge
05a2732851 deinterlace: Order caps by amount of contained information 2010-04-29 19:28:24 +02:00
Sebastian Dröge
cb789617f9 deinterlace: Properly set interlaced field in getcaps 2010-04-29 19:28:24 +02:00
Sebastian Dröge
03a8379e20 deinterlace: Add support for Y444, Y42B, I420, YV12 and Y41B
The vfir method supports them and will be used until something else
supports it.
2010-04-29 19:28:23 +02:00
Sebastian Dröge
600f82fbfe deinterlace: Move deinterlacing methods to their own file 2010-04-29 19:28:23 +02:00
Sebastian Dröge
a405d5a4f1 deinterlace: Simplify passthrough mode detection 2010-04-29 19:28:23 +02:00
Sebastian Dröge
3dc7215492 deinterlace: Refactor deinterlacing as preparation for supporting more color formats 2010-04-29 19:28:23 +02:00
David Schleef
289f69eb84 deinterlace: Only check interlaced flag in sink caps
Fixes #615460.
2010-04-11 13:15:32 -07:00
Stefan Kost
e7a5ff40bd docs: trim sections file more
Rename some defines and move some itesm to *.c files. Add more items to internal
subsection.
2010-04-08 12:14:07 +03:00
Tim-Philipp Müller
5908c40405 deinterlace: fix typo in mode enum description 2009-11-20 10:30:00 +00:00
Jan Schmidt
baa79ffecb Use new still-frame API from gst-plugins-base 2009-11-18 03:09:06 +00:00
Tim-Philipp Müller
65a1db99eb deinterlace: remove pointless call to gst_element_no_more_pads() 2009-11-02 08:45:53 +00:00
Sebastian Dröge
810c60a6f3 deinterlace: Remove the interlaced field from the output caps if deinterlacing is enabled 2009-06-23 17:52:29 +02:00
Sebastian Dröge
ceea8dc15f deinterlace: Implement upstream negotation 2009-06-11 10:06:04 +02:00
Sebastian Dröge
cd33adbd96 deinterlace: Improve debugging and clean up some code 2009-06-11 10:06:04 +02:00
Sebastian Dröge
101123a75d deinterlace: Clip buffers to the current segment if possible 2009-06-11 10:06:04 +02:00
Sebastian Dröge
4710a91cbf deinterlace: Clean up includes and clean up order of instance struct fields 2009-06-11 10:06:03 +02:00
Sebastian Dröge
f297e37b5e deinterlace: Fix QoS calculations
The diff is a signed integer, not an unsigned one of course.

In modes other than GST_DEINTERLACE_ALL every frame has twice the
duration of the field duration.
2009-06-10 09:57:31 +02:00