qtdemux: Provide a 2 frames lead-in for audio decoders

AAC and various other audio codecs need a couple frames of lead-in to
decode it properly. The parser elements like aacparse take care of it
via gst_base_parse_set_frame_rate, but when inside a container, the
demuxer is doing the seek segment handling and never gives lead-in
data downstream.
Handle this similar to going back to a keyframe with video, in the
same place. Without a lead-in, the start of the segment is silence,
when it shouldn't, which becomes especially evident in NLE use cases.
19 jobs for qtdemux-audio-leadin in 78 minutes and 32 seconds (queued for 2 seconds)
latest
Status Job ID Name Coverage
  Preparation
passed #340795
gst indent

00:00:32

passed #340794
manifest

00:00:38

 
  Build
passed #340796
build fedora x86_64

00:08:35

passed #340797
build nodebug fedora x86_64

00:08:28

passed #340939
build static fedora x86_64

00:11:32

passed #340799
build static nodebug fedora x86_64

00:13:44

failed #340798
build static fedora x86_64

00:11:06

 
  Test
passed #340800
check fedora

00:04:03

passed #340801
integration testsuites fedora

00:08:46

 
  Full Builds
passed #340803
build cerbero android universal

00:11:39

passed #340804
build cerbero cross win32

00:06:58

passed #340805
build cerbero cross win64

00:06:28

passed #340802
build cerbero fedora x86_64

00:07:27

passed #340806
gst-macos-10.14
build cerbero macos x86_64

00:13:10

 
  Full Tests
passed #340808
valgrind bad

00:04:34

passed #340809
valgrind ges

00:08:43

passed #340807
valgrind good

00:20:16

 
  Apps
passed #340810
android universal examples

00:05:53

 
  Deploy
passed #340811
documentation

00:04:28