如何在 gstreamer 中设置通道位置/通道掩码？

Question

我试图从 a 中获取 28 个通道jackaudiosrc，但是一旦我连接了一个deinterleave元素，源就会回退到无需任何通道掩码即可工作的 2 通道设置。我设置通道掩码的尝试是徒劳的。

发生的情况如下：

potential caps: audio/x-raw-float, endianness=(int)1234, width=(int)32, rate=(int)48000, channels=(int)[ 1, 28 ]
filter caps: audio/x-raw-float, channels=(int)28, channel-positions=(GstAudioChannelPosition)< GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE >
actual caps: audio/x-raw-float, endianness=(int)1234, width=(int)32, rate=(int)48000, channels=(int)2, channel-positions=(GstAudioChannelPosition)< GST_AUDIO_CHANNEL_POSITION_FRONT_LEFT, GST_AUDIO_CHANNEL_POSITION_FRONT_RIGHT >

下面是生成该输出的 Python 代码：

potential caps: audio/x-raw-float, endianness=(int)1234, width=(int)32, rate=(int)48000, channels=(int)[ 1, 28 ]
filter caps: audio/x-raw-float, channels=(int)28, channel-positions=(GstAudioChannelPosition)< GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE, GST_AUDIO_CHANNEL_POSITION_NONE >
actual caps: audio/x-raw-float, endianness=(int)1234, width=(int)32, rate=(int)48000, channels=(int)2, channel-positions=(GstAudioChannelPosition)< GST_AUDIO_CHANNEL_POSITION_FRONT_LEFT, GST_AUDIO_CHANNEL_POSITION_FRONT_RIGHT >

我还尝试直接设置输出的上限，而不是使用过滤器，但这也没有做任何事情。

这是最后一个 bin-to-dot-file 行的输出（单击放大）。

这是一条gst-launch在没有实际多通道音频设置的情况下工作的行，但实际上并没有重现错误，我相信是因为该interleave元素默认情况下正确设置了 NONE 上限，而 却jackaudiosrc没有。也许它仍然可以帮助那些由于技术限制而无法自己重现这个问题的人解决这个问题。感谢http://tristanswork.blogspot.de/2008/08/multichannel-audio-with-gstreamer.html我改编的原始启动线。

gst-launch-0.10 interleave name=i ! audioconvert ! audioresample ! queue ! deinterleave name=o audiotestsrc ! audioconvert ! queue ! i. audiotestsrc ! audioconvert ! queue ! i. audiotestsrc ! audioconvert ! queue ! i. o. ! queue ! fakesink dump=true o. ! queue ! fakesink dump=true o. ! queue ! fakesink dump=true

编辑：我现在尝试在 GStreamer 1.x 中做同样的事情。数据tuple类型不再被接受，因此这是我在查看其他人直接将通道位置写入函数后得出的结论get_caps_from_string。

from gi.repository import GObject, Gst, GstAudio
GObject.threads_init()
Gst.init(None)
pipeline = Gst.Pipeline("mypipeline")
src = Gst.ElementFactory.make("jackaudiosrc", "jacksrc")
pipeline.add(src)
srcpad = src.get_static_pad("src")
pipeline.set_state(Gst.State.PLAYING)
print 'potential caps:', srcpad.query_caps()
sink = Gst.ElementFactory.make("fakesink", "sink")
pipeline.add(sink)
filtercaps = Gst.caps_from_string("audio/x-raw, channels=(int)28, channel-positions=(int)< 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28 >")
print "filter caps:", filtercaps
src.link_filtered(sink, filtercaps)
pipeline.set_state(Gst.State.PLAYING)
print 'actual caps:', srcpad.get_current_caps()
Gst.debug_bin_to_dot_file_with_ts(pipeline, Gst.DebugGraphDetails.ALL, 'caps')

文本形式和图表的输出惊人地相似：

potential caps: audio/x-raw, format=(string)F32LE, layout=(string)interleaved, rate=(int)48000, channels=(int)[ 1, 28 ]
filter caps: audio/x-raw, channels=(int)28, channel-positions=(int)< 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28 >
actual caps: audio/x-raw, format=(string)F32LE, layout=(string)interleaved, rate=(int)48000, channels=(int)2