Support S24_P32/S32/FLOAT sample formats on Pulse

This is based on a patch from Ian Scott in 2014. It was never committed, so I figured I'd fix the outstanding issue and resubmit it. https://www.mail-archive.com/mpd-devel%40musicpd.org/msg00139.html

Support S24_P32/S32/FLOAT sample formats on Pulse
62000670 · Nils Schneider · Max Kellermann · ac49043f · 62000670 · 62000670
Commit 62000670 authored Aug 14, 2016 by Nils Schneider Committed by Max Kellermann Aug 15, 2016
Hide whitespace changes
Inline Side-by-side

Showing with 24 additions and 4 deletions

NEWS NEWS +2 -0

PulseOutputPlugin.cxx src/output/plugins/PulseOutputPlugin.cxx +22 -4

No files found.
--- a/NEWS
+++ b/NEWS
 ver 0.19.19 (not yet released)
+* output
+  - pulse: support 32 bit, 24 bit and floating point playback

 ver 0.19.18 (2016/08/05)
 * decoder

--- a/src/output/plugins/PulseOutputPlugin.cxx
+++ b/src/output/plugins/PulseOutputPlugin.cxx
@@ -574,12 +574,30 @@ pulse_output_open(AudioOutput *ao, AudioFormat &audio_format,
 		return false;
 	}

-	/* MPD doesn't support the other pulseaudio sample formats, so
-	   we just force MPD to send us everything as 16 bit */
-	audio_format.format = SampleFormat::S16;
+	/* Use the sample formats that our version of PulseAudio and MPD
+	   have in common, otherwise force MPD to send 16 bit */

 	pa_sample_spec ss;
-	ss.format = PA_SAMPLE_S16NE;
+
+	switch (audio_format.format) {
+	case SampleFormat::FLOAT:
+		ss.format = PA_SAMPLE_FLOAT32NE;
+		break;
+	case SampleFormat::S32:
+		ss.format = PA_SAMPLE_S32NE;
+		break;
+	case SampleFormat::S24_P32:
+		ss.format = PA_SAMPLE_S24_32NE;
+		break;
+	case SampleFormat::S16:
+		ss.format = PA_SAMPLE_S16NE;
+		break;
+	default:
+		audio_format.format = SampleFormat::S16;
+		ss.format = PA_SAMPLE_S16NE;
+		break;
+	}
+
 	ss.rate = audio_format.sample_rate;
 	ss.channels = audio_format.channels;