summary | shortlog | log | commit | commitdiff | tree
raw | patch | inline | side by side (parent: 8fcb607)
raw | patch | inline | side by side (parent: 8fcb607)
author | Mark Nauwelaerts <mark.nauwelaerts@collabora.co.uk> | |
Mon, 26 Sep 2011 10:07:15 +0000 (12:07 +0200) | ||
committer | Nikhil Devshatwar <a0132237@ti.com> | |
Fri, 17 May 2013 09:40:42 +0000 (15:10 +0530) |
... which simply comes down to requesting one frame of input data at a time,
since the encoder nicely turns this into 1 encoded frame.
since the encoder nicely turns this into 1 encoded frame.
ext/twolame/gsttwolame.c | patch | blob | history |
index 3006f1aa275ac8e23cfb2b4b35e3054f51e06c5e..bc2c07c5e12e2052ba0f0059890407fa3f407452 100644 (file)
--- a/ext/twolame/gsttwolame.c
+++ b/ext/twolame/gsttwolame.c
gst_pad_set_caps (GST_AUDIO_ENCODER_SRC_PAD (twolame), othercaps);
gst_caps_unref (othercaps);
- /* not much base class feedback:
- * - we will handle buffers, just hand us all available */
+ /* report needs to base class:
+ * hand one frame at a time, if we are pretty sure what a frame is */
+ if (out_samplerate == twolame->samplerate) {
+ gst_audio_encoder_set_frame_samples (enc, 1152);
+ gst_audio_encoder_set_frame_max (enc, 1);
+ }
return TRUE;