mirrors/gstreamer
mirrors/gstreamer
1
1
Fork 0
mirror of https://gitlab.freedesktop.org/gstreamer/gstreamer.git synced 2024-11-27 12:11:13 +00:00
Code Issues 1 Projects Releases Wiki Activity

audioconvert: fix passthrough operation

Browse source 
We did not take the sample size into account. Rearrange the tests to have more
conversion test and an extra test case for passthrough operations.

Fixes #759890
This commit is contained in:
Stefan Sauer 2015-12-29 14:23:59 +01:00
parent f2ecf85103
commit 7bbfa39ada
2 changed files with 161 additions and 66 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
gst-libs/gst/audio/audio-converter.c
View file

@ -869,6 +869,7 @@ gst_audio_converter_samples (GstAudioConverter * convert,
in_samples = MIN (in_samples, out_samples);
if (in_samples == 0) {
GST_LOG ("skipping empty buffer");
*in_consumed = 0;
*out_produced = 0;
return TRUE;
@ -877,8 +878,13 @@ gst_audio_converter_samples (GstAudioConverter * convert,
chain = convert->pack_chain;
if (convert->passthrough) {
gsize bytes = in_samples * chain->inc *
(convert->in.bpf / convert->in.channels);
GST_LOG ("passthrough: %" G_GSIZE_FORMAT " / %" G_GSIZE_FORMAT " bytes",
in_samples, bytes);
for (i = 0; i < chain->blocks; i++)
memcpy (out[i], in[i], in_samples * chain->inc);
memcpy (out[i], in[i], bytes);
*out_produced = in_samples;
*in_consumed = in_samples;
return TRUE;

219
tests/check/elements/audioconvert.c
View file

@ -106,7 +106,7 @@ get_int_caps (guint channels, gint endianness, guint width,
fmt = gst_audio_format_build_integer (signedness, endianness, width, depth);
gst_audio_info_init (&info);
gst_audio_info_set_format (&info, fmt, 48000, channels, NULL);
gst_audio_info_set_format (&info, fmt, GST_AUDIO_DEF_RATE, channels, NULL);
caps = gst_audio_info_to_caps (&info);
fail_unless (caps != NULL);
@ -141,8 +141,8 @@ get_float_caps (guint channels, gint endianness, guint width)
g_assert (channels <= 2);
gst_audio_info_init (&info);
gst_audio_info_set_format (&info, get_float_format (endianness, width), 48000,
channels, NULL);
gst_audio_info_set_format (&info, get_float_format (endianness, width),
GST_AUDIO_DEF_RATE, channels, NULL);
caps = gst_audio_info_to_caps (&info);
fail_unless (caps != NULL);
@ -338,10 +338,10 @@ get_float_mc_caps (guint channels, gint endianness, guint width,
if (position) {
gst_audio_info_set_format (&info, get_float_format (endianness, width),
48000, channels, position);
GST_AUDIO_DEF_RATE, channels, position);
} else if (channels <= 6) {
gst_audio_info_set_format (&info, get_float_format (endianness, width),
48000, channels, channelpositions[channels - 1]);
GST_AUDIO_DEF_RATE, channels, channelpositions[channels - 1]);
} else {
GstAudioChannelPosition pos[64];
gint i;
@ -349,7 +349,7 @@ get_float_mc_caps (guint channels, gint endianness, guint width,
for (i = 0; i < 64; i++)
pos[i] = GST_AUDIO_CHANNEL_POSITION_NONE;
gst_audio_info_set_format (&info, get_float_format (endianness, width),
48000, channels, pos);
GST_AUDIO_DEF_RATE, channels, pos);
}
caps = gst_audio_info_to_caps (&info);
@ -373,9 +373,10 @@ get_int_mc_caps (guint channels, gint endianness, guint width,
gst_audio_info_init (&info);
if (position) {
gst_audio_info_set_format (&info, fmt, 48000, channels, position);
gst_audio_info_set_format (&info, fmt, GST_AUDIO_DEF_RATE, channels,
position);
} else if (channels <= 6) {
gst_audio_info_set_format (&info, fmt, 48000, channels,
gst_audio_info_set_format (&info, fmt, GST_AUDIO_DEF_RATE, channels,
channelpositions[channels - 1]);
} else {
GstAudioChannelPosition pos[64];
@ -383,7 +384,7 @@ get_int_mc_caps (guint channels, gint endianness, guint width,
for (i = 0; i < 64; i++)
pos[i] = GST_AUDIO_CHANNEL_POSITION_NONE;
gst_audio_info_set_format (&info, fmt, 48000, channels, pos);
gst_audio_info_set_format (&info, fmt, GST_AUDIO_DEF_RATE, channels, pos);
}
caps = gst_audio_info_to_caps (&info);
@ -398,7 +399,7 @@ get_int_mc_caps (guint channels, gint endianness, guint width,
static void
verify_convert (const gchar * which, void *in, int inlength,
GstCaps * incaps, void *out, int outlength, GstCaps * outcaps,
GstFlowReturn expected_flow)
GstFlowReturn expected_flow, gboolean in_place_allowed)
{
GstBuffer *inbuffer, *outbuffer;
GstElement *audioconvert;
@ -422,6 +423,13 @@ verify_convert (const gchar * which, void *in, int inlength,
gst_buffer_fill (inbuffer, 0, in, inlength);
ASSERT_BUFFER_REFCOUNT (inbuffer, "inbuffer", 1);
if (!in_place_allowed) {
/* take extra ref to force processing */
gst_buffer_ref (inbuffer);
gst_base_transform_set_passthrough (GST_BASE_TRANSFORM (audioconvert),
FALSE);
}
/* pushing gives away my reference ... */
GST_DEBUG ("push it");
fail_unless_equals_int (gst_pad_push (mysrcpad, inbuffer), expected_flow);
@ -434,6 +442,11 @@ verify_convert (const gchar * which, void *in, int inlength,
fail_unless (g_list_length (buffers) == 1);
fail_if ((outbuffer = (GstBuffer *) buffers->data) == NULL);
if (!in_place_allowed) {
/* release extra ref */
gst_buffer_unref (inbuffer);
}
ASSERT_BUFFER_REFCOUNT (outbuffer, "outbuffer", 1);
fail_unless_equals_int (gst_buffer_get_size (outbuffer), outlength);
@ -477,11 +490,18 @@ done:
#define RUN_CONVERSION(which, inarray, in_get_caps, outarray, out_get_caps) \
verify_convert (which, inarray, sizeof (inarray), \
in_get_caps, outarray, sizeof (outarray), out_get_caps, GST_FLOW_OK)
in_get_caps, outarray, sizeof (outarray), out_get_caps, GST_FLOW_OK, \
TRUE)
#define RUN_CONVERSION_TO_FAIL(which, inarray, in_caps, outarray, out_caps) \
verify_convert (which, inarray, sizeof (inarray), \
in_caps, outarray, sizeof (outarray), out_caps, GST_FLOW_NOT_NEGOTIATED)
in_caps, outarray, sizeof (outarray), out_caps, \
GST_FLOW_NOT_NEGOTIATED, TRUE)
#define RUN_CONVERSION_NOT_INPLACE(which, inarray, in_get_caps, outarray, out_get_caps) \
verify_convert (which, inarray, sizeof (inarray), \
in_get_caps, outarray, sizeof (outarray), out_get_caps, GST_FLOW_OK, \
FALSE)
GST_START_TEST (test_int16)
@ -671,6 +691,63 @@ GST_START_TEST (test_int_conversion)
GST_END_TEST;
GST_START_TEST (test_float_conversion)
{
/* 64-bit float <-> 32-bit float */
{
gdouble in[] = { 0.0, 1.0, -1.0, 0.5, -0.5 };
gfloat out[] = { 0.0, 1.0, -1.0, 0.5, -0.5 };
RUN_CONVERSION ("64 float to 32 float",
in, get_float_caps (1, G_BYTE_ORDER, 64),
out, get_float_caps (1, G_BYTE_ORDER, 32));
RUN_CONVERSION ("32 float to 64 float",
out, get_float_caps (1, G_BYTE_ORDER, 32),
in, get_float_caps (1, G_BYTE_ORDER, 64));
}
/* 32-bit float little endian <-> big endian */
{
gfloat le[] = { GFLOAT_TO_LE (0.0), GFLOAT_TO_LE (1.0), GFLOAT_TO_LE (-1.0),
GFLOAT_TO_LE (0.5), GFLOAT_TO_LE (-0.5)
};
gfloat be[] = { GFLOAT_TO_BE (0.0), GFLOAT_TO_BE (1.0), GFLOAT_TO_BE (-1.0),
GFLOAT_TO_BE (0.5), GFLOAT_TO_BE (-0.5)
};
RUN_CONVERSION ("32 float LE to BE",
le, get_float_caps (1, G_LITTLE_ENDIAN, 32),
be, get_float_caps (1, G_BIG_ENDIAN, 32));
RUN_CONVERSION ("32 float BE to LE",
be, get_float_caps (1, G_BIG_ENDIAN, 32),
le, get_float_caps (1, G_LITTLE_ENDIAN, 32));
}
/* 64-bit float little endian <-> big endian */
{
gdouble le[] =
{ GDOUBLE_TO_LE (0.0), GDOUBLE_TO_LE (1.0), GDOUBLE_TO_LE (-1.0),
GDOUBLE_TO_LE (0.5), GDOUBLE_TO_LE (-0.5)
};
gdouble be[] =
{ GDOUBLE_TO_BE (0.0), GDOUBLE_TO_BE (1.0), GDOUBLE_TO_BE (-1.0),
GDOUBLE_TO_BE (0.5), GDOUBLE_TO_BE (-0.5)
};
RUN_CONVERSION ("64 float LE to BE",
le, get_float_caps (1, G_LITTLE_ENDIAN, 64),
be, get_float_caps (1, G_BIG_ENDIAN, 64));
RUN_CONVERSION ("64 float BE to LE",
be, get_float_caps (1, G_BIG_ENDIAN, 64),
le, get_float_caps (1, G_LITTLE_ENDIAN, 64));
}
}
GST_END_TEST;
GST_START_TEST (test_int_float_conversion)
{
/* 32 float <-> 16 signed */
/* NOTE: if audioconvert was doing dithering we'd have a problem */
@ -756,63 +833,10 @@ GST_START_TEST (test_float_conversion)
in, get_int_caps (1, G_BYTE_ORDER, 32, 32, TRUE),
out, get_float_caps (1, G_BYTE_ORDER, 64));
}
/* 64-bit float <-> 32-bit float */
{
gdouble in[] = { 0.0, 1.0, -1.0, 0.5, -0.5 };
gfloat out[] = { 0.0, 1.0, -1.0, 0.5, -0.5 };
RUN_CONVERSION ("64 float to 32 float",
in, get_float_caps (1, G_BYTE_ORDER, 64),
out, get_float_caps (1, G_BYTE_ORDER, 32));
RUN_CONVERSION ("32 float to 64 float",
out, get_float_caps (1, G_BYTE_ORDER, 32),
in, get_float_caps (1, G_BYTE_ORDER, 64));
}
/* 32-bit float little endian <-> big endian */
{
gfloat le[] = { GFLOAT_TO_LE (0.0), GFLOAT_TO_LE (1.0), GFLOAT_TO_LE (-1.0),
GFLOAT_TO_LE (0.5), GFLOAT_TO_LE (-0.5)
};
gfloat be[] = { GFLOAT_TO_BE (0.0), GFLOAT_TO_BE (1.0), GFLOAT_TO_BE (-1.0),
GFLOAT_TO_BE (0.5), GFLOAT_TO_BE (-0.5)
};
RUN_CONVERSION ("32 float LE to BE",
le, get_float_caps (1, G_LITTLE_ENDIAN, 32),
be, get_float_caps (1, G_BIG_ENDIAN, 32));
RUN_CONVERSION ("32 float BE to LE",
be, get_float_caps (1, G_BIG_ENDIAN, 32),
le, get_float_caps (1, G_LITTLE_ENDIAN, 32));
}
/* 64-bit float little endian <-> big endian */
{
gdouble le[] =
{ GDOUBLE_TO_LE (0.0), GDOUBLE_TO_LE (1.0), GDOUBLE_TO_LE (-1.0),
GDOUBLE_TO_LE (0.5), GDOUBLE_TO_LE (-0.5)
};
gdouble be[] =
{ GDOUBLE_TO_BE (0.0), GDOUBLE_TO_BE (1.0), GDOUBLE_TO_BE (-1.0),
GDOUBLE_TO_BE (0.5), GDOUBLE_TO_BE (-0.5)
};
RUN_CONVERSION ("64 float LE to BE",
le, get_float_caps (1, G_LITTLE_ENDIAN, 64),
be, get_float_caps (1, G_BIG_ENDIAN, 64));
RUN_CONVERSION ("64 float BE to LE",
be, get_float_caps (1, G_BIG_ENDIAN, 64),
le, get_float_caps (1, G_LITTLE_ENDIAN, 64));
}
}
GST_END_TEST;
GST_START_TEST (test_multichannel_conversion)
{
{
@ -1010,6 +1034,68 @@ GST_START_TEST (test_multichannel_conversion)
GST_END_TEST;
GST_START_TEST (test_passthrough)
{
/* int 8 bit */
{
gint8 data[] = { 0, 1, 2, 127, -127 };
RUN_CONVERSION ("int 8 bit passthrough",
data, get_int_caps (1, G_BYTE_ORDER, 8, 8, TRUE),
data, get_int_caps (1, G_BYTE_ORDER, 8, 8, TRUE)
);
RUN_CONVERSION_NOT_INPLACE ("int 8 bit passthrough",
data, get_int_caps (1, G_BYTE_ORDER, 8, 8, TRUE),
data, get_int_caps (1, G_BYTE_ORDER, 8, 8, TRUE)
);
}
/* int 16 bit signed */
{
gint16 data[] = { 0, 256, 512, 32512, -32512 };
RUN_CONVERSION ("int 16 bit signed passthrough",
data, get_int_caps (1, G_BYTE_ORDER, 16, 16, TRUE),
data, get_int_caps (1, G_BYTE_ORDER, 16, 16, TRUE)
);
RUN_CONVERSION_NOT_INPLACE ("int 16 bit signed passthrough",
data, get_int_caps (1, G_BYTE_ORDER, 16, 16, TRUE),
data, get_int_caps (1, G_BYTE_ORDER, 16, 16, TRUE)
);
}
/* int 32 bit signed */
{
gint32 data[] = { 0, G_MININT32, G_MAXINT32,
(32 << 16), (32 << 16) + (1 << 15), (32 << 16) - (1 << 15),
(32 << 16) + (2 << 15), (32 << 16) - (2 << 15),
(-32 << 16) + (1 << 15), (-32 << 16) - (1 << 15),
(-32 << 16) + (2 << 15), (-32 << 16) - (2 << 15),
(-32 << 16)
};
RUN_CONVERSION ("int 32 bit signed passthrough",
data, get_int_caps (1, G_BYTE_ORDER, 32, 32, TRUE),
data, get_int_caps (1, G_BYTE_ORDER, 32, 32, TRUE)
);
RUN_CONVERSION_NOT_INPLACE ("int 32 bit signed passthrough",
data, get_int_caps (1, G_BYTE_ORDER, 32, 32, TRUE),
data, get_int_caps (1, G_BYTE_ORDER, 32, 32, TRUE)
);
}
/* int 16 bit signed stereo */
{
gint16 data[] = { 0, 0, 1, 1, 2, 2, 3, 3 };
RUN_CONVERSION ("int 16 bit signed 2 channel passthrough",
data, get_int_caps (2, G_BYTE_ORDER, 16, 16, TRUE),
data, get_int_caps (2, G_BYTE_ORDER, 16, 16, TRUE));
RUN_CONVERSION_NOT_INPLACE ("int 16 bit signed 2 channel passthrough",
data, get_int_caps (2, G_BYTE_ORDER, 16, 16, TRUE),
data, get_int_caps (2, G_BYTE_ORDER, 16, 16, TRUE));
}
}
GST_END_TEST;
GST_START_TEST (test_caps_negotiation)
{
GstElement *src, *ac1, *ac2, *ac3, *sink;
@ -1500,11 +1586,14 @@ audioconvert_suite (void)
tcase_add_test (tc_chain, test_float32);
tcase_add_test (tc_chain, test_int_conversion);
tcase_add_test (tc_chain, test_float_conversion);
tcase_add_test (tc_chain, test_int_float_conversion);
tcase_add_test (tc_chain, test_multichannel_conversion);
tcase_add_test (tc_chain, test_passthrough);
tcase_add_test (tc_chain, test_caps_negotiation);
tcase_add_test (tc_chain, test_convert_undefined_multichannel);
tcase_add_test (tc_chain, test_preserve_width);
return s;
}