zqz/platform-external-webrtc
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

ACM: Removing runtime APIs related to playout mode

Browse Source 
The playout mode in NetEq can still be set through the constructor
configuration.

BUG=webrtc:3520

Review URL: https://codereview.webrtc.org/1362943004

Cr-Commit-Position: refs/heads/master@{#10089}
This commit is contained in:
henrik.lundin
2015-09-28 06:12:17 -07:00
committed by Commit bot
parent d417523194
commit 1bd0e03ce5
11 changed files with 13 additions and 258 deletions
Download Patch File Download Diff File Expand all files Collapse all files

37
webrtc/modules/audio_coding/main/interface/audio_coding_module.h
View File

@ -605,43 +605,6 @@ class AudioCodingModule {
// TODO(tlegrand): Change function to return the timestamp.
virtual int32_t PlayoutTimestamp(uint32_t* timestamp) = 0;
///////////////////////////////////////////////////////////////////////////
// int32_t SetPlayoutMode()
// Call this API to set the playout mode. Playout mode could be optimized
// for i) voice, ii) FAX or iii) streaming. In Voice mode, NetEQ is
// optimized to deliver highest audio quality while maintaining a minimum
// delay. In FAX mode, NetEQ is optimized to have few delay changes as
// possible and maintain a constant delay, perhaps large relative to voice
// mode, to avoid PLC. In streaming mode, we tolerate a little more delay
// to achieve better jitter robustness.
//
// Input:
// -mode : playout mode. Possible inputs are:
// "voice",
// "fax" and
// "streaming".
//
// Return value:
// -1 if failed to set the mode,
// 0 if succeeding.
//
virtual int32_t SetPlayoutMode(const AudioPlayoutMode mode) = 0;
///////////////////////////////////////////////////////////////////////////
// AudioPlayoutMode PlayoutMode()
// Get playout mode, i.e. whether it is speech, FAX or streaming. See
// audio_coding_module_typedefs.h for definition of AudioPlayoutMode.
//
// Return value:
// voice: is for voice output,
// fax: a mode that is optimized for receiving FAX signals.
// In this mode NetEq tries to maintain a constant high
// delay to avoid PLC if possible.
// streaming: a mode that is suitable for streaming. In this mode we
// accept longer delay to improve jitter robustness.
//
virtual AudioPlayoutMode PlayoutMode() const = 0;
///////////////////////////////////////////////////////////////////////////
// int32_t PlayoutData10Ms(
// Get 10 milliseconds of raw audio data for playout, at the given sampling

29
webrtc/modules/audio_coding/main/interface/audio_coding_module_typedefs.h
View File

@ -18,35 +18,6 @@
namespace webrtc {
///////////////////////////////////////////////////////////////////////////
// enum AudioPlayoutMode
// An enumerator for different playout modes.
//
// -voice : This is the standard mode for VoIP calls. The trade-off
// between low delay and jitter robustness is optimized
// for high-quality two-way communication.
// NetEQs packet loss concealment and signal processing
// capabilities are fully employed.
// -fax : The fax mode is optimized for decodability of fax signals
// rather than for perceived audio quality. When this mode
// is selected, NetEQ will do as few delay changes as possible,
// trying to maintain a high and constant delay. Meanwhile,
// the packet loss concealment efforts are reduced.
//
// -streaming : In the case of one-way communication such as passive
// conference participant, a webinar, or a streaming application,
// this mode can be used to improve the jitter robustness at
// the cost of increased delay.
// -off : Turns off most of NetEQ's features. Stuffs zeros for lost
// packets and during buffer increases.
//
enum AudioPlayoutMode {
voice = 0,
fax = 1,
streaming = 2,
off = 3,
};
///////////////////////////////////////////////////////////////////////////
// enum ACMSpeechType
// An enumerator for possible labels of a decoded frame.