AGC2 as a new APM sub-module operating with hard-coded gain.

webrtc/modules/audio_processing/agc2/gain_controller2.h

+/*
+ *  Copyright (c) 2017 The WebRTC project authors. All Rights Reserved.
+ *
+ *  Use of this source code is governed by a BSD-style license
+ *  that can be found in the LICENSE file in the root of the source
+ *  tree. An additional intellectual property rights grant can be found
+ *  in the file PATENTS.  All contributing project authors may
+ *  be found in the AUTHORS file in the root of the source tree.
+ */
+
+#ifndef WEBRTC_MODULES_AUDIO_PROCESSING_AGC2_GAIN_CONTROLLER2_H_
+#define WEBRTC_MODULES_AUDIO_PROCESSING_AGC2_GAIN_CONTROLLER2_H_
+
+#include <memory>
+#include <string>
+
+#include "webrtc/base/constructormagic.h"
+#include "webrtc/modules/audio_processing/include/audio_processing.h"
+#include "webrtc/modules/audio_processing/level_controller/gain_applier.h"

[peah-webrtc, 2017/05/16 13:06:58]

I don't think we should reuse level_controller code like this, even if it is a
prototype. This is not part of the API surface of LevelController. If we would
want to make it general, I'd rather prefer to move this class out of the level
controller before starting to use it elsewhere.
I'm not saying it should be made general though. 

[AleBzk, 2017/05/18 08:31:37]

Thanks for this comment.
I wrote a new class named DigitalGainApplier in the AGC2 folder.
Since AGC2 has digital and analog gain, I think it's better to add the "Digital"
prefix.
The class is not as fancy as that in LC. When more features are needed, we will
add them.

[peah-webrtc, 2017/05/18 10:51:18]

That sounds good. 

+
+namespace webrtc {
+
+class ApmDataDumper;
+class AudioBuffer;
+
+// Gain Controller 2 aims to automatically adjusting levels by acting on the
+// microphone gain and/or applying digital gain. It does not use the band split

[peah-webrtc, 2017/05/16 13:06:57]

I don't think you should mention the band-split nor the issues with that in the
code. Instead write that in the CL description. Same thing applies to the
floating point interface.

[AleBzk, 2017/05/18 08:31:37]

Done.

+// filter, in order to avoid spectral leakage issues and unnecessary complexity.
+// Finally, it only provides the floating point interface.
+//
+// It temporarily implements a hard-coded gain mode only.
+class GainController2 {
+ public:
+  explicit GainController2(int sample_rate_hz);
+  ~GainController2();
+
+  void Process(AudioBuffer* audio);
+
+  static bool Validate(const AudioProcessing::Config::GainController2& config);
+  static std::string ToString(
+      const AudioProcessing::Config::GainController2& config);
+
+ private:
+  std::unique_ptr<ApmDataDumper> data_dumper_;
+  GainApplier gain_applier_;
+  static int instance_count_;
+
+  // TODO(...): Remove once a meaningful gain controller mode is implemented.

[peah-webrtc, 2017/05/16 13:06:58]

You need to have a name for the todo.

[AleBzk, 2017/05/18 08:31:37]

Done.

+  const float hard_coded_gain_;

[peah-webrtc, 2017/05/16 13:06:57]

fixed_gain_ sounds better. WDYT?

[AleBzk, 2017/05/18 08:31:37]

I started with that naming, but I didn't want the reader thinking that the fixed
digital gain is implemented (it's not). "Hard-coded" should convey more the
correct interpretation.

I also added _digital_ to further clarify.

Once this CL is lgtm'ed, the next will replace the hard-coded gain mode with a
first fixed digital gain implementation.

If you think I'm picky, no worries; |fixed_gain_| is also fine :)

[peah-webrtc, 2017/05/18 10:51:18]

Np :-) At this stage, anything is actually fine since this variable will anyway
be removed in upcoming CLs. So go for that.

+

[peah-webrtc, 2017/05/16 13:06:58]

Please remove duplicate empty line.

[AleBzk, 2017/05/18 08:31:37]

Done.

+
+  RTC_DISALLOW_COPY_AND_ASSIGN(GainController2);
+};
+
+}  // namespace webrtc
+
+#endif  // WEBRTC_MODULES_AUDIO_PROCESSING_AGC2_GAIN_CONTROLLER2_H_