(OBSOLETE) Increasing audio buffer on Iphone 4S.

Increasing audio buffer on Iphone 4S.

Iphone 4S is a slow. To allow complicated work in WebRTC audio processing, we need to increase the audio buffer.

BUG=

[minyue-webrtc, 2016-06-16 14:29:33]

Patchset #1 (id:1) has been deleted

[minyue-webrtc, 2016-06-16 17:38:49]

Description was changed from

==========
adding buffer

BUG=
==========

to

==========
Increasing audio buffer on Iphone 4S.

Iphone 4S is a slow. To allow complicated work in WebRTC audio processing, we
need to increase the audio buffer.

BUG=
==========

[minyue-webrtc, 2016-06-16 17:38:49]

minyue@webrtc.org changed reviewers:
+ henrika@webrtc.org

[minyue-webrtc, 2016-06-16 17:39:18]

Henrik,

Would you take a look at this?

[henrika_webrtc, 2016-06-17 07:37:37]

Thanks for looking into this. I am fine with the change but have some comments.

https://codereview.webrtc.org/2066993005/diff/20001/webrtc/modules/audio_devi...
File webrtc/modules/audio_device/ios/objc/RTCAudioSessionConfiguration.m
(right):

https://codereview.webrtc.org/2066993005/diff/20001/webrtc/modules/audio_devi...
webrtc/modules/audio_device/ios/objc/RTCAudioSessionConfiguration.m:50: // e.g.,
encoding long frames with Opus codec.
nit, I would say large frames

https://codereview.webrtc.org/2066993005/diff/20001/webrtc/modules/audio_devi...
webrtc/modules/audio_device/ios/objc/RTCAudioSessionConfiguration.m:91: struct
utsname systemInfo;
Please compare with existing methods in
https://cs.chromium.org/chromium/src/third_party/webrtc/modules/utility/sourc....
They are designed for mixed C++/ObjC but should be useful here as well.

[minyue-webrtc, 2016-06-23 10:57:29]

Patchset #2 (id:40001) has been deleted

[minyue-webrtc, 2016-06-23 11:06:03]

Patchset #2 (id:50001) has been deleted

[minyue-webrtc, 2016-06-23 11:35:07]

Hi Henrik,

I updated this CL. I have to rename .m to .mm to be able to use C code.

To make review easier, I use patch set 2 just for renaming, and patch set 3
contains updates.

https://codereview.webrtc.org/2066993005/diff/20001/webrtc/modules/audio_devi...
File webrtc/modules/audio_device/ios/objc/RTCAudioSessionConfiguration.m
(right):

https://codereview.webrtc.org/2066993005/diff/20001/webrtc/modules/audio_devi...
webrtc/modules/audio_device/ios/objc/RTCAudioSessionConfiguration.m:50: // e.g.,
encoding long frames with Opus codec.
On 2016/06/17 07:37:37, henrika_webrtc wrote:
> nit, I would say large frames

Done.

https://codereview.webrtc.org/2066993005/diff/20001/webrtc/modules/audio_devi...
webrtc/modules/audio_device/ios/objc/RTCAudioSessionConfiguration.m:91: struct
utsname systemInfo;
On 2016/06/17 07:37:37, henrika_webrtc wrote:
> Please compare with existing methods in
>
https://cs.chromium.org/chromium/src/third_party/webrtc/modules/utility/sourc....
> They are designed for mixed C++/ObjC but should be useful here as well.

Done.

[henrika_webrtc, 2016-06-23 11:49:13]

Sorry if my initial round of comments were not clear but I don't think we should
change from .m to .mm here.
I meant that the existing code could be used as base for doing the same thing in
pure ObjC but I must
refer to tkchin@ who knows ObjC (and mix of ObjC and C++) much better than I do.

Zeke, this is a trivial patch but I would appreciate some advice on how to
proceed to make it perfect.

[henrika_webrtc, 2016-06-23 11:49:37]

henrika@webrtc.org changed reviewers:
+ tkchin@webrtc.org

[tkchin_webrtc, 2016-06-24 00:30:48]

https://codereview.webrtc.org/2066993005/diff/90001/webrtc/modules/audio_devi...
File webrtc/modules/audio_device/ios/objc/RTCAudioSessionConfiguration.mm
(right):

https://codereview.webrtc.org/2066993005/diff/90001/webrtc/modules/audio_devi...
webrtc/modules/audio_device/ios/objc/RTCAudioSessionConfiguration.mm:16:
#include "webrtc/modules/utility/include/helpers_ios.h"
Let's avoid using this file. I want to move helpers away to sdk/objc.

https://codereview.webrtc.org/2066993005/diff/90001/webrtc/modules/audio_devi...
webrtc/modules/audio_device/ios/objc/RTCAudioSessionConfiguration.mm:92: if
(webrtc::ios::GetDeviceName() == "iPhone 4S") {
Wouldn't use strings like this. We should enumerate devices elsewhere.

I think the better things to do is to add a category to UIDevice, e.g.
UIDevice+WebRTC.h, define an enum for all the known device types, and then have
a method that does GetDeviceName() except it returns the type instead of the
string.

It should go into webrtc/sdk/objc/Framework/Headers/WebRTC/