Android: Fix chroma sampling bug when converting from texture to YUV

Android: Fix chroma sampling bug when converting from texture to YUV

The xUnit for the UV channels in SurfaceTextureHelper.YuvConverter is
currently calculated from 1 / (2 * width). It should be 1 / (width / 2)
instead.

R=nisse@webrtc.org

Committed: https://chromium.googlesource.com/external/webrtc/+/fae14804ee553cc6aa5de594eafe234f32e16bfd

[magjed_webrtc, 2016-04-06 10:40:29]

Description was changed from

==========
Android: Fix chroma sampling bug when converting from texture to YUV

The xUnit for the UV channels in SurfaceTextureHelper.YuvConverter is
currently calculated from 1 / (2 * width). It should be 1 / (width / 2)
instead.
==========

to

==========
Android: Fix chroma sampling bug when converting from texture to YUV

The xUnit for the UV channels in SurfaceTextureHelper.YuvConverter is
currently calculated from 1 / (2 * width). It should be 1 / (width / 2)
instead.
==========

[magjed_webrtc, 2016-04-06 10:40:30]

magjed@webrtc.org changed reviewers:
+ nisse@webrtc.org

[magjed_webrtc, 2016-04-06 10:40:42]

nisse - Please take a look.

[nisse-webrtc, 2016-04-06 10:55:39]

lgtm. Apply nit if you think it makes the code more readably.

https://codereview.webrtc.org/1862003002/diff/1/webrtc/api/java/android/org/w...
File webrtc/api/java/android/org/webrtc/SurfaceTextureHelper.java (right):

https://codereview.webrtc.org/1862003002/diff/1/webrtc/api/java/android/org/w...
webrtc/api/java/android/org/webrtc/SurfaceTextureHelper.java:256: 2 *
transformMatrix[0] / width,
Maybe write 2.0, to make it more obvious that the division is a floating point
op, not integer?

The previous code used 2.0f, matching the precision of the matrix elements. But
that was quite pointless. I don't remember the fine details in current C/C++,
but it seems likely that everything is promoted to double anyway (and then
truncating it to float for the function call, since I think glUniform2f really
takes floats, not doubles).

[magjed_webrtc, 2016-04-06 14:00:29]

https://codereview.webrtc.org/1862003002/diff/1/webrtc/api/java/android/org/w...
File webrtc/api/java/android/org/webrtc/SurfaceTextureHelper.java (right):

https://codereview.webrtc.org/1862003002/diff/1/webrtc/api/java/android/org/w...
webrtc/api/java/android/org/webrtc/SurfaceTextureHelper.java:256: 2 *
transformMatrix[0] / width,
On 2016/04/06 10:55:39, nisse-webrtc wrote:
> Maybe write 2.0, to make it more obvious that the division is a floating point
> op, not integer?
> 
> The previous code used 2.0f, matching the precision of the matrix elements.
But
> that was quite pointless. I don't remember the fine details in current C/C++,
> but it seems likely that everything is promoted to double anyway (and then
> truncating it to float for the function call, since I think glUniform2f really
> takes floats, not doubles).

Done, I changed to float literals. Previously it was promoted to float anyway
because |transformMatrix| is a float array.

[magjed_webrtc, 2016-04-06 14:00:33]

The CQ bit was checked by magjed@webrtc.org

[magjed_webrtc, 2016-04-06 14:00:33]

The patchset sent to the CQ was uploaded after l-g-t-m from nisse@webrtc.org
Link to the patchset: https://codereview.webrtc.org/1862003002/#ps20001 (title:
"Change 2 to 2.0f")

[commit-bot: I haz the power, 2016-04-06 14:00:36]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1862003002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1862003002/20001

[commit-bot: I haz the power, 2016-04-06 16:01:01]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-04-06 16:01:02]

Try jobs failed on following builders:
  win_baremetal on tryserver.webrtc (JOB_TIMED_OUT, no build URL)

[magjed_webrtc, 2016-04-07 07:26:57]

Description was changed from

==========
Android: Fix chroma sampling bug when converting from texture to YUV

The xUnit for the UV channels in SurfaceTextureHelper.YuvConverter is
currently calculated from 1 / (2 * width). It should be 1 / (width / 2)
instead.
==========

to

==========
Android: Fix chroma sampling bug when converting from texture to YUV

The xUnit for the UV channels in SurfaceTextureHelper.YuvConverter is
currently calculated from 1 / (2 * width). It should be 1 / (width / 2)
instead.

R=nisse@webrtc.org

Committed:
https://chromium.googlesource.com/external/webrtc/+/fae14804ee553cc6aa5de594e...
==========

[commit-bot: I haz the power, 2016-04-07 07:26:59]

Description was changed from

==========
Android: Fix chroma sampling bug when converting from texture to YUV

The xUnit for the UV channels in SurfaceTextureHelper.YuvConverter is
currently calculated from 1 / (2 * width). It should be 1 / (width / 2)
instead.

R=nisse@webrtc.org

Committed:
https://chromium.googlesource.com/external/webrtc/+/fae14804ee553cc6aa5de594e...
==========

to

==========
Android: Fix chroma sampling bug when converting from texture to YUV

The xUnit for the UV channels in SurfaceTextureHelper.YuvConverter is
currently calculated from 1 / (2 * width). It should be 1 / (width / 2)
instead.

R=nisse@webrtc.org

Committed: https://crrev.com/fae14804ee553cc6aa5de594eafe234f32e16bfd
Cr-Commit-Position: refs/heads/master@{#12274}
==========

[magjed_webrtc, 2016-04-07 07:27:00]

Committed patchset #2 (id:20001) manually as
fae14804ee553cc6aa5de594eafe234f32e16bfd (presubmit successful).

[commit-bot: I haz the power, 2016-04-07 07:27:03]

Patchset 2 (id:??) landed as
https://crrev.com/fae14804ee553cc6aa5de594eafe234f32e16bfd
Cr-Commit-Position: refs/heads/master@{#12274}