Add a 4k 60 fps vp9 smpte bars video to telemetry tests.

Add a 4k 60 fps vp9 smpte bars video to telemetry tests.

This video is 10 seconds of SMPTE color bars.
The load on the vp9 decoder is relatively low, in this way we can better
measure the load on the compositor.

The video has been generated using libavfilter.
BUG=

Committed: https://crrev.com/8c579956a575d458dd4fb6231f8f341be136c5c7
Cr-Commit-Position: refs/heads/master@{#329332}

[Daniele Castagna, 2015-04-14 22:34:50]

dcastagna@chromium.org changed reviewers:
+ prabhur@chromium.org

[prabhur1, 2015-04-14 22:56:06]

prabhur@chromium.org changed reviewers:
+ dalecurtis@chromium.org

[prabhur1, 2015-04-14 22:56:06]

dalecurtis@, could you take a look at the choice of the 4k60fps video being
added? Is there any specific requirements around how it needs to be encoded ?

[DaleCurtis, 2015-04-14 22:57:48]

dalecurtis@chromium.org changed reviewers:
+ fgalligan@chromium.org, renganathan@chromium.org

[DaleCurtis, 2015-04-14 22:57:49]

Can you post the encoding settings you used? +pangu, fgalligan for encoding
recommendations.

[Daniele Castagna, 2015-04-14 23:00:59]

On 2015/04/14 at 22:57:49, dalecurtis wrote:
> Can you post the encoding settings you used? +pangu, fgalligan for encoding
recommendations.

The video has been generated with:

ffmpeg -f lavfi -i smptebars=duration=10:size=4096x2160:rate=60 -vf
drawtext='fontsize=2000:fontfile=/usr/share/fonts/truetype/freefont/FreeSans.ttf:x=10:y=10:text=%{n}'
-c:v libvpx-vp9 -pass 2 -b:v 1000K -threads 8 -speed 1  -tile-columns 6
-frame-parallel 1 -auto-alt-ref 1 -lag-in-frames 25 -c:a libopus -b:a 64k -f
webm smpte_4k60_vp9.webm

the encoding settings are copied from here:
http://wiki.webmproject.org/ffmpeg/vp9-encoding-guide

[fgalligan1, 2015-04-15 01:12:01]

On 2015/04/14 23:00:59, Daniele Castagna wrote:
> On 2015/04/14 at 22:57:49, dalecurtis wrote:
> > Can you post the encoding settings you used? +pangu, fgalligan for encoding
> recommendations.
> 
> The video has been generated with:
> 
> ffmpeg -f lavfi -i smptebars=duration=10:size=4096x2160:rate=60 -vf
>
drawtext='fontsize=2000:fontfile=/usr/share/fonts/truetype/freefont/FreeSans.ttf:x=10:y=10:text=%{n}'
> -c:v libvpx-vp9 -pass 2 -b:v 1000K -threads 8 -speed 1  -tile-columns 6
> -frame-parallel 1 -auto-alt-ref 1 -lag-in-frames 25 -c:a libopus -b:a 64k -f
> webm smpte_4k60_vp9.webm
> 
> the encoding settings are copied from here:
> http://wiki.webmproject.org/ffmpeg/vp9-encoding-guide

Did you do a one pass first? I guess it doesn't matter, if you are satisfied
with the quality.

Should you change the name of the file to webm smpte_4096x2160_60fps_vp9.webm or
something? I just don't want to confuse someone later that might think the file
is a different resolution like 3840x2160.

Also why not at 3840x2160? Should we have both? I know displaying on TVs you
will need to scale the 4096x2160 to fit UHD, which is more overhead.

[Daniele Castagna, 2015-04-15 18:35:02]

On 2015/04/15 at 01:12:01, fgalligan wrote:
> On 2015/04/14 23:00:59, Daniele Castagna wrote:
> > On 2015/04/14 at 22:57:49, dalecurtis wrote:
> > > Can you post the encoding settings you used? +pangu, fgalligan for
encoding
> > recommendations.
> > 
> > The video has been generated with:
> > 
> > ffmpeg -f lavfi -i smptebars=duration=10:size=4096x2160:rate=60 -vf
> >
drawtext='fontsize=2000:fontfile=/usr/share/fonts/truetype/freefont/FreeSans.ttf:x=10:y=10:text=%{n}'
> > -c:v libvpx-vp9 -pass 2 -b:v 1000K -threads 8 -speed 1  -tile-columns 6
> > -frame-parallel 1 -auto-alt-ref 1 -lag-in-frames 25 -c:a libopus -b:a 64k -f
> > webm smpte_4k60_vp9.webm
> > 
> > the encoding settings are copied from here:
> > http://wiki.webmproject.org/ffmpeg/vp9-encoding-guide
> 
> Did you do a one pass first? I guess it doesn't matter, if you are satisfied
with the quality.
> 
Yes I did, just forgot to copy and paste.

> Should you change the name of the file to webm smpte_4096x2160_60fps_vp9.webm
or something? I just don't want to confuse someone later that might think the
file is a different resolution like 3840x2160.
> 
Sure.

> Also why not at 3840x2160? Should we have both? I know displaying on TVs you
will need to scale the 4096x2160 to fit UHD, which is more overhead.
Good point, the HD videos on youtube seems to be 3840x2160. Not sure if we
should keep 4096x2160 too since the difference with 3840x2160 is minimal.

[fgalligan1, 2015-04-15 23:57:35]

On 2015/04/15 18:35:02, Daniele Castagna wrote:
> On 2015/04/15 at 01:12:01, fgalligan wrote:
> > On 2015/04/14 23:00:59, Daniele Castagna wrote:
> > > On 2015/04/14 at 22:57:49, dalecurtis wrote:
> > > > Can you post the encoding settings you used? +pangu, fgalligan for
> encoding
> > > recommendations.
> > > 
> > > The video has been generated with:
> > > 
> > > ffmpeg -f lavfi -i smptebars=duration=10:size=4096x2160:rate=60 -vf
> > >
>
drawtext='fontsize=2000:fontfile=/usr/share/fonts/truetype/freefont/FreeSans.ttf:x=10:y=10:text=%{n}'
> > > -c:v libvpx-vp9 -pass 2 -b:v 1000K -threads 8 -speed 1  -tile-columns 6
> > > -frame-parallel 1 -auto-alt-ref 1 -lag-in-frames 25 -c:a libopus -b:a 64k
-f
> > > webm smpte_4k60_vp9.webm
> > > 
> > > the encoding settings are copied from here:
> > > http://wiki.webmproject.org/ffmpeg/vp9-encoding-guide
> > 
> > Did you do a one pass first? I guess it doesn't matter, if you are satisfied
> with the quality.
> > 
> Yes I did, just forgot to copy and paste.
> 
> > Should you change the name of the file to webm
smpte_4096x2160_60fps_vp9.webm
> or something? I just don't want to confuse someone later that might think the
> file is a different resolution like 3840x2160.
> > 
> Sure.
> 
> > Also why not at 3840x2160? Should we have both? I know displaying on TVs you
> will need to scale the 4096x2160 to fit UHD, which is more overhead.
> Good point, the HD videos on youtube seems to be 3840x2160. Not sure if we
> should keep 4096x2160 too since the difference with 3840x2160 is minimal.

The problem with 4096x2160 is that you will need to scale down as most 4k
devices are 3840x2160 (at least it was 6-12 months ago). As you know, just doing
copies at 4k can be very expensive. Doing a scale in software will be more
expensive than a copy. If all the scaling is done in hardware then it might not
matter.

I'm pretty sure Chrome will have to handle both resolutions. If we can only
accommodate one right now, then I would probably go with 3840x2160.

[Daniele Castagna, 2015-04-16 16:13:26]

On 2015/04/15 at 23:57:35, fgalligan wrote:
> On 2015/04/15 18:35:02, Daniele Castagna wrote:
> > On 2015/04/15 at 01:12:01, fgalligan wrote:
> > > On 2015/04/14 23:00:59, Daniele Castagna wrote:
> > > > On 2015/04/14 at 22:57:49, dalecurtis wrote:
> > > > > Can you post the encoding settings you used? +pangu, fgalligan for
> > encoding
> > > > recommendations.
> > > > 
> > > > The video has been generated with:
> > > > 
> > > > ffmpeg -f lavfi -i smptebars=duration=10:size=4096x2160:rate=60 -vf
> > > >
> >
drawtext='fontsize=2000:fontfile=/usr/share/fonts/truetype/freefont/FreeSans.ttf:x=10:y=10:text=%{n}'
> > > > -c:v libvpx-vp9 -pass 2 -b:v 1000K -threads 8 -speed 1  -tile-columns 6
> > > > -frame-parallel 1 -auto-alt-ref 1 -lag-in-frames 25 -c:a libopus -b:a
64k -f
> > > > webm smpte_4k60_vp9.webm
> > > > 
> > > > the encoding settings are copied from here:
> > > > http://wiki.webmproject.org/ffmpeg/vp9-encoding-guide
> > > 
> > > Did you do a one pass first? I guess it doesn't matter, if you are
satisfied
> > with the quality.
> > > 
> > Yes I did, just forgot to copy and paste.
> > 
> > > Should you change the name of the file to webm
smpte_4096x2160_60fps_vp9.webm
> > or something? I just don't want to confuse someone later that might think
the
> > file is a different resolution like 3840x2160.
> > > 
> > Sure.
> > 
> > > Also why not at 3840x2160? Should we have both? I know displaying on TVs
you
> > will need to scale the 4096x2160 to fit UHD, which is more overhead.
> > Good point, the HD videos on youtube seems to be 3840x2160. Not sure if we
> > should keep 4096x2160 too since the difference with 3840x2160 is minimal.
> 
> The problem with 4096x2160 is that you will need to scale down as most 4k
devices are 3840x2160 (at least it was 6-12 months ago). As you know, just doing
copies at 4k can be very expensive. Doing a scale in software will be more
expensive than a copy. If all the scaling is done in hardware then it might not
matter.
> 
> I'm pretty sure Chrome will have to handle both resolutions. If we can only
accommodate one right now, then I would probably go with 3840x2160.

Let's start with 3840x2160 then.

[fgalligan1, 2015-04-17 17:09:17]

On 2015/04/16 16:13:26, Daniele Castagna wrote:
> On 2015/04/15 at 23:57:35, fgalligan wrote:
> > On 2015/04/15 18:35:02, Daniele Castagna wrote:
> > > On 2015/04/15 at 01:12:01, fgalligan wrote:
> > > > On 2015/04/14 23:00:59, Daniele Castagna wrote:
> > > > > On 2015/04/14 at 22:57:49, dalecurtis wrote:
> > > > > > Can you post the encoding settings you used? +pangu, fgalligan for
> > > encoding
> > > > > recommendations.
> > > > > 
> > > > > The video has been generated with:
> > > > > 
> > > > > ffmpeg -f lavfi -i smptebars=duration=10:size=4096x2160:rate=60 -vf
> > > > >
> > >
>
drawtext='fontsize=2000:fontfile=/usr/share/fonts/truetype/freefont/FreeSans.ttf:x=10:y=10:text=%{n}'
> > > > > -c:v libvpx-vp9 -pass 2 -b:v 1000K -threads 8 -speed 1  -tile-columns
6
> > > > > -frame-parallel 1 -auto-alt-ref 1 -lag-in-frames 25 -c:a libopus -b:a
> 64k -f
> > > > > webm smpte_4k60_vp9.webm
> > > > > 
> > > > > the encoding settings are copied from here:
> > > > > http://wiki.webmproject.org/ffmpeg/vp9-encoding-guide
> > > > 
> > > > Did you do a one pass first? I guess it doesn't matter, if you are
> satisfied
> > > with the quality.
> > > > 
> > > Yes I did, just forgot to copy and paste.
> > > 
> > > > Should you change the name of the file to webm
> smpte_4096x2160_60fps_vp9.webm
> > > or something? I just don't want to confuse someone later that might think
> the
> > > file is a different resolution like 3840x2160.
> > > > 
> > > Sure.
> > > 
> > > > Also why not at 3840x2160? Should we have both? I know displaying on TVs
> you
> > > will need to scale the 4096x2160 to fit UHD, which is more overhead.
> > > Good point, the HD videos on youtube seems to be 3840x2160. Not sure if we
> > > should keep 4096x2160 too since the difference with 3840x2160 is minimal.
> > 
> > The problem with 4096x2160 is that you will need to scale down as most 4k
> devices are 3840x2160 (at least it was 6-12 months ago). As you know, just
doing
> copies at 4k can be very expensive. Doing a scale in software will be more
> expensive than a copy. If all the scaling is done in hardware then it might
not
> matter.
> > 
> > I'm pretty sure Chrome will have to handle both resolutions. If we can only
> accommodate one right now, then I would probably go with 3840x2160.
> 
> Let's start with 3840x2160 then.

LGTM

[DaleCurtis, 2015-04-17 17:10:54]

lgtm

[Daniele Castagna, 2015-04-17 17:25:15]

The CQ bit was checked by dcastagna@chromium.org

[commit-bot: I haz the power, 2015-04-17 17:26:02]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1083073003/100001

[commit-bot: I haz the power, 2015-04-17 17:38:01]

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

[commit-bot: I haz the power, 2015-04-17 17:38:02]

Try jobs failed on following builders:
  chromium_presubmit on tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromium_presub...)

[Daniele Castagna, 2015-04-17 17:56:14]

dcastagna@chromium.org changed reviewers:
+ sullivan@chromium.org

[Daniele Castagna, 2015-04-17 17:56:14]

+sullivan for owner's approval.

[sullivan, 2015-04-17 17:58:51]

lgtm

Please ensure this passes on perf try bots before submitting:
http://www.chromium.org/developers/performance-try-bots

[Daniele Castagna, 2015-05-10 20:50:53]

Patchset #7 (id:120001) has been deleted

[Daniele Castagna, 2015-05-12 01:49:16]

The CQ bit was checked by dcastagna@chromium.org

[Daniele Castagna, 2015-05-12 01:49:16]

The patchset sent to the CQ was uploaded after l-g-t-m from
sullivan@chromium.org, fgalligan@chromium.org, dalecurtis@chromium.org
Link to the patchset: https://codereview.chromium.org/1083073003/#ps140001
(title: "Default to 0 if metric is not present in dict.")

[commit-bot: I haz the power, 2015-05-12 01:50:43]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1083073003/140001

[commit-bot: I haz the power, 2015-05-12 02:56:33]

Committed patchset #7 (id:140001)

[commit-bot: I haz the power, 2015-05-12 02:57:21]

Patchset 7 (id:??) landed as
https://crrev.com/8c579956a575d458dd4fb6231f8f341be136c5c7
Cr-Commit-Position: refs/heads/master@{#329332}