Linux Media Controller development
 help / color / mirror / Atom feed
From: Kamil Debski <k.debski@samsung.com>
To: 'Kiran Avnd' <kiran@chromium.org>
Cc: 'Kiran AVND' <avnd.kiran@samsung.com>,
	linux-media@vger.kernel.org, wuchengli@chromium.org,
	posciak@chromium.org, 'Arun Mankuzhi' <arun.m@samsung.com>,
	ihf@chromium.org, 'Prathyush' <prathyush.k@samsung.com>,
	'Arun Kumar' <arun.kk@samsung.com>
Subject: RE: [PATCH 03/17] [media] s5p-mfc: set B-frames as 2 while encoding
Date: Tue, 16 Sep 2014 16:20:17 +0200	[thread overview]
Message-ID: <02ba01cfd1b9$57162de0$054289a0$%debski@samsung.com> (raw)
In-Reply-To: <CAFHkUjjjK4TOSfYj=xS12LPcUAusKrMaDG1WYiVegG+B9K6w3A@mail.gmail.com>

Hi Kiran,

> From: Kiran Avnd [mailto:kiran@chromium.org]
> Sent: Tuesday, September 16, 2014 10:10 AM
> 
> Hi Kamil,
> 
> 
> On Mon, Sep 15, 2014 at 8:12 PM, Kamil Debski <k.debski@samsung.com>
> wrote:
> 
> 
> 	Hi,
> 
> 	> From: Kiran AVND [mailto:avnd.kiran@samsung.com]
> 	> Sent: Monday, September 15, 2014 8:43 AM
> 	>
> 	> set default number of B-frames as 2 while encoding for better
> 	> compression. This User can however change this setting using
> 	> V4L2_CID_MPEG_VIDEO_B_FRAMES.
> 
> 	The last sentence should be rephrased, as it is not clear.
> 
> 
> 
> Sure. Will do.
> 
> 
> 	The tougher question is what should be the default?
> 	In my opinion the default should produce a stream that is as
> 	simple as it gets. It should be playable on most hardware,
> 	some of which may not support B frames.
> 
> 	Anyway - this setting can be changed by the user to 2 using
> 	V4L2_CID_MPEG_VIDEO_B_FRAMES ;)
> 
> 
> 
> 
> Its empirical, and this setting has better compression quality and bit
> rate than otherwise, in our testing.

It doesn't surprise me - introducing B frames should increase both quality
and compression efficiency at the cost of computation complexity and
limiting compatibility.

The thing I am against is changing the driver so that it forces such
value of the encoding parameter on the user. Default number of B frames
should remain 0. The application should consciously choose to use B frames.

Best wishes,
--
Kamil Debski
Samsung R&D Institute Poland


> 
> Regards,
> Kiran
> 
> 
> 
> 
> 	>
> 	> Signed-off-by: Kiran AVND <avnd.kiran@samsung.com>
> 	> ---
> 	>  drivers/media/platform/s5p-mfc/s5p_mfc_enc.c |    2 +-
> 	>  1 files changed, 1 insertions(+), 1 deletions(-)
> 	>
> 	> diff --git a/drivers/media/platform/s5p-mfc/s5p_mfc_enc.c
> 	> b/drivers/media/platform/s5p-mfc/s5p_mfc_enc.c
> 	> index cd1b2a2..f7a6f68 100644
> 	> --- a/drivers/media/platform/s5p-mfc/s5p_mfc_enc.c
> 	> +++ b/drivers/media/platform/s5p-mfc/s5p_mfc_enc.c
> 	> @@ -280,7 +280,7 @@ static struct mfc_control controls[] = {
> 	>               .minimum = 0,
> 	>               .maximum = 2,
> 	>               .step = 1,
> 	> -             .default_value = 0,
> 	> +             .default_value = 2,
> 	>       },
> 	>       {
> 	>               .id = V4L2_CID_MPEG_VIDEO_H264_PROFILE,
> 	> --
> 	> 1.7.3.rc2
> 
> 	--
> 	To unsubscribe from this list: send the line "unsubscribe linux-
> media" in
> 	the body of a message to majordomo@vger.kernel.org
> 	More majordomo info at  http://vger.kernel.org/majordomo-
> info.html
> 
> 



  parent reply	other threads:[~2014-09-16 14:20 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-09-15  6:42 [PATCH 00/17] Fixes from Chrome OS tree for MFC driver Kiran AVND
2014-09-15  6:42 ` [PATCH 01/17] [media] s5p-mfc: support MIN_BUFFERS query for encoder Kiran AVND
2014-09-15  6:42 ` [PATCH 02/17] [media] s5p-mfc: Fix REQBUFS(0) " Kiran AVND
2014-09-15  6:42 ` [PATCH 03/17] [media] s5p-mfc: set B-frames as 2 while encoding Kiran AVND
2014-09-15 14:42   ` Kamil Debski
     [not found]     ` <CAFHkUjjjK4TOSfYj=xS12LPcUAusKrMaDG1WYiVegG+B9K6w3A@mail.gmail.com>
2014-09-16 14:20       ` Kamil Debski [this message]
2014-09-15  6:42 ` [PATCH 04/17] [media] s5p-mfc: clear 'enter_suspend' flag if suspend fails Kiran AVND
2014-09-15  6:43 ` [PATCH 05/17] [media] s5p-mfc: don't disable clock when next ctx is pending Kiran AVND
2014-09-16 14:20   ` Kamil Debski
2014-09-17  9:25     ` Kiran Avnd
2014-09-17  9:57       ` Kamil Debski
2014-09-17 10:55         ` Kiran Avnd
2014-09-15  6:43 ` [PATCH 06/17] [media] s5p-mfc: Only set timestamp/timecode for new frames Kiran AVND
2014-09-15  6:43 ` [PATCH 07/17] [media] s5p-mfc: keep RISC ON during reset for V7/V8 Kiran AVND
2014-09-15  6:43 ` [PATCH 08/17] [media] s5p-mfc: check mfc bus ctrl before reset Kiran AVND
2014-09-15  6:43 ` [PATCH 09/17] [media] s5p-mfc: Don't crash the kernel if the watchdog kicks in Kiran AVND
2014-09-15  6:43 ` [PATCH 10/17] [media] s5p-mfc: modify mfc wakeup sequence for V8 Kiran AVND
2014-09-15 14:42   ` Kamil Debski
2014-09-15  6:43 ` [PATCH 11/17] [media] s5p-mfc: De-init MFC when watchdog kicks in Kiran AVND
2014-09-15  6:43 ` [PATCH 12/17] [media] s5p-mfc: flush dpbs when resolution changes Kiran AVND
2014-09-15  6:43 ` [PATCH 13/17] [media] s5p-mfc: Remove unused alloc field from private buffer struct Kiran AVND
2014-09-15  6:43 ` [PATCH 14/17] [media] s5p-mfc: fix V4L2_CID_MIN_BUFFERS_FOR_CAPTURE on resolution change Kiran AVND
2014-09-15  6:43 ` [PATCH 15/17] [media] s5p-mfc: remove reduntant clock on & clock off Kiran AVND
2014-09-15 14:42   ` Kamil Debski
2014-09-15  6:43 ` [PATCH 16/17] [media] s5p-mfc: fix a race in interrupt flags handling Kiran AVND
2014-09-15 14:42   ` Kamil Debski
2014-09-15  6:43 ` [PATCH 17/17] [media] s5p-mfc: Don't change the image size to smaller than the request Kiran AVND

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to='02ba01cfd1b9$57162de0$054289a0$%debski@samsung.com' \
    --to=k.debski@samsung.com \
    --cc=arun.kk@samsung.com \
    --cc=arun.m@samsung.com \
    --cc=avnd.kiran@samsung.com \
    --cc=ihf@chromium.org \
    --cc=kiran@chromium.org \
    --cc=linux-media@vger.kernel.org \
    --cc=posciak@chromium.org \
    --cc=prathyush.k@samsung.com \
    --cc=wuchengli@chromium.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox