Re: [PATCH v2] media: v4l2-ctrls: add control for dpcm predictor

[Sakari Ailus]

Hi Hans and Prabhakar,

Hans Verkuil wrote:
> Hi Prabhakar!
>
> I've got some documentation review comments below...
>
> On Thu August 30 2012 00:58:16 Prabhakar Lad wrote:
> > From: Lad, Prabhakar <prabhakar.lad@xxxxxx>
> >
> > add V4L2_CID_DPCM_PREDICTOR control of type menu, which
> > determines the dpcm predictor. The predictor can be either
> > simple or advanced.
> >
> > Signed-off-by: Lad, Prabhakar <prabhakar.lad@xxxxxx>
> > Signed-off-by: Manjunath Hadli <manjunath.hadli@xxxxxx>
> > Cc: Sakari Ailus <sakari.ailus@xxxxxx>
> > Cc: Hans Verkuil <hans.verkuil@xxxxxxxxx>
> > Cc: Laurent Pinchart <laurent.pinchart@xxxxxxxxxxxxxxxx>
> > Cc: Mauro Carvalho Chehab <mchehab@xxxxxxxxxxxxx>
> > Cc: Sylwester Nawrocki <s.nawrocki@xxxxxxxxxxx>
> > Cc: Hans de Goede <hdegoede@xxxxxxxxxx>
> > Cc: Kyungmin Park <kyungmin.park@xxxxxxxxxxx>
> > ---
> > This patches has one checkpatch warning for line over
> > 80 characters altough it can be avoided I have kept it
> > for consistency.
> >
> > Changes for v2:
> > 1: Added documentaion in controls.xml pointed by Sylwester.
> > 2: Chnaged V4L2_DPCM_PREDICTOR_ADVANCE to V4L2_DPCM_PREDICTOR_ADVANCED
> >     pointed by Sakari.
> >
> >   Documentation/DocBook/media/v4l/controls.xml |   25 ++++++++++++++++++++++++-
> >   drivers/media/v4l2-core/v4l2-ctrls.c         |    9 +++++++++
> >   include/linux/videodev2.h                    |    5 +++++
> >   3 files changed, 38 insertions(+), 1 deletions(-)
> >
> > diff --git a/Documentation/DocBook/media/v4l/controls.xml b/Documentation/DocBook/media/v4l/controls.xml
> > index 93b9c68..84746d0 100644
> > --- a/Documentation/DocBook/media/v4l/controls.xml
> > +++ b/Documentation/DocBook/media/v4l/controls.xml
> > @@ -4267,7 +4267,30 @@ interface and may change in the future.</para>
> >   	    pixels / second.
> >   	    </entry>
> >   	  </row>
> > -	  <row><entry></entry></row>
> > +	  <row>
> > +	    <entry spanname="id"><constant>V4L2_CID_DPCM_PREDICTOR</constant></entry>
> > +	    <entry>menu</entry>
> > +	  </row>
> > +	  <row id="v4l2-dpcm-predictor">
> > +	    <entry spanname="descr"> DPCM Predictor: depicts what type of prediction
> > +	    is used simple or advanced.
>
> This is not useful information. It basically just rephrases the name of the
> define without actually telling me anything.
>
> I would expect to see here at least the following:
>
> - what the DPCM abbreviation stands for
> - a link or bibliography reference to the relevant standard (if there is any)

There's a Wikipedia article:

<URL:http://en.wikipedia.org/wiki/Differential_pulse-code_modulation>

It's the same DPCM encoding as in many of the compressed raw bayer formats.

> - a high-level explanation of what this do and what the difference is between
>   simple and advanced.

That's somewhat subjective and hardware dependent. If I understand this 
correctly, the "advanced" should differ only quality-wise from "simple" 
option. Why one would use "simple" instead of "advanced" then? Perhaps 
mostly for testing purposes; it might be that the advanced predictor 
could have issues in certain cases where the simple would not. However 
this isn't the only piece of hardware where I see that this is 
configurable, and the simple one was even the default.

> If this is part of a video compression standard, then this control would probably
> belong to the MPEG control class as well.

It's not --- DPCM compression is typically used on the bus between the 
sensor and the receiver to compress the data as the bus is often a 
limiting factor in the transfer rate. DPCM compression can be used to 
squeeze the data from 10 to 8 bits without much loss in quality or 
dynamic range.

Cheers,

--
Sakari Ailus
sakari.ailus@xxxxxx

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/