W3C home > Mailing lists > Public > www-svg@w3.org > August 2012

Re: SVG Compositing Details

From: Rik Cabanier <cabanier@gmail.com>
Date: Tue, 14 Aug 2012 10:57:19 -0700
Message-ID: <CAGN7qDCd52dePbTbEXYM142AiUiSFjeoqT1tzQCkyFjGTN2E-A@mail.gmail.com>
To: Bob Holmes <rangsynth@gmail.com>
Cc: Nikos Andronikos <Nikos.Andronikos@cisra.canon.com.au>, www-svg@w3.org
On Tue, Aug 14, 2012 at 5:22 AM, Bob Holmes <rangsynth@gmail.com> wrote:

> Suggestions
>
> A) In Section 1.2
> While this specification provides the author the choice of many Porter
> Duff compositing operators and many blending modes, the simple alpha
> compositing model forced a single Porter Duff compositing operator and
> a single blend mode.
>
> While this specification provides the author the choice of many Porter
> Duff compositing operators and many blending modes, the simple alpha
> compositing model forced a single Porter Duff compositing operator and
> a single blend mode, WHICH IS SPECIFICALLY SRCOVER COMPOSITE MODE AND
> NORMAL BLEND MODE RESPECTIVELY.
>
> More emphasis needs to be placed on this factor. Implementors with
> decent design should not be concerned with simple alpha compositing or
> the formula. Only the formula for SRCOVER and NORMAL blend mode
> surely?(Because is same thing Yes?)
>

Yes, it is the same.
the first section talks about the current state of the world (where there
is no blending or advanced compositing).
It feels a bit strange to jump ahead here...


>
>
> --------------------------------------------------------------------------------------------------
>
> B) Section 6.1.1
> This is the default attribute which specifies no blending. The
> blending formula simply select the source color.
>
> SHOULD IT NOT SAY "SIMPLY SELECTS THE SOURCE COLOR"?
>
>
> --------------------------------------------------------------------------------------------------
> C) 5.1.7. Destination In
>
> NEED TO INSERT A LINE BREAK BEFORE THE ao
> Fa = 0 Fb = s
>     co = b x Cb x s o = b x s
>
> SHOULD BE
> Fa = 0 Fb = s
>     co = b x Cb x s
>     o = b x s
>

Thanks!
I updated the spec.


>
> -------------------------------------------------------------------------
> D) Reference bitmaps
> Would be nice to have reference bitmaps.
>

A set of basic tests are submitted to the W3C but they are not reviewed yet.


>
> -------------------------------------------------------------------------
> E) Alpha Adjustment/Masking/Opacity
> It would be nice if there was a section or some more emphasis as a
> guide to implementors on the fact that any alpha adjustments can
> simply be made before the blending function. Drawing to a layer with
> some masks then works fine. Then drawing the layer to the background
> using "srcover" composite mode allows the opacity and any further
> masking to be applied just before the srcover and all works well.
>

that would be an optimization that might not be applicable to all rendering
engines.

Rik
Received on Tuesday, 14 August 2012 17:57:48 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:52 GMT