From mboxrd@z Thu Jan  1 00:00:00 1970
MIME-Version: 1.0
In-Reply-To: <14ec7b180903051921u10524f16naad654b297a67de3@mail.gmail.com>
References: <14ec7b180903041807q58ae6a8fh31350709edd0c0c@mail.gmail.com>
	 <dc58bd25d2fbefabd5d0f4d965c72a8e@quanstro.net>
	 <14ec7b180903051921u10524f16naad654b297a67de3@mail.gmail.com>
Date: Fri,  6 Mar 2009 12:48:31 -0800
Message-ID: <dd6fe68a0903061248q70a02f56q7b350645fabed8e4@mail.gmail.com>
Subject: Re: [9fans] Porter-Duff alpha blending
From: Russ Cox <rsc@swtch.com>
To: Fans of the OS Plan 9 from Bell Labs <9fans@9fans.net>
Content-Type: multipart/mixed; boundary=000e0cd146be5a2c980464796839
Topicbox-Message-UUID: b5059d34-ead4-11e9-9d60-3106f5b1d025

--000e0cd146be5a2c980464796839
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: quoted-printable

Thanks for tracking this down and pointing out where
the error is.  There are actually a handful of errors
in that code.

Assume a source value sv, source alpha sa,
mask alpha ma, destination value dv, and
destination alpha da.  The values sv and dv
are stored premultiplied by their corresponding
alphas (the one true way).

Given those values, the correct new values
for the destination pixel in an S over D op are:

    dv =3D (sv*ma + dv*(255-sa*ma)) / 255
    da =3D (sa*ma + da*(255-sa*ma)) / 255

Bug #1: The current draw.c does the division
separately on the two halves:

    dv =3D (sv*ma)/255 + (dv*(255-sa*ma))/255

This can be off by one if the remainders from the
two divisions sum to >=3D 255.

Bug #2: The MUL0123 macro assumes four
values are packed into a 32-bit int and runs
them in simultaneous pairs as 32-bit operations
(MUL02 and MUL13) operating on 16-bit halves
of the word.  Those two don't use the right rounding
for the bitwise op implementation of /255.  On
a single value, the implementation is

    x / 255 =3D=3D (t =3D x+1, (t+(t>>8))>>8)
    (x+127) / 255 =3D=3D (t =3D x+128, (t+(t>>8))>>8)

These calculations only need 16 bits so you can
run two of them in the two halves of a 32-bit word.
The second implements round-to-nearest and
is what the draw code tries to do in this case.
But it only adds 128 (0x80), so it only rounds
the bottom half correctly.  It needs to add 0x00800080,
which would round both of them.  This explains:

src  rFF gFF bFF =CE=B1FF
mask  kFF =CE=B1FF
dst  r00 g00 b00 =CE=B1FF
dst after calc  rFE gFE bFF =CE=B1FF

Bug #3: MUL0123 is enabled whenever the src
and dst both have 32-bit pixel width, but there is
no check that the sub-channels are in the same
order.  You don't say what the image chans were
in your test, but this:

src  rFF g00 b00 =CE=B1FF
mask  kFF =CE=B1FF
dst  r00 g00 b00 =CE=B1FF
dst after calc  rFE gFE b00 =CE=B100

would be explained by, say, src=3D=3DARGB and
dst=3D=3DRGBA.  The A and R values in src became
the R and G chans in dst.  (In fact, since the dst
R and G are FE not FF, that's almost certainly
the scenario, modulo the little-endian draw names.)

This one is similarly explained:

src  rCC g00 b00 =CE=B1CC
mask  kFF =CE=B1FF
dst  r00 g00 b00 =CE=B1FF
dst after calc  rCB gCB b00 =CE=B133

The destination got scaled by 0x33/0xFF,
leaving 00 00 00 33, and then the source,
cc 00 00 cc, was added in the wrong place,
using the incorrect rounding, to produce
cb cb 00 33.

I have corrected these bugs in the plan9port
copy of src/libmemdraw/draw.c and finally
get the right answer for Andrey's test (attached).
I leave it as an exercise to the interested reader to
port the changes to the other dozen copies
of libmemdraw that are floating around,
or to unify them all.

Russ

--000e0cd146be5a2c980464796839
Content-Type: image/png; name="andrey.png"
Content-Disposition: attachment; filename="andrey.png"
Content-Transfer-Encoding: base64
X-Attachment-Id: f_frzc8vrx0
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=
--000e0cd146be5a2c980464796839--