--===============1171786648==
Content-Type: multipart/signed;
  boundary="nextPart1706242.ulFNaDuBgx";
  protocol="application/pgp-signature";
  micalg=pgp-sha1
Content-Transfer-Encoding: 7bit

--nextPart1706242.ulFNaDuBgx
Content-Type: text/plain;
  charset="iso-8859-15"
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline

On Wednesday 14 March 2007, Steve Thompson wrote:
> Boudewijn -
>
> On 3/14/07, Boudewijn Rempt <boud@valdyas.org> wrote:
> > Eek, yes... Could I trouble you for the original file & the .kra file in
> > lab format, so I can experiment and see what's going on? Nice photo, btw
> > :-)
>
> I've attached the original.  I didn't create a .kra for it, but I
> attached an image of the kind of LAB adjustment I applied to the a and
> b channels.  Let me know what you come up with, as working with LAB in
> Krita is such a pleasure.

I cannot recreate exactly your effect, but I noticed the following: if I=20
convert to rgb I lose some of the sparkle, too, but that depends on the rgb=
=20
profile I use. By default (stupid default, our bad), the Adobe RGB profile =
is=20
selected when you convert to rgb. Much better results can be had with the=20
lcms internal sRGB profile. (This is logical, since Adobe RGB is a pretty b=
ad=20
profile, for fun & kicks, create a white image and then switch to Adobe RGB=
=20
as your display profile.)


=2D-=20
Boudewijn Rempt=20
http://www.valdyas.org/fading/index.cgi

--nextPart1706242.ulFNaDuBgx
Content-Type: application/pgp-signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.3 (GNU/Linux)

iD8DBQBF+FS6daCcgCmN5d8RAiH5AKDVeYUg4oo344KeJBJnwg6BLv5jNQCeN0kp
QWvG+xMi2arPTvjvOQyr3UU=
=CCAR
-----END PGP SIGNATURE-----

--nextPart1706242.ulFNaDuBgx--


--===============1171786648==
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

____________________________________
koffice mailing list
koffice@kde.org
To unsubscribe please visit:
https://mail.kde.org/mailman/listinfo/koffice

--===============1171786648==--