Looks fantastic! Will have a play this weekend with them!

Sent from my iPhone

On 29 Jul 2011, at 14:25, Tavmjong Bah <tavmjong@...8...> wrote:

Hi,

We'll be discussing Advanced Gradients in today's SVG Working Group meeting. As an experiment, I've made a modified version of Inkscape that can read in the proposed syntax for Mesh Gradients and display the meshes. You can look at the results at:

http://tavmjong.free.fr/SVG/MESH/Mesh_Inkscape.html

                   Tav

---

Got Input? Slashdot Needs You. Take our quick survey online. Come on, we don't ask for help often. Plus, you'll get a chance to win $100 to spend on ThinkGeek. http://p.sf.net/sfu/slashdot-survey _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel

On Fri, 2011-07-29 at 21:38 +0200, Chris Lilley wrote:

On Friday, July 29, 2011, 3:25:52 PM, Tavmjong wrote:

TB> We'll be discussing Advanced Gradients in today's SVG Working Group TB> meeting.

It was a good discussion and your write up of your experiments was very helpful.

TB> As an experiment, I've made a modified version of Inkscape that TB> can read in the proposed syntax for Mesh Gradients and display the TB> meshes.

Is it possible to get hold of the modified source so that tI can experiment further?

TB> You can look at the results at:

TB> http://tavmjong.free.fr/SVG/MESH/Mesh_Inkscape.html

Yes, give me a day or two and I'll put in in a publicly available place.

Tav

On 2011-07-29 23:01, Joel Holdsworth wrote:

On 29/07/11 14:25, Tavmjong Bah wrote:

...

Hi,

We'll be discussing Advanced Gradients in today's SVG Working Group meeting. As an experiment, I've made a modified version of Inkscape that can read in the proposed syntax for Mesh Gradients and display the meshes. You can look at the results at:

http://tavmjong.free.fr/SVG/MESH/Mesh_Inkscape.html

				Tav

Have the merits of mesh gradients been discussed over the merits of diffusion curves?

http://wiki.inkscape.org/wiki/index.php/Diffusion_Curves http://artis.imag.fr/Publications/2008/OBWBTS08/

Which of the two concepts is more powerful?

Neither :) Both have their strengths and weaknesses. In general diffusion curves as currently defined have more of a tendency "to do the right thing" and are WAY easier to handle. On the other hand, mesh gradients (especially tensor product patches) give more control. Now, there are some ideas floating around to enhance diffusion curves to the point where there might be little benefit to using meshes, but even so, given practical considerations and the wide spread use of meshes it will no doubt be useful to support meshes as well.

Also, keep in mind that diffusion curves are relatively new and there are still some issues in rendering them (it should be possible, but so far no renderer has been made, that I'm aware of, that is both fast and accurate). Also, while they often "do the right thing" there are some definite issues with non-color boundaries (the current solution is definitely not optimal) and they can have the tendency to over-smooth certain images.

It would be very interesting to see what gradient meshes do with color opacity !

ivan

________________________________ De : Jon Cruz <jon@...18...> À : Jasper van de Gronde <th.v.d.gronde@...528...> Cc : inkscape-devel@lists.sourceforge.net Envoyé le : Dimanche 31 Juillet 2011 10h50 Objet : Re: [Inkscape-devel] Mesh Gradients

On Jul 31, 2011, at 1:33 AM, Jasper van de Gronde wrote:

Neither :) Both have their strengths and weaknesses. In general diffusion curves as currently defined have more of a tendency "to do the right thing" and are WAY easier to handle. On the other hand, mesh gradients (especially tensor product patches) give more control. Now, there are some ideas floating around to enhance diffusion curves to the point where there might be little benefit to using meshes, but even so, given practical considerations and the wide spread use of meshes it will no doubt be useful to support meshes as well.

Also, keep in mind that diffusion curves are relatively new and there are still some issues in rendering them (it should be possible, but so far no renderer has been made, that I'm aware of, that is both fast and accurate). Also, while they often "do the right thing" there are some definite issues with non-color boundaries (the current solution is definitely not optimal) and they can have the tendency to over-smooth certain images.

Also keep in mind visual issues.

In all the examples I have seen, diffusion curves have had a distinct, if sometimes subtle, unique look to them. Something akin to looking at a watercolor and being able to say "the artists used watercolors to do this painting".

I've not seen the same for general gradient meshes.

------------------------------------------------------------------------------ Got Input?  Slashdot Needs You. Take our quick survey online.  Come on, we don't ask for help often. Plus, you'll get a chance to win $100 to spend on ThinkGeek. http://p.sf.net/sfu/slashdot-survey _______________________________________________ Inkscape-devel mailing list Inkscape-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/inkscape-devel

On 31/07/11 21:26, Ivan Louette wrote:

It would be very interesting to see what gradient meshes do with color opacity !

ivan

Here are a couple of examples from cairo. The first example contains two patches in the mesh. The second contains one patch that folds over on itself.

http://people.freedesktop.org/~ajohnson/mesh/mesh-gradient.png

http://people.freedesktop.org/~ajohnson/mesh/mesh-gradient-overlap-transp.pn...