W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > July to September 2002

RE: do vector graphics enhance our concept of self?

From: Danny Ayers <danny666@virgilio.it>
Date: Tue, 30 Jul 2002 00:01:10 +0200
To: "jonathan chetwynd" <j.chetwynd@btinternet.com>, "Charles McCathieNevile" <charles@w3.org>
Cc: "Jon Hanna" <jon@spin.ie>, "WAI List \(E-mail\)" <w3c-wai-ig@w3.org>
Message-ID: <EBEPLGMHCDOJJJPCFHEFKEHMHEAA.danny666@virgilio.it>

I've just been playing with Celinea, and haven't managed to achieve what
you're looking for. The most realistic I've got around 50k appear like
particularly bad 256-colour gifs. The big problem with the Celinea-produced
images is that the flesh tones change abruptly rather than being smooth
blends. It wasn't what I expected, but the more detailed bits appear a lot
more acceptable. As SVG supports blends, I'm certain it will be possible to
produce reasonable photo-realism in a small size, though a bit more number
crunching is needed.
I've not tried yet, but maybe doing some kind of half-toning/pixellating
first may be a workaround.

Cheers,
Danny.

---
Danny Ayers
<stuff> http://www.isacat.net </stuff>

Idea maps for the Semantic Web
http://www.isacat.net/ideagraph


>-----Original Message-----
>From: jonathan chetwynd [mailto:j.chetwynd@btinternet.com]
>Sent: 29 July 2002 22:18
>To: Charles McCathieNevile
>Cc: Danny Ayers; Jon Hanna; WAI List (E-mail)
>Subject: Re: do vector graphics enhance our concept of self?
>
>
>chaals,
>
>the problem staring us in the face, is that jpegs dont scale, like vectors.
>there are some vector faces here:
>http://www.eboy.com/pages/works/vectors/vectors_09.html
>they have a large file size(30K), and whilst probably recognisable are
>certainly not photographic.
>
>Can anyone point to realistic vector portraits, with a small file size?
>
>jonathan
>
>
>
>----- Original Message -----
>From: "Charles McCathieNevile" <charles@w3.org>
>To: "jonathan chetwynd" <j.chetwynd@btinternet.com>
>Cc: "Danny Ayers" <danny666@virgilio.it>; "Jon Hanna" <jon@spin.ie>; "WAI
>List (E-mail)" <w3c-wai-ig@w3.org>
>Sent: Monday, July 29, 2002 1:27 PM
>Subject: Re: do vector graphics enhance our concept of self?
>
>
>>
>> I think the situation is a bit more complex than you seem to suggest.
>>
>> There are technical reasons why jpeg doesn't have transparency, although
>it
>> does have good compression for photograph type images. GIF and
>PNG, on the
>> other hand, have transparency, but very poor compression for photographs
>(by
>> comparison). Essentially using these types of technology you have to make
>a
>> trade-off - you get one or the other.
>>
>> In some cases SVG should be used to replace PNG, JPEG or GIF images
>> (perhaps most cases where png or gif are a good choice SVG is a better
>> technical choice, and as user agents and authoring tools become more
>common
>> will be a better overall choice).
>>
>> In other cases SVG allows you to combine the best features of
>SVG with the
>> best features of JPEG (its compression for certain types of image) or
>PNG/GIF
>> (Hmmm. The more I think about it the more I wonder what you
>really gain by
>> keeping those formats. I am sure there are relevant cases though).
>>
>> I am not familiar with all the SVG authoring tools. However, to create a
>tool
>> which allows you to trace a section of a jpeg and say "show me just that
>> section" isn't hard. Jim Ley has such a tool available, in an online
>version
>> written in SVG. (He uses it for providing more useful searchable
>information
>> about the bit of the image that you trace, but it can also provide a
>clipped
>> region).
>>
>> This thing can be animated or scripted to move around, and again Jim has
>> demos. (Because these are things he works on for a bit of fun, the script
>> changes from time to time. But I believe that he is happy for people to
>use
>> it).
>>
>> It is possible to use this to create a complete user interface, but it
>takes
>> some programming time - generally available at the whim of a
>programmer or
>by
>> paying for it.
>>
>>
>> Technically, following the approach Jim has taken is the best
>method I can
>> think of. In terms of getting a tool that makes it easy, it also
>seems the
>> best approach I can think of. If you want something that a person with
>> relatively basic programming skills can work on, it is perhaps also the
>best
>> option. I you want me to program it for you then all you need to do is
>find
>> me the time ;-)
>>
>> Seriously, I appreciate that you want better tools. Me too. And world
>peace.
>> I think the best approach is to follow the mixed SVG/JPEG method,
>especially
>> if you want to do animation.
>>
>> Cheers
>>
>> Chaals
>>
>> On Sat, 27 Jul 2002, jonathan chetwynd wrote:
>>
>> >
>> >Yes Chaals,
>> >but the unfortunate fact is that with a gif one can just choose a 'magic
>> >wand tool' and the job is kind of done.
>> >whereas for svg, one remains in a kind of fantasy land, ie highly work
>> >intensive, and only partially meeting the needs*.
>> >
>> >Jon,
>> >as far as png files go, as far as my limited (ie one file) test showed
>there
>> >was a 5-10% compression saving on gif, hardly comparable with ~300% for
>> >jpeg.
>> >
>> >Surely it must be plain that with all the open source productivity
>available
>> >a concerted effort to either get the jpeg source released, reverse
>engineer,
>> >or create a new source is a reasonable project?
>> >SVG is not intended to replace jpeg or gif as I understand it, or did I
>get
>> >this wrong to?
>> >
>> >thanks again
>> >
>> >Jonathan
>> >
>> >On degredation, transparency is lost, so all the tracing goes to waste.
>> >Transparency is fairly fundamental to sprites.
>> >Its this degredation, that is the concern, or more significantly, the
>> >attitude to it, for meta-freaks.
>> >
>> >----- Original Message -----
>> >From: "Danny Ayers" <danny666@virgilio.it>
>> >To: "Charles McCathieNevile" <charles@w3.org>; "Jon Hanna" <jon@spin.ie>
>> >Cc: "WAI List (E-mail)" <w3c-wai-ig@w3.org>
>> >Sent: Friday, July 26, 2002 7:27 PM
>> >Subject: RE: do vector graphics enhance our concept of self?
>> >
>> >
>> >>
>> >> <clippath d="[[here you need a path around the bit of the image that
>> >> >you want - this can be generated quickly by tracing out the image,
>> >>
>> >> Amaya's good for that...
>> >>
>> >>
>> >
>> >
>>
>> --
>> Charles McCathieNevile    http://www.w3.org/People/Charles
>phone: +61 409
>134 136
>> W3C Web Accessibility Initiative     http://www.w3.org/WAI  fax: +33 4 92
>38 78 22
>> Location: 21 Mitchell street FOOTSCRAY Vic 3011, Australia
>> (or W3C INRIA, Route des Lucioles, BP 93, 06902 Sophia Antipolis Cedex,
>France)
>>
>
>
Received on Monday, 29 July 2002 18:09:22 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:14:05 GMT