matplotlib-devel Mailing List for matplotlib

On 2008年4月22日 15:41:26 -0400, Michael Droettboom wrote:
> My concern with subpixel rendering has always been that it's fine for
> interactive use, but for producing plots that may end up on someone
> else's screen or on a printer, it can actually make matters much worse.
Naturally.
As such, cairo lets you explicitly choose what you want to draw by
setting one of the following values in your font options:
 CAIRO_ANTIALIAS_NONE: Use a bilevel alpha mask.
 CAIRO_ANTIALIAS_GRAY: Perform single-color antialiasing (using
 shades of gray for black text on a white background, for example).
 CAIRO_ANTIALIAS_SUBPIXEL: Perform antialiasing by taking
 advantage of the order of subpixel elements on devices
 such as LCD panels
And most importantly, the default value is as follows:
 CAIRO_ANTIALIAS_DEFAULT: Use the default antialiasing for
 the subsystem and target device
So, (excepting bugs), you should really have to go out of your way to
get cairo to do the wrong thing here.
-Carl

My concern with subpixel rendering has always been that it's fine for 
interactive use, but for producing plots that may end up on someone 
else's screen or on a printer, it can actually make matters much worse.
Agg (well freetype really) could do subpixel font rendering as well, but 
I deliberately stayed away from it last time I looked at fonts. We 
could have some mechanism to determine whether the render is intended 
for a GUI or a file, but that would have to be dealt with carefully so 
as not to introduce more backend selection and context related problems.
Cheers,
Mike
Darren Dale wrote:
> On Tuesday 22 April 2008 02:54:09 pm Eric Firing wrote:
> 
>> In case any devels are interested in making more backends work on the
>> trunk:
>>
>> Running examples/backend_driver.py yields quite a few svg plots that are
>> incorrect, with things like colorbars missing their colors.
>>
>> The cairo backend does better, but also needs a little work. Cairo is
>> looking more interesting with the latest release; the notes say that eps
>> output is now supported.
>> 
>
> Also, sibpixel rendering of fonts is supported in the new Cairo. On an LCD 
> monitor, fonts look sharper with Cairo than they do with AGG.
>
> -------------------------------------------------------------------------
> This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
> Don't miss this year's exciting event. There's still time to save 100ドル. 
> Use priority code J8TL2D2. 
> http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
> 
-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

On Tuesday 22 April 2008 02:54:09 pm Eric Firing wrote:
> In case any devels are interested in making more backends work on the
> trunk:
>
> Running examples/backend_driver.py yields quite a few svg plots that are
> incorrect, with things like colorbars missing their colors.
>
> The cairo backend does better, but also needs a little work. Cairo is
> looking more interesting with the latest release; the notes say that eps
> output is now supported.
Also, sibpixel rendering of fonts is supported in the new Cairo. On an LCD 
monitor, fonts look sharper with Cairo than they do with AGG.

In case any devels are interested in making more backends work on the trunk:
Running examples/backend_driver.py yields quite a few svg plots that are 
incorrect, with things like colorbars missing their colors.
The cairo backend does better, but also needs a little work. Cairo is 
looking more interesting with the latest release; the notes say that eps 
output is now supported.
Eric

This is inherent; mpl 0.90.1 is permanently incompatible with numpy 1.1, 
short of each user making the change suggested below. The earliest mpl 
that should work with numpy 1.1 is 0.91.2.
The change in masked array module is a major reason why numpy is getting 
a version bump to 1.1.0 instead of 1.0.5.
Eric
Michael Droettboom wrote:
> I will forward it to the matplotlib-devel mailing list on your behalf.
> 
> Cheers,
> Mike
> 
> lorenzo bolla wrote:
>> Hello,
>>
>> the latest svn numpy version 1.1.0.dev5061 does not work with 
>> matplotlib 0.90.1 (version shipped with enthought distribution), 
>> unless a change in 
>> Python25/Lib/site-packages/matplotlib-0.90.1.0003-py2.5-win32.egg/matplotlib/numerix/ma/__init__.py 
>> is done:
>>
>> $ diff __init__.py.orig __init__.py
>> 12c12
>> < from numpy.core.ma <http://numpy.core.ma> import *
>> ---
>>> from numpy.ma <http://numpy.ma> import *
>> Maybe this should be forwarded to the pylab mailing list, but I'm not 
>> subscribed there...
>>
>> L.
>>
>> ------------------------------------------------------------------------
>>
>> _______________________________________________
>> Numpy-discussion mailing list
>> Num...@sc...
>> http://projects.scipy.org/mailman/listinfo/numpy-discussion
>> 
> 

I will forward it to the matplotlib-devel mailing list on your behalf.
Cheers,
Mike
lorenzo bolla wrote:
> Hello,
>
> the latest svn numpy version 1.1.0.dev5061 does not work with 
> matplotlib 0.90.1 (version shipped with enthought distribution), 
> unless a change in 
> Python25/Lib/site-packages/matplotlib-0.90.1.0003-py2.5-win32.egg/matplotlib/numerix/ma/__init__.py 
> is done:
>
> $ diff __init__.py.orig __init__.py
> 12c12
> < from numpy.core.ma <http://numpy.core.ma> import *
> ---
> > from numpy.ma <http://numpy.ma> import *
>
> Maybe this should be forwarded to the pylab mailing list, but I'm not 
> subscribed there...
>
> L.
>
> ------------------------------------------------------------------------
>
> _______________________________________________
> Numpy-discussion mailing list
> Num...@sc...
> http://projects.scipy.org/mailman/listinfo/numpy-discussion
> 
-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA