« Return to Thread: A small optimization to gradient_linear_color
Re: A small optimization to gradient_linear_color
by Petr Kobalíček
::
Rate this Message:
Hi Vinnie,
if you are using only 1 pixel wide lines, I think that WU algorithm
can be really fast compared to antigrain. Remember that antigrain
producess scanline and span objects, so it needs first to serialize
data thats rendered later.
The SSE2 code I posted here contains bugs, I recommend to extract it
from SVN I also posted (there are fixed all of them, i hope).
> agg::rasterizer_cells_aa<>::line()
>
> This function looks complex and it seems to do a lot of work. What does it do? What is the meaning of the parameter list (x1, y1, x2, y2) ?
I don't know exactly, but I think that it will process one line from
list of lines that was generated from curves (line is main rasterizer
object). Debug versions are slow, but Release should be fast (faster
than cairo or gdi+). You can also try to switch compiler to use SSE2,
this should also help.
- Petr
2009/5/9 Vinnie <thevinn@...>:
>
>> From: Petr Kobal??ek <kobalicek.petr@...>
>> I'm repeating here, but some time ago I wrote MMX and
>> SSE2 optimized interpolation between two colors based on
>> linear increments
>
> Those are some nice optimizations! The problem is that my app is spending by far, the majority of its time in:
>
> agg::rasterizer_cells_aa<>::line()
>
> This function looks complex and it seems to do a lot of work. What does it do? What is the meaning of the parameter list (x1, y1, x2, y2) ?
>
> ------------------------------------------------------------------------------
> The NEW KODAK i700 Series Scanners deliver under ANY circumstances! Your
> production scanning environment may not be a perfect world - but thanks to
> Kodak, there's a perfect scanner to get the job done! With the NEW KODAK i700
> Series Scanner you'll get full speed at 300 dpi even with all image
> processing features enabled. http://p.sf.net/sfu/kodak-com
> _______________________________________________
> Vector-agg-general mailing list
> Vector-agg-general@...
> https://lists.sourceforge.net/lists/listinfo/vector-agg-general
>
------------------------------------------------------------------------------
The NEW KODAK i700 Series Scanners deliver under ANY circumstances! Your
production scanning environment may not be a perfect world - but thanks to
Kodak, there's a perfect scanner to get the job done! With the NEW KODAK i700
Series Scanner you'll get full speed at 300 dpi even with all image
processing features enabled. http://p.sf.net/sfu/kodak-com
_______________________________________________
Vector-agg-general mailing list
Vector-agg-general@...
https://lists.sourceforge.net/lists/listinfo/vector-agg-general
if you are using only 1 pixel wide lines, I think that WU algorithm
can be really fast compared to antigrain. Remember that antigrain
producess scanline and span objects, so it needs first to serialize
data thats rendered later.
The SSE2 code I posted here contains bugs, I recommend to extract it
from SVN I also posted (there are fixed all of them, i hope).
> agg::rasterizer_cells_aa<>::line()
>
> This function looks complex and it seems to do a lot of work. What does it do? What is the meaning of the parameter list (x1, y1, x2, y2) ?
I don't know exactly, but I think that it will process one line from
list of lines that was generated from curves (line is main rasterizer
object). Debug versions are slow, but Release should be fast (faster
than cairo or gdi+). You can also try to switch compiler to use SSE2,
this should also help.
- Petr
2009/5/9 Vinnie <thevinn@...>:
>
>> From: Petr Kobal??ek <kobalicek.petr@...>
>> I'm repeating here, but some time ago I wrote MMX and
>> SSE2 optimized interpolation between two colors based on
>> linear increments
>
> Those are some nice optimizations! The problem is that my app is spending by far, the majority of its time in:
>
> agg::rasterizer_cells_aa<>::line()
>
> This function looks complex and it seems to do a lot of work. What does it do? What is the meaning of the parameter list (x1, y1, x2, y2) ?
>
> ------------------------------------------------------------------------------
> The NEW KODAK i700 Series Scanners deliver under ANY circumstances! Your
> production scanning environment may not be a perfect world - but thanks to
> Kodak, there's a perfect scanner to get the job done! With the NEW KODAK i700
> Series Scanner you'll get full speed at 300 dpi even with all image
> processing features enabled. http://p.sf.net/sfu/kodak-com
> _______________________________________________
> Vector-agg-general mailing list
> Vector-agg-general@...
> https://lists.sourceforge.net/lists/listinfo/vector-agg-general
>
------------------------------------------------------------------------------
The NEW KODAK i700 Series Scanners deliver under ANY circumstances! Your
production scanning environment may not be a perfect world - but thanks to
Kodak, there's a perfect scanner to get the job done! With the NEW KODAK i700
Series Scanner you'll get full speed at 300 dpi even with all image
processing features enabled. http://p.sf.net/sfu/kodak-com
_______________________________________________
Vector-agg-general mailing list
Vector-agg-general@...
https://lists.sourceforge.net/lists/listinfo/vector-agg-general
« Return to Thread: A small optimization to gradient_linear_color
| Free embeddable forum powered by Nabble | Forum Help |
