RE: render.c:GetPixelOpacity() optimization ideas?
Bill Radcliffe <[email protected]>
| Newsgroups | gmane.comp.video.graphicsmagick.core |
|---|---|
| Message-ID | <306F012D0A00D611BC9D0008C79192740D12F227@seamail1.continuum.corbis.corp> |
> floating point when in doubt, which seems to be the current philosophy. I believe that this is an accurate statement. It is tough to know what the round off situation would be switching from double to float on this. I am confident that Bob will weigh in shortly. > Would you be interested in working on such improvements as a (paid) Actually Bob would be a better choice then me. He is already taking contract work and my cycles for GM work have been nearly 0 of late. I have been darn near useless due to other work pressures during the last two months or more. -----Original Message----- From: [email protected] [mailto:[email protected]] On Behalf Of Wolfgang Spraul Sent: Thursday, January 15, 2004 4:28 PM To: [email protected] Subject: Re: [GM-core] render.c:GetPixelOpacity() optimization ideas? Bill - > I believe that the code you talking about is dealing with anti-aliasing. If that's the case, would it be possible to render all primitives that don't need anti-aliasing or where it's turned off without all those calculations? > I can say for sure, the there is an assumption in that code that floating > point is no more expensive then integer math, which is largely true on any > the usual CPU's we run on. Other can weigh in on this, but you could > probably generalize this assumption out to the entire package. Most of the > code does not hesitate to use floating point were integer might do just as > well. I am certain there are some exceptions, so don't jump all over me > with those. I am just talking broad brush. Well, that's not too good for me ;-))) What could be done about it? Would it be possible to define macros to switch from using type 'double' to type 'float', for example (float is faster on some platforms and in some software implementations)? Or to switch to using (much) more integer whereever possible, instead of use floating point when in doubt, which seems to be the current philosophy. Would you be interested in working on such improvements as a (paid) consultant? (I would need the results in about 4, at most 6 weeks though) Regards, Wolfgang ------------------------------------------------------- The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA. http://www.eclipsecon.org/osdn _______________________________________________ Graphicsmagick-core mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/graphicsmagick-core ------------------------------------------------------- The SF.Net email is sponsored by EclipseCon 2004 Premiere Conference on Open Tools Development and Integration See the breadth of Eclipse activity. February 3-5 in Anaheim, CA. http://www.eclipsecon.org/osdn