Hi Dar, > Date: Wed, 31 Dec 2003 01:01:18 -0700 > From: Dar Scott <[EMAIL PROTECTED]> > Subject: Re: More image processing > > > On Wednesday, December 31, 2003, at 12:33 AM, Ken Norris wrote: > >> put numToChar(0) into char x of binDat > > This takes a long time. ------------ How long? Surely not *that* long. Not more than a few minutes, although I agree it's slow. ------------ > Bugzilla 586 is an enhancement suggestion that > will drop the time to do this drastically. > > Instead build a new value with "put ... after ...". That would also > allow you to use a faster repeat. ------------ Well, it would go faster if I could find the rectangle of the painted section pasted into the image. That way I could use a formula to find the 'pixelToStart' and go from there to a 'pixelToStop' and be done.
OTOH, I once saw that the image object had concatenated itself to the size of the pasted area, which is much smaller of course, but I don't know how it happened, and I don't know if I could still set the windowShape to it. ------------ > BTW, I only see 0 and 255 in maskData. ------------ Right. There are only three pixel colors, black, white, and transparent. I'm trying to use only the black for the demask. I tried several methods to recover the white value with no success. It is apparently *not* 1. However, I *know* the black has to be the last value, or 255, in Rev's way of looking at things, so if I set everything else to 0 then it should leave only the black, which is what I want. ------------ > You might want to test with small images until those speed problems are > tracked down. ------------ OK, and I'll try something with the other method as well. What a pain this is. All I want is an 8-bit mask shape to fit my 32-bit graphic. Maybe I can bring it over from SC. I haven't tried that. Ken N. _______________________________________________ use-revolution mailing list [EMAIL PROTECTED] http://lists.runrev.com/mailman/listinfo/use-revolution
