> -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of Bob Talbot > Sent: Wednesday, October 12, 2005 11:15 PM > To: [email protected] > Subject: [inbox] Re: EOS 1D Mark II > > > >Photoshop books and courses I've been to all suggested > > applying many successive small increments > > rather than large increments to the final resolution. > > I've read this before and can't help the immediate feeling of "it must > be BS". > > From a pure data viewpoint every step MUST result in loss of colour > information as every step involves integer rounding. > > From a pure "bicubic (spline) interpolation" viewpoint I'm also > struggling to understand how multiple small step could be better than > one step anyway (unless the interpolation algorithm was poor). > > > Maybe I'll google for an objective comparison > Perhaps it's time to do a little home-brew comparison myself. > > > "I'll be back" :o) >
Hi BobT, I thought the same thing when I read the incremental resampling thing but as it turns out there was both practical evidence and mathematical proof for why this is so. I can't find the links to what I read but it's out there. I've tried it myself and it works as advertised. The differences are not huge but seeable if you have to upres more than 200% or so as I recall. But once I tested QImage I had to but it as there is basically no way to get upresed images to look as good, try it on large prints where you are resampling 200% or more, do not just go by what you see on the screen as there is more to it than this. Cheers/Chip * **** ******* *********************************************************** * For list instructions, including unsubscribe, see: * http://www.a1.nl/phomepag/markerink/eos_list.htm ***********************************************************
