| Author | Thread |
|
|
10/18/2002 03:41:43 PM · #1 |
OK, I know this is going to be over some people's heads, but I'm gonna ask anyway since I know some of you know what I'm talking about.
Is there a magic number to use for step interpolation increments? I can see the difference between 5% increments and a 100% jump so I'm a believer in SI. However, I recently decided to try increasing by 1% increments (if little is better, even smaller must be even better) and discovered 1% results in a much more pixellated image than 5%. It makes me wonder if maybe 7.5% would be better than 5%, or 10% or 12.1325%.....
Anybody know???
|
|
|
|
10/18/2002 03:57:52 PM · #2 |
you are doing interpolation, so doing it on even pixel boundaries
will tend to introduce less averaging/ rounding errors.
Too small will cause this so 1% probably gives bad results.
Much of this sort of image processing stuff isn't actaully a very
precise science (even though all of the underlying principles are
well defined mathematically and understood) Once you start turning
it into quantised values for actual pixels, all the maths mostly goes
out of the window, via the door of human perception.
For the many years I've been learning, teaching and earning money doing
image processing, the maxim 'if it looks good enough then it is good'
has held true. I'd suggest trying different values and seeing how you
like the results.
|
|
|
|
10/18/2002 04:00:11 PM · #3 |
:-P
Gordon, *whine* I want an easy answer that I can apply to everything!!!
*grin*
Seriously, thanks, I was just wondering if there was any special numbers, but if there isn't, I'll keep playing with 5% since that's what I've set up all my actions to do...
|
|
|
|
10/18/2002 04:01:22 PM · #4 |
Fred Miranda sells something that does this in photoshop, i think it has increments of 10% with an unsharp mask every time the image size doubles.
Originally posted by Patella:
:-P
Gordon, *whine* I want an easy answer that I can apply to everything!!!
*grin*
Seriously, thanks, I was just wondering if there was any special numbers, but if there isn't, I'll keep playing with 5% since that's what I've set up all my actions to do...
|
|
|
|
10/18/2002 04:06:49 PM · #5 |
Pag,
Please join us in chat.
dpcfanatics
|
|
|
|
10/18/2002 04:16:23 PM · #6 |
I tried some various SI methods and, while I think there is a lot to it, I have found it to be just a little tedious for me. I have had great results using Genuine Fractals. Not always in the same way, however. I usually do not like to sharpen my photos after I have applied GF. I have had some very good results by sharpening the photo just a little bit before I apply GF. I keep trying to come up with one universal method for upsizing and sharpening but I have not yet found it, either.
T
|
|
|
|
10/18/2002 04:38:52 PM · #7 |
Originally posted by Patella:
:-P
Gordon, *whine* I want an easy answer that I can apply to everything!!!
*grin*
Seriously, thanks, I was just wondering if there was any special numbers, but if there isn't, I'll keep playing with 5% since that's what I've set up all my actions to do...
Well - there will be an optimum number, but it'll be based on your image
size, rather than a percentage size. Without really sitting down and
working it out, I can see that 10% increase on a 100 pixel image will
probably be fine, whereas a 10% increase on a 133 pixel image will
probably have nasty errors (due to that 13.3 pixel sample region)
That isn't the clearest description of the problem, but it probably
shows the issue.
|
|
Home -
Challenges -
Community -
League -
Photos -
Cameras -
Lenses -
Learn -
Help -
Terms of Use -
Privacy -
Top ^
DPChallenge, and website content and design, Copyright © 2001-2025 Challenging Technologies, LLC.
All digital photo copyrights belong to the photographers and may not be used without permission.
Current Server Time: 08/26/2025 02:34:47 PM EDT.
