| Author | Thread |
|
|
07/15/2003 03:26:44 PM · #1 |
Hi...
Just wanted to ask a question... How do you get thumbnails of pictures without them blurring? I have a Fuji FinePix Pro and when i try to shrink down the pics to thumbnails (for my website) they blur... Dunno what I am doing wrong or if the pictures are just too big to be resized down so small!!
Can someone help me out please??
Thanks in advance
Terri |
|
|
|
07/15/2003 03:49:57 PM · #2 |
Resizing any picture dramatically will always result in blurring. It's unavoidable.
Just sharpen your images after resizing.
|
|
|
|
07/15/2003 04:17:45 PM · #3 |
| Shrinking a photo to make a thumbnail shouldn't cause a lot of blurring, though... my guess is that you're possibly saving the .jpg at too compressed a level. Check the .jpg properties and see if you can bring the quality up a bit when you're saving the thumbnails. |
|
|
|
07/15/2003 04:38:07 PM · #4 |
| imagine 4 pixels in a 2x2 square.. top left white, top right black, bottom left blue, bottom left red.. Now reduce this image to 25%.. what colour is the one pixel you have left? This is exactly why you get blurring when you reduce beyond a certain level.. go with roley and use a sharpen filter and you should get SOME of the quality back.. |
|
|
|
07/15/2003 04:40:45 PM · #5 |
Originally posted by Alpine99:
imagine 4 pixels in a 2x2 square.. top left white, top right black, bottom left blue, bottom left red.. Now reduce this image to 25%.. what colour is the one pixel you have left? |
Depends what resampling algorithm you use too...
|
|
|
|
07/15/2003 05:24:27 PM · #6 |
Originally posted by Gordon:
Originally posted by Alpine99:
imagine 4 pixels in a 2x2 square.. top left white, top right black, bottom left blue, bottom left red.. Now reduce this image to 25%.. what colour is the one pixel you have left? |
Depends what resampling algorithm you use too... |
This got me wondering so I did just this in photoshop.
The resulting color of the pixel:
bicubic - light purple (#894989)
bilinear - same
nearest neighbor - almost black (#222222)
|
|
|
|
07/15/2003 05:41:53 PM · #7 |
| I seem to remember Photoshop uses something like the four corners and center pixel of a 3x3 array, but I've avoided knowing the details until now .... it just seems logical that if you use an exact reduction ratio (2:1, 3:1, 4:1, etc.) you would have more accurate results and faster computations--but it could be just the opposite! |
|
|
|
07/15/2003 05:45:40 PM · #8 |
Originally posted by greenem2:
Originally posted by Gordon:
Originally posted by Alpine99:
imagine 4 pixels in a 2x2 square.. top left white, top right black, bottom left blue, bottom left red.. Now reduce this image to 25%.. what colour is the one pixel you have left? |
Depends what resampling algorithm you use too... |
This got me wondering so I did just this in photoshop.
The resulting color of the pixel:
bicubic - light purple (#894989)
bilinear - same
nearest neighbor - almost black (#222222) |
nearest neighbour is going to get you one of the original pixels, bilinear gets you an average of 4 pixels, bicubic will give you something different if you have more pixels, as it uses basically a more complex average function over a wider set of pixels.
nearest neighbour tends to give the harshest results, but is much faster, as all you are doing is throwing pixels away. It also tends to work best for integer resizing.
Bilinear tends to smooth edges, which is why you need to sharpen images afterwards (e.g., a sharp black to white edge becomes a averaged, softer grey edge)
Bicubic also does this softening but in a more intelligent, and computationally more expensive (hence slower) way.
|
|
|
|
07/15/2003 05:53:31 PM · #9 |
| I've never had anyone tell be to use anything other than bicubic, including Adobe employees who weren't quite sure why the other methods continued to be included. |
|
|
|
07/15/2003 06:16:05 PM · #10 |
| nearest neighbor is very useful for enlarging images for web use, when you want to show what is happening at the pixel level (photoshop needs it when you zoom in for the exact same reason) |
|
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: 08/27/2025 07:28:40 PM EDT.
