Welcome to TalkGraphics.com
Page 2 of 2 FirstFirst 12
Results 11 to 17 of 17

Thread: Jagged fill

  1. #11
    Join Date
    Aug 2010
    Location
    Sofia, Bulgaria
    Posts
    19

    Default Re: Jagged fill

    Funny, because i just did the same type of gradient on PS, and it doesn't show any banding, allClick image for larger version. 

Name:	Untitled-1.jpg 
Views:	101 
Size:	7.5 KB 
ID:	87204 on the same monitor.

  2. #12
    Join Date
    Feb 2007
    Location
    UK
    Posts
    21,283

    Default Re: Jagged fill

    ah well - vector and bitmap[pixel] are different.. right?

    a pixel gradient such as you have in PS is not made from incremental steps that produce bands, its a continuous 'stream' of pixel matrix, dithered or otherwise

    the steps that produce the bands are a vector 'limitiation' - vectors are mathematically calculated objects which show up as the bands..

    one reason why it's common to use bitmaps as texture
    -------------------------------
    Nothing lasts forever...

  3. #13
    Join Date
    Aug 2010
    Location
    Sofia, Bulgaria
    Posts
    19

    Default Re: Jagged fill

    That is correct.

  4. #14

    Default Re: Jagged fill

    A couple thoughts on banding in vector-based apps.

    If the
    1. re is too little color shift, banding will be more apparent.
    2. Banding is worse when using a single color (say only the K component) and dropping only 15% or less from one pole to the other over a long enough distance.
    3. You will almost always have a step in the transition if you drop to white (0,0,0,0). So with a K only gradient, stop at 5% or less (but not 0%). The film at 150 lpi or greater will be almost white (paper color) at 5%.
    4. Outputting to a level 3 postscript produces "smooth shading" which boosts the inherent limitation of the vector gradient from 8-bit to 11-bit. The result will be a smoother interpretation of the bitmap dump that hits the output device if going to a rip. I.e., a postscript laser printer may not show how smooth the gradient can be, but a rip will. Doesn't help much on lower-end digital printers, but does with better equipment.
    5. If at all possible, include more than K in a gradient. A richer black once output will likely print better than it looks on-screen because of the screening in the rip.


    There is a reason that designers often resort to PhotoShop or another good image editor for gradients--which one can also do within XDP. With added noise and a low amount of gaussian blur, a gradient has a better chance of avoiding banding noticeable, especially if combined with output to a PS level 3 device.

    Enough babble for now. Take care, Mike

  5. #15
    Join Date
    Feb 2007
    Location
    UK
    Posts
    21,283

    Default Re: Jagged fill

    thanks Mike..
    -------------------------------
    Nothing lasts forever...

  6. #16
    Join Date
    Aug 2000
    Location
    Virginia
    Posts
    4,432

    Default Re: Jagged fill

    Useful info, Mike. Thanks.

  7. #17

    Default Re: Jagged fill

    the first point should be--but I cannot edit the post--is:

    If there is too little color shift, banding will be more apparent.

    Sorry about that and thank you for the kind words.

    Mike

 

 

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •