SM> FW: Re: PWG Review of the IPP Color and Imaging Attributes versio n 0.1, Thur, 12/5/02, 1-2 PM EST

SM> FW: Re: PWG Review of the IPP Color and Imaging Attributes versio n 0.1, Thur, 12/5/02, 1-2 PM EST

SM> FW: Re: PWG Review of the IPP Color and Imaging Attributes versio n 0.1, Thur, 12/5/02, 1-2 PM EST

Hastings, Tom N hastings at cp10.es.xerox.com
Wed Dec 4 17:07:34 EST 2002


Here is a suggestion from Michael Sweet to add a 'bi-cubic' value to the
"resample-method" Job Template attribute proposal.  The "resample-method"
Job Template attribute is itself proposed to be added to the IPP Color and
Imaging Spec.

Comments?

Tom

-----Original Message-----
From: Hastings, Tom N 
Sent: Wednesday, December 04, 2002 13:56
To: 'digital_printing at cip4.org'
Subject: RE: CIP4 Forum Digital Printing: Re: PWG Review of the IPP
Color and Imaging Attributes version 0.1, Thur, 12/5/02, 1-2 PM EST


How about this definition for "bicubic" (no hyphen) from Delmar's Dictionary
of Digital Printing & Publishing:

Bicubic Interpolation - The use of two cubic functions to generate
interpolated values along a geometric surface.  Cubic functions are more
accurate than linear functions, and more difficult to compute, but yield
superior results.

So making it parallel to the bi-linear definition of:

'bi-linear'  A method used to resample image data (pixels) from one
resolution to another that is accomplished by using the weighted sum of the
four nearest pixel values in the source image to compute the replacement
pixel in the output (resampled) image.  This method would give higher image
quality than nearest-neighbor but would take more time to compute.
The algorithm to determine the "weighted sum" is implementation dependent.

we'd have:

'bi-cubic'  A method used to resample image data (pixels) from one
resolution to another that is accomplished by using the weighted sum of two
cubic functions of pixel values in the source image to compute the
replacement pixel in the output (resampled) image.  This method would give
higher image quality than 'bi-linear' but would take more time to compute.

The algorithm to determine the "weighted sum" and the two cubic functions is
implementation dependent.

Tom


-----Original Message-----
From: mike at easysw.com [mailto:mike at easysw.com]
Sent: Wednesday, December 04, 2002 13:22
To: hastings at cp10.es.xerox.com
Subject: CIP4 Forum Digital Printing: Re: PWG Review of the IPP Color
and Imaging Attributes version 0.1, T hur, 12/5/02, 1-2 PM EST


Hastings, Tom N wrote:
> ...
> We need to consider adding 'linear' and 'cubic' (or bicubic) with good
> definitions.  

bi-cubic should definitely be added; linear and cubic are probably
not appropriate unless you have a 1-pixel wide or height image.

> ...
> Also Acrobat can apply resampling independently to color bit map images,
> gray-scaled bit map images, and monochrome bit map images.

I think we can leave that to the PDL to define; the resample-method
attribute can specify the default setting and the PDL data can/may
override it.

-- 
______________________________________________________________________
Michael Sweet, Easy Software Products                  mike at easysw.com
Printing Software for UNIX                       http://www.easysw.com



-----
This message was submitted by 
Message URL: http://www.cip4.org/intern/groups/email_details.php?eid=4184
Forum URL:   http://www.cip4.org/intern/groups/email_forum.php?gid=2




More information about the Sm mailing list