[Date Prev][Date Next] [Thread Prev][Thread Next]
[Date Index] [Thread Index] [New search]

RE: PDF graphics bewilderment



It's the downsampling that's killing your screenshots. Bitmap formats just
don't survive that sampling. Stick with 8-bit Zip compression. It's reported
to be lossless.


Jim Stauffer
Senior Technical Writer
Pacific Broadband Communications 
3103 North First Street 
San Jose, CA 95134 USA 
408.468.6182 tel 
408.468.6297 fax 
jstauffer@pbc.com 
http://www.pacificbroadband.com 



> -----Original Message-----
> From: Thomas Michanek [mailto:thomas.michanek@iar.se]
> Sent: Wednesday, July 25, 2001 9:29 AM
> To: framers@omsys.com
> Subject: PDF graphics bewilderment
> 
> 
> Dear Framers and Adobers,
> 
> I cannot get any sense out of my PDF experiments from FrameMaker.
> Why does some color graphics end up "blocky" and "blurry", while
> others end up crisp and clear? Please enlighten me on how Acrobat
> Distiller is supposed to handle color graphics. And excuse me for
> a long message; I want to provide as much details as necessary.
> 
> Background: I use FM 5.5.6 on Windows 2000 SP1 with Acrobat 4.05a.
> The FM files contain color screendumps taken with PaintShop Pro,
> saved as 24-bit TIFFs and also as 8-bit and 4-bit color-reduced
> TIFFs, and imported at 96/150/200 dpi (depending on window size).
> I create PDFs by printing a PS file using the Acrobat Distiller
> printer, created using AdobePS 5.1.2 (UI 1.0.2) and Adist4.ppd.
> I view the PDFs with Smoothing turned on, and when I talk about
> "blurry" graphics, it's not the effect of smoothing in Acrobat.
> 
> I want to create PDFs for on-line viewing purposes. I use a
> pre-made Distiller joboptions file with the following settings:
> * Resolution 600 dpi, Binary format, non-optimized
> * Downsampling to 72 dpi, Automatic compression, Medium quality
> (I think that's all the relevant ones in this case)
> 
> I thought that downsampling to 72 dpi would reduce the image
> data, so that zooming in on a high-dpi image would reveal no
> further details, but only make the image "blocky". However, this
> is not the case: the 150/200-dpi 8-bit and 4-bit TIFFs can be
> zoomed in to reveal all pixels (details) of the original image.
> Have I misunderstood what downsampling to 72 dpi means?
> 
> Further, the high-dpi 24-bit TIFFs look very bad: they are
> pixelated (low-res), blurry and contain color artifacts.
> Text and icons cannot be distinguished. None of this affects
> the 8/4-bit TIFFs or the 96-dpi 24-bit TIFFs.
> Why are some 24-bit TIFFs handled differently???
> 
> To overcome the problems with 24-bit TIFFs, I have tried using
> downsampling to 144 dpi, as well as no downsampling at all.
> Both results in larger PDF files (expected), and the 24-bit TIFFs
> now have the same quality and "behavior" as the 8/4-bit TIFFs.
> Great, but why doesn't this work for 72 dpi downsampling?
> 
> I have also experimented with downsampling to 10 dpi, just to
> check the downsampling. The result baffles me: only the 24-bit
> TIFFs are affected in the way I would expect, but the 8/4-bit
> TIFFs are unaffected and appear in full glory!
> 
> This makes me wonder: are there bugs or undocumented limitations
> on how the Distiller handles 24-bit color graphics (as opposed
> to 8/4-bit color graphics), and how downsampling works?
> Are these problems in any way related to Windows 2000?
> Or have I completely misunderstood these concepts?
> 
> I can provide sample files, if someone needs to see the
> effects I'm trying to describe, or to trouble-shoot for me.
> 
> Very many thanks in advance,
> 
> - - - - - - - - - - - - - - - - - - - - - -
> Thomas Michanek, Technical Writer
> IAR Systems AB, Sweden: http://www.iar.com
> mailto:Thomas.Michanek@iar.se
> Tel: +46 18 167800, Fax: +46 18 167838
> - - - - - - - - - - - - - - - - - - - - - -
> 
> ** To unsubscribe, send a message to majordomo@omsys.com **
> ** with "unsubscribe framers" (no quotes) in the body.   **
> 

** To unsubscribe, send a message to majordomo@omsys.com **
** with "unsubscribe framers" (no quotes) in the body.   **