Building fast box to open pdf files

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View
I'm in the process of building a new PC. The PC needs to be able to
open large Adobe Acrobat pdf files as quickly as possible (the pdf
files are generated from a scanner). What will be the bottleneck in the
system? Stated another way, I'm willing to spend extra $$ on the ONE
component that will make things go faster. Where should I spend the
extra money? Thanks.


Re: Building fast box to open pdf files

Quoted text here. Click to load it

What do you call "large"?

Don't you mean make large PDF files, not "open"?

Are you making images or text documents via OCR?

How much money are you spending on the scanner?  That could very well
be the bottleneck.  (there are batch scanners that do 4000 sheets per
hour, including OCR.)

a d y k e s @ p a n i x . c o m

Don't blame me. I voted for Gore.

Re: Building fast box to open pdf files

Thanks for the reply. Here are the answers to your questions:
  Large = 10 to 20 MB
  I mean "open" as in "opening them up to view them"
  I am making images, not text documents or doing OCR.
  The scanner isn't the bottleneck. Opening pre-existing pdf files is
too slow.


Re: Building fast box to open pdf files

Quoted text here. Click to load it
Not much you can do about it. Acroread is a pig. THere are applications
out there to speed it up and tweak guides on the net but thats about


Same shit, different day.

Re: Building fast box to open pdf files

On 4 Mar 2006 20:16:29 -0800, wrote:

Quoted text here. Click to load it

There are two issues:

1)  For a moment, forget that it's an image, or a PDF, or
Acrobat, and just do same things you'd do to optimize any
system.  Ensure enough memory in the system for the things
you're running (total memory usage to allow caching files
and minimizing pagefile hits), ample unfragmented HDD space,
sufficiently modern (performance) hard drives with enough
free space that all the files are on the faster front end of
the drive.  

2)  Disable Acrobat's unused plugins.  This is the most
significant for typical use.  Acrobat has grown quite
bloated and seldom do people need ALL those plugins but they
take up most the time to load Acrobat.   See the following
page for the manual directions to do it, though with a
Google search you should also find separate utilities to
automate the process some.

Re: Building fast box to open pdf files


Quoted text here. Click to load it

        Simple.  Acrobat Reader is too slow.

        Try Foxit Reader ( ) for a much faster
PDF reader.


Re: Building fast box to open pdf files

Quoted text here. Click to load it
Hi H. D\load Foxit PDF Reader ( it`s free !), and see if it will
do all you want. It knocks Acrobat reader sideways.


Re: Building fast box to open pdf files

I'd like to use "Foxit PDF Reader", but that solution doesn't work for
me. My PaperPort Pro 10.0 software is integrated with Acrobat.


Re: Building fast box to open pdf files

Quoted text here. Click to load it


a d y k e s @ p a n i x . c o m

Don't blame me. I voted for Gore.

Re: Building fast box to open pdf files

PapperPort Pro 10.0 uses pdf as its default format for its files. I
also use Acrobat for lots of other purposes. Bottom line: I need help
in designing a box that's optimized to OPEN pdf files. Thanks.


Re: Building fast box to open pdf files wrote:
Quoted text here. Click to load it

Any of the latest systems will open pdfs fast.
I have no problems opening a 150mb Acrobat document with a
Pentium 4 640 and a Gig of RAM.

Re: Building fast box to open pdf files wrote:
Quoted text here. Click to load it
I can understand your wanting to keep Acrobat but the others are
correct, the easiest way to speed things up with a different pdf viewer.
Also, Paperport can be configured to use any PDF viewer, not just
Acrobat, and do it seamlessly. How much ram are you using now?

Re: Building fast box to open pdf files

I'm using 1.5 GB of RAM. How can I re-configure PaperPort to use an
alternative PDF viewer?


Re: Building fast box to open pdf files

On 5 Mar 2006 19:54:37 -0800, wrote:

Quoted text here. Click to load it

First try disabling the unnecessary plugins for Acrobat.
There may not be any need to install another 3rd party

If you find Acrobat is still too slow (or to then compare
the more optimized Acrobat configuration), try installing
your choice of other PDF viewer.   Typically installing such
a viewer would make PDFs open with it by default, or if
necessary you can right-click on a PDF file while holding
down the keyboard <SHIFT> key, choosing "Open with", and
browsing to the newly installed PDF viewer application if it
isn't on the list Explorer presents, then choose to make it
the default handler for that file type.

An alternate way to associate the files is through Explorer
(open any folder), Tools->Folder Options->File Types

Re: Building fast box to open pdf files

Thanks for the software hints. I'm going to them implement today.

However, back to the question about building a box that will quickly
open big PDF files. Am I correct in assuming that I'm not going to get
too much "bang for the buck"? My sense is that a Falcon NW box isn't
going to blow away an eMachines box when it comes to opening an Acrobat
file. Three seconds isn't worth a $1000, but 30 seconds is worth it. Am
I interepreting the comments on this board correctly or is there a
single piece of hardware that will give me the performance boost that
I'm seeking?

Re: Building fast box to open pdf files

On 6 Mar 2006 08:39:09 -0800, wrote:

Quoted text here. Click to load it

You have some kind of hardware misconfiguration perhaps?
It doesn't take 3 seconds TOTAL to open a typical 10MB
PDF/scan with Acrobat once the unnecessary plugins are
disabled unless one used a very slow/old system (like 6+
years old).

How long is it typically taking?  Are any other applications
behaving so badly?  Have you tried uninstalling then
reinstalling Acrobat?  Which version?

So yes, you should not get much bang for the buck buying a
new system unless it's to replace one very old, and even
then, maybe 3 seconds difference between the two.  That's
the first time it's loaded.  Subsequent loads should be even
faster, it doesn't take more than 3 seconds to load a 10MB
file here when it's being accessed across a 100MB lan if
Acrobat had already been running (and subsequently exited).

Re: Building fast box to open pdf files wrote:

Quoted text here. Click to load it

Maybe the Paperport program generates inefficient PDF files ?

I have a copy of Acrobat Distiller here, but I also did a test
with the free Ghostscript program.

I started with a 92MB source file, a scan of a portion of a map.

Using Photoshop, I printed the file to PostScript. The output
file is 278MB or so.

I did two Distiller runs. In one run, I disabled image compression.
In the other run, I enabled LZW 8 bit compression for color

The output files were about 100MB or so. Surprisingly, there
wasn't that large of a difference in file size between the
two of them.

Using Acrobat Reader, I opened the files. They don't open
such that the whole image is visible. As a test, I timed from
the time I select "fit to page" (i.e. view the full image)
until the page is finished redrawing. The uncompressed file
redraws in 3 seconds. The compressed file redraws in 7 seconds.

And that was for a 92MB 24 bit color source image.

Is it actually taking 30 seconds to view a 10MB image ?

I suspect there must be some non-trivial differences in
the execution environment here. While I agree with the
general comments that Acrobat Reader is a pig (non-scalable
with increased CPU clock speed), taking 30 seconds seems a
bit extreme.

The GhostScript program has a script called "ps2pdf" and
offers several flavors "ps2pdf13" etc. It is not limited
to dealing with PostScript as a source. You can also feed
it PDF and have PDF as output. There are also command line
options you can feed to the script, which give you partial
control of the output. For example, the following
command would prevent the image from being compressed.

ps2pdf13 -dEncodeColorImages=false output.pdf

278MB PostScript source file
   Uncompressed run, Distiller execution time 2:01 min, 102.5MB output
   Compressed run, Distiller execution time 2:07 min, 87.9MB output
   Convert compressed to uncompressed with Ghostscript
      87.9MB source --> 102.5MB output PDF, time 0:26 min

Time to zoom out so full page can be seen:

   Distiller uncompressed image, 3-4 seconds.
   Distiller compressed image,   7 seconds
   Ghostscript conversion of compressed to uncompressed, 3-4 seconds.

You must have a major anti-virus scanner "helping" you open
the files or something. My machine is not recent hardware
by any stretch of the imagination. I don't use anti-virus
software on this computer.

If you want to experiment with GhostScript reencoding the
PDF files for you, the latest one gs853w32.exe is here:

Install the program. Have it install the fonts as well.
Open a Command Prompt:

   path=%path%;C:\Program Files\gs\gs8.53\bin
   path=%path%;C:\Program Files\gs\gs8.53\lib
   cd <where_the_source_PostScript_or_PDF_file_is>
   ps2pdf13 "-dEncodeColorImages=false" input.pdf output.pdf

That should allow you to experiment with reencoding the files.
The "Ps2pdf" file in the "doc" directory of the install, lists
some of the options available.

Note - My understanding of MSDOS is pretty limited, and I
used the double quotes around the passed parameter, as
it didn't work right otherwise. It could be the equals sign
confused things.

If you have some web space somewhere, it might be interesting
to examine a _small_ scanned image (in PDF form) from your
PaperPort software. A small image may open too quickly to make
a good benchmark, but looking at the code inside the small PDF
file, might suggest what the PaperPort software has done. Due
to the binary representation used in PDF, I don't know if
all of the file can be read by a human, but portions of it are
readable text.

For example, inside one of my test PDF files, I can see
this chunk of code. This is what GhostScript uses as
an uncompressed specification.

/ColorSpace 8 0 R
/Width 4979
/Height 7200
/BitsPerComponent 8
/Length 107546400>>stream

and that is a specification for how to decode the following
stream of binary image information.

This is what an uncompressed PDF file from Acrobat Distiller uses.
The colorspace option likely will cause some kind of color
correction after the image is loaded.

/Type /XObject
/Subtype /Image
/Name /Im1
/Width 4979
/Height 7200
/BitsPerComponent 8
/ColorSpace 2 0 R
/Length 107546401
Quoted text here. Click to load it

This is a compressed PDF file from Acrobat Distiller:
The LZWDecode is consistent with my "Job Option" selection
of LZW as a compression method within Distiller.

/Type /XObject
/Subtype /Image
/Name /Im1
/Width 4979
/Height 7200
/BitsPerComponent 8
/ColorSpace 2 0 R
/Length 92211452
/Filter /LZWDecode
Quoted text here. Click to load it


Re: Building fast box to open pdf files

You've nailed the issue Paul. The problem/choke point is PaperPort Pro
10's rendering of PDF files. Files flew open with my old copy of
Acrobat Distiller and an ancient version of Ghostscript.

My mistake is that I stupidly assumed that PaperPort was using Acrobat
to render to PDF files. When I double clicked on my PDF files in
Explorer, they flew open with Acrobat. When I double clicked on the
file while in PaperPort's desktop, the system choked. I then changed
the default parameters in PP so that Acrobat is now used to render PDF
files within PP (I use the term "within" very loosely) instead of PP's
ineffficient algorithm. Everything is back on the fast lane.

I'm still going to build the box, but now I can focus more important
issues like backup (RAID 1) and sound proofing.

Site Timeline