Scanned PDF is always crooked (need to rotate PDF a degree or three)

[Fran Jones]

My HP Laserjet 3200m printer/scanner often scans pages off kilter.
It's noticeable, but slight (maybe a couple of degrees).

If it were a JPEG, I could use The Gimp freeware to straighten
but it's a PDF. If I must, I could scan to JPEG and then use
cutePDF freeware or Adobe Acrobat 6.0 Standard payware to then
print the straightened JPG to PDF - but that seems convoluted.

Is there freeware that will rotate a PDF page by just a degree
or three (just like The Gimp rotates an image)?

[Robert Heller]

At Tue, 8 Oct 2013 22:45:36 +0000 (UTC) Fran Jones <Fran...@is.invalid> wrote:

>
> My HP Laserjet 3200m printer/scanner often scans pages off kilter.
> It's noticeable, but slight (maybe a couple of degrees).
>
> If it were a JPEG, I could use The Gimp freeware to straighten
> but it's a PDF. If I must, I could scan to JPEG and then use
> cutePDF freeware or Adobe Acrobat 6.0 Standard payware to then
> print the straightened JPG to PDF - but that seems convoluted.

Not when you consider that the PDF is probably just a wrapper around a jpeg
image.

>
> Is there freeware that will rotate a PDF page by just a degree
> or three (just like The Gimp rotates an image)?
>
>

--
Robert Heller -- 978-544-6933 / hel...@deepsoft.com
Deepwoods Software -- http://www.deepsoft.com/
() ascii ribbon campaign -- against html e-mail
/\ www.asciiribbon.org -- against proprietary attachments

[sandy]

On Tue, 08 Oct 2013 19:45:14 -0500, Robert Heller wrote:

> Not when you consider that the PDF is probably just a
> wrapper around a jpeg image.

If you can rotate 180 degrees, why not 1 degree?

[Wilfried]

I do not know any method to do this by just modifying the pdf. Rotation
by modifying the pdf is restricted to 90°, 180°, 270° AFAIK

The only way to do what you need is to render the pdf to a bitmap,
rotate the bitmap, and write the rotated bitmap to a pdf.
However this could be done in one command (in the command prompt or a
batch file) if you install ImageMagick and GhostScript.
The following command calls ImageMagick's convert which in turn uses
Ghostscript to convert the pdf to bitmap and the bitmap to pdf
(command is here broken into three lines but must be entered in one
command line):

convert -density 300 tilted.pdf -rotate 1.5 -alpha off -monochrome
-units PixelsPerInch -page 2480x3508 -density 300
-compress Group4 corrected.pdf

The options "density", "page" and "monochrome" / "colorspace" must be
set according to the original pdf and what you want the result to be,
the option "rotate" defines the rotation angle in degrees (here 1.5 deg
clockwise, use negative number for counter-clockwise), and the option
"compress" must match the image type ("Group4" for monochrome, "jpeg"
for color).
I had to experiment with the settings to achieve an output pdf with a
file size and resolution comparable to the original. ImageMagick tends
to create output files much larger than the input.

HTH
--
Wilfried Hennings
please reply in the newsgroup, the e-mail address is invalid

[Peter Flynn]

The PDF is just an enclosure. The image is inside it. What you need to
do is extract the image, rotate it, and recreate the PDF, eg

$ pdfimages -j scanfile.pdf scan
$ convert scan-001.jpg -rotate 2 scanfile.pdf

pdfimages is part of the Poppler Utilities; the -j forces JPG output;
the third argument is a base-name for the extracted image (images lose
their names when you embed them in a PDF, so they need naming when they
are extracted). The first (or only) file will use this base-name plus a
hyphen and a serial number starting at 1, plus the extension.
convert is the ImageMagick conversion program.

///Peter

[occam]

I do not use 3200, but a different HP product (3-in-one). After a scan,
one of the options I am offered (apart from rotate 90% left, right) is
'Skewscan'. This corrects (automatically) the issue you are describing.
Are you sure you do not have a similar post-scan option?

[Joe Beanfish]

Increments of 90 degrees are programmatically trivial. Other increments
are more difficult and require math. So the answer is probably laziness.

[Anonymous]

>> Is there freeware that will rotate a PDF page by just a degree
>> or three (just like The Gimp rotates an image)?
>
>I do not know any method to do this by just modifying the pdf. Rotation
>by modifying the pdf is restricted to 90°, 180°, 270° AFAIK

The commercial tool "PaperPort (tm)" can do it. It can do it
automatically.. and if the automatic approach fails, then you can pull
out a line drawing tool where you drop a line along the bottom of some
text, and it will align based on that.

AFAIK, there is no freeware that does it.

[Axel Berger]

Joe Beanfish wrote:
> So the answer is probably laziness.

Not in this case. An embedded image is a rectangle with height and width
and that doesn't change with rotations through right angles. Any other
angle creates crooked shapes that are much harder to deal with.

[yufan]

Perhaps Foxit Reader may fit your needs. It provides PDF creation / PDF
conversion. Most important----it is free.
--
yufan
stev...@yahoo.com

[yufan]

Why not have a try with Foxit PDF Reader. It's more fast and safe.
--
yufan
stev...@yahoo.com

[Joe Beanfish]

On Fri, 11 Oct 2013 22:00:07 +0200, Axel Berger wrote:
> Joe Beanfish wrote:
> > > If you can rotate 180 degrees, why not 1 degree?
> >
> > Increments of 90 degrees are programmatically trivial. Other increments

> > are more difficult and require math. So the answer is probably laziness.

>
> Not in this case. An embedded image is a rectangle with height and width
> and that doesn't change with rotations through right angles. Any other
> angle creates crooked shapes that are much harder to deal with.

So, in your first sentence you disagree with me. Then in the remainder
you reiterate and elaborate on my point. Make up your mind.

Part of the non-90 degree rotation process is to effectively trim and fill
keeping the whole a rectangle.

[Axel Berger]

Joe Beanfish wrote:
> Part of the non-90 degree rotation process is to effectively
> trim and fill keeping the whole a rectangle.

Exactly. A task like that is far into the realm of a graphics package
and strictly out of the task a something, whose whole job is to take an
image and embed it.

Axel

[jo...@coherentgraphics.co.uk]

Hi,

On Tuesday, October 8, 2013 11:45:36 PM UTC+1, Fran Jones wrote:
> My HP Laserjet 3200m printer/scanner often scans pages off kilter.
>
> It's noticeable, but slight (maybe a couple of degrees).

> Is there freeware that will rotate a PDF page by just a degree
>
> or three (just like The Gimp rotates an image)?

To rotate three degress clockwise about the centre of each page:

cpdf -rotate-contents 3 in.pdf -o out.pdf

http://community.coherentpdf.com/

Free for non-commercial use.

If your background is white, this ought to look ok. If not, it will be rather obvious,

John Whitington

[Nomen Nescio]

>Perhaps Foxit Reader may fit your needs. It provides PDF creation / PDF
>conversion. Most important----it is free.

To be clear, free as in gratis, not free as in libre.

>Why not have a try with Foxit PDF Reader. It's more fast and safe.

How can you say it's safe? It's proprietary and closed source, and
also comes bundled with some craplets.

The only *safe* option is using the pdfimages+ImageMagik tool chain.

If you mean safe in the sense that it works, perhaps. Certainly the
pdfimages+ImageMagik tool chain fails if the PDF has been OCRd.

[DKra...@lawrenceville.org]

I had the same problem and have been looking for a non-adobe solution for a while. Just tried pdfescape.com -- it works great and no install! Of course, if you have a ton of docs, doing 1 at a time via pdfescape.com is too slow, but for an occasional straightening job, it's awesome.
Fyi -- it's a bit hard to find the "Deskew" button -- it is under Pages >> More

[erde...@gmail.com]

That sites sending a Bad Gateway error for me. www.pdfescape.com right?

[tlvp]

You'd be better off feeding your printer/scanner the document again, trying
to minimize the page-skew of its document feeder :-) . HTH. Cheers, -- tlvp
--
Avant de repondre, jeter la poubelle, SVP.

[Anonymous]

You can do that with LaTeX. E.g. if you have a file named "my.pdf"
and want to simply rotate 37 degrees:

\includegraphics[angle=37]{my}

Note that the pdf extension is not needed. You can do more, for
example, crop, rotate, and scale:

\includegraphics[clip,scale=0.75,angle=-37,viewport=20mm 128mm 200mm 280mm]{my}

[Frank L. Thiel]

GIMP will allow you to import a PDF, rotate it, and export it as a PDF.
Scanning to an image format is not required.

[Axel Berger]

"Frank L. Thiel" wrote:
> Scanning to an image format is not required.

And once more for the record:

You CAN'T scan to PDF. A scanner is a raster machine and what it sees is
and always will be a raster format. You can through a PDF mantle around
it or, seen the other way, include a raster image inside a PDF, but it
will stay a raster format.

What's more important and worse, all those "intelligent" machines (if
you defer to their superior intelligence, what does that say about you?)
outputting a PDF in a single operation generally allow you NO control
whatever about the internal raster format used. Usually that will be a
Jpeg with 16 million different colours at an atrociously low resolution
and a high compression setting losing content and inserting highly
visible artefacts. For 95 % of all pages I scan that is the worst
possible choice of format. I tend to get much better quality at much
smaller filesizes.

THat said, taking the PDF, unpacking all the images, roting as needed,
and repacking again, is one of the simpler tasks batch files were
invented for. Of course "intelligent" machines, that won't allow you
tell them what to do and make you click icons just like pigeons do in
biology labs reduce your capabilities to those of one of those pigeons,
but that seems to be what people want from "user friendly" gadgetry
nowadays.

Axel

[p-0''0-h the cat (UK)]

On Sun, 16 Nov 2014 19:03:33 +0000 (UTC), Dustin wrote:

> I have nothing of value to contribute to this discussion

I agree with every opinion you provided and just wanted to
say, "Damn... When you're right, you're right!"

Poser.

--
p-0.0-h the cat

Internet Terrorist, Mass sock puppeteer, Agent provocateur, Gutter rat,
Devil incarnate, Linux user#666, BaStarD hacker, Resident evil, Monkey Boy,
Certifiable criminal, Spineless cowardly scum, textbook Psychopath,
the SCOURGE, l33t p00h d3 tr0ll, p00h == lam3r, p00h == tr0ll, troll
infâme, the OVERCAT [The BEARPAIR are dead, and we are its
murderers], lowlife troll, shyster [pending approval by STATE_TERROR],
cripple, sociopath, kook, smug prick, smartarse, arsehole, moron, idiot,
imbecile, snittish scumbag, liar, total ******* retard, and shill.

Honorary SHYSTER and FRAUD awarded for services to Haberdashery.
By Appointment to God Frank-Lin.

Signature integrity check
md5 Checksum: be0b2a8c486d83ce7db9a459b26c4896

[Fritz Wuehler]

> You CAN'T scan to PDF. A scanner is a raster machine and what it
> sees is and always will be a raster format.

Some devices are more sophisticated than others. You're talking about
the simplest of scanners.

> You can through a PDF mantle around it or, seen the other way,
> include a raster image inside a PDF, but it will stay a raster
> format.

In principle, there's nothing the stop a scanner from looking for
lines instead of pixels, and composing a vector image. While I'm not
sure if any such thing exists, certainly there are scanners that
derive a raw bitmap just as a first step, but then run it through an
OCR algorithm that results in a collection of text and graphics.

> What's more important and worse, all those "intelligent" machines
> (if you defer to their superior intelligence, what does that say
> about you?) outputting a PDF in a single operation generally allow
> you NO control whatever about the internal raster format
> used.

Actually there are scanners that give copious control over many
variables.

More importantly, secondary utilities like the OCR tool can give
feedback to the light rod such as adjust pixel density to optimize for
that particular OCR algorithm, etc. Making everything a multi-stage
operation forces the human to do the job of a machine, and manually
repeat steps because the feedback loop is severed. If the user sets
the OCR language to Thaana, for example, maybe a different pixel
density works better. While the user should be able to override the
setting, the default should be the most sensible.

Information should go the other direction as well. PDF metadata can
store details about the scan not otherwise possible.

> Usually that will be a Jpeg with 16 million different colours
> at an atrociously low resolution and a high compression setting
> losing content and inserting highly visible artefacts.

Versatile scanners give you the choice. Users should have control
over the quality. It's obviously unreasonable for the scanner to
always produce the highest quality image saved in the raw internal
bitmap for every page, when a user may be scanning a 250 page text
document that then needs to go over the LAN. By the time they do the
final processing, it's too late, the network congestion has already
happened.

> THat said, taking the PDF, unpacking all the images, roting as
> needed, and repacking again, is one of the simpler tasks batch files
> were invented for.

Batch files? Are you talking DOS batch files? How do you unpack the
text that has already been linked to images in the PDF container? I'm
aware of pdftotext, but that loses the linkage, no?

[Axel Berger]

Fritz Wuehler wrote:
> certainly there are scanners that
> derive a raw bitmap just as a first step, but then run it through an
> OCR algorithm that results in a collection of text and graphics.

Which is far worse, because it potentially introduces more errors. Some
adminstrations scan letters and discard the originals and some scanners
do not just make numbers illegible or hard to read but substitute one
clear perfectly legible number for another. If you read German look at

http://www.dkriesel.com/blog/2013/0802_xerox-workcentres_are_switching_written_numbers_when_scanning

If you don't, do so anyway, the example images speak for themselves.
Look for "6" substituted by "8".
Don't rely on machines too much and don't allow them to think for you.

All that said, I do admit that some are better than others or at least
less bad. I get and read a lot of scans and the bad ones predominate.

Axel

[Poutnik]

Dne 11/16/2014 v 10:58 PM Fritz Wuehler napsal(a):

>> You CAN'T scan to PDF. A scanner is a raster machine and what it
>> sees is and always will be a raster format.
>
> Some devices are more sophisticated than others. You're talking about
> the simplest of scanners.
>

It is a pity a DJVU data format, developed especially
for scanned documents, is not widely used.
Scanned texts of same quality are several times smaller,
compared to PDF.

When I once scanned 340 page B/W book with Czech text and some drawings,
old DJVUsolo free SW produced about 1.5 MB small DJVU file
with estimated ratio 2-2.5 bytes per typed character,
with great reading quality, even without OCR.

--
Poutnik

[Poutnik]

Dne 11/17/2014 v 9:01 AM Axel Berger napsal(a):

> Fritz Wuehler wrote:
>> certainly there are scanners that
>> derive a raw bitmap just as a first step, but then run it through an
>> OCR algorithm that results in a collection of text and graphics.
>
> Which is far worse, because it potentially introduces more errors. Some
> adminstrations scan letters and discard the originals and some scanners
> do not just make numbers illegible or hard to read but substitute one
> clear perfectly legible number for another. If you read German look at
>

That is why I prefer well processed DJVU without OCR.
The size is not much bigger ( 2-2.5 bytes/char for B/W text )
and more readable.

--
Poutnik

[Jack Ryan]

> That is why I prefer well processed DJVU without OCR.
> The size is not much bigger ( 2-2.5 bytes/char for B/W text )
> and more readable.

DJVU has the best quality/size ratio possible. But how can you
search/extract/translate text without OCR?

[Poutnik]

Dne 11/17/2014 v 9:09 PM Jack Ryan napsal(a):

I do not say I do. I just read.
If I could , I would have OCR already done.

Some paid DJVU related products include OCR processing
and add OCR processed layer in DJVU documents,
but I do not used them.


--
Poutnik

[Thomas Kaiser]

Axel Berger wrote in <news:5468F435...@B.Maus.De>

> What's more important and worse, all those "intelligent" machines (if
> you defer to their superior intelligence, what does that say about
> you?) outputting a PDF in a single operation generally allow you NO
> control whatever about the internal raster format used.

Fortunately there exist intelligent open source approaches that help in
such situations (capturing/scanning book pages and creating the best
possible output). For example SpreadPi:

http://spreads.readthedocs.org/en/latest/gui.html

Before and after:

http://postimg.org/image/tfyqatxir/
http://postimg.org/image/d2v8hmpsf/

Regards,

Thomas

[ericap...@gmail.com]

On Tuesday, October 8, 2013 at 3:45:36 PM UTC-7, Fran Jones wrote:
> My HP Laserjet 3200m printer/scanner often scans pages off kilter.
> It's noticeable, but slight (maybe a couple of degrees).
>
> If it were a JPEG, I could use The Gimp freeware to straighten
> but it's a PDF. If I must, I could scan to JPEG and then use
> cutePDF freeware or Adobe Acrobat 6.0 Standard payware to then
> print the straightened JPG to PDF - but that seems convoluted.
>
> Is there freeware that will rotate a PDF page by just a degree
> or three (just like The Gimp rotates an image)?

Buy your high quality real or fake passport,(fani...@gmail.com) Counterfeit Bills,Real and Fake Driver's licenses, ID cards, visas, stamps, diploma, certificates, degrees, citizenship and other products for a number of countries like: USA, Australia, Belgium, Brazil, Canada, Italy, Finland, France, Germany, Israel, Russia,Mexico, Finland,Netherlands ,South Africa,Spain,United Kingdom.Japan when producing; magnetic encoded strips and/or scan able bar-code. UV-spectrum analysis test standards,magnetic strip,

Watch video here for more details........... http://vimeo.com/82973635

Contact us............... fani...@gmail.com

Email.......................... fani...@yahoo.com

SKYPE US for quick chat ................. fandena.fandena

SKYPE US for quick chat ................. fandena.fandena

SKYPE US for quick chat ................. fandena.fandena

Contact e-mails: fani...@gmail.com Technical support: fani...@gmail.com

[david....@gmail.com]

On Wednesday, 9 October 2013 11:45:36 UTC+13, Fran Jones wrote:
> My HP Laserjet 3200m printer/scanner often scans pages off kilter.
> It's noticeable, but slight (maybe a couple of degrees).
>
> If it were a JPEG, I could use The Gimp freeware to straighten
> but it's a PDF. If I must, I could scan to JPEG and then use
> cutePDF freeware or Adobe Acrobat 6.0 Standard payware to then
> print the straightened JPG to PDF - but that seems convoluted.
>
> Is there freeware that will rotate a PDF page by just a degree
> or three (just like The Gimp rotates an image)?

One approach is to convert each page to an XObject Form, then render that rotated a few degrees. A Perl solution, using PDF::API2 follows:
====================
#! /usr/bin/perl
use warnings; use strict;
use PDF::API2;
use Getopt::Long;

my $degrees = 3;
my $scale = 1.0;
my $x = 0;
my $y = 0;
GetOptions ("rotate=i" => \$degrees, "scale=f" => \$scale, "x=f" => \$x, "y=f" => \$y)
or die "usage: $0 IN_PDF OUT_PDF --rotate=DEG --scale=ALPHA --x=POINTS --y=POINTS";

my $infile = shift (@ARGV);
my $outfile = shift (@ARGV);

my $pdf_in = PDF::API2->open($infile);
my $pdf_out = PDF::API2->new;

foreach my $pagenum (1 .. $pdf_in->pages) {

my $page_in = $pdf_in->openpage($pagenum);
#
# create a new page
#
my $page_out = $pdf_out->page(0);

my @mbox = $page_in->get_mediabox;
$page_out->mediabox(@mbox);

my $xo = $pdf_out->importPageIntoForm($pdf_in, $pagenum);
#
# lay up the input page in the output page
# note that you can adjust the position and scale, if required
#
my $gfx = $page_out->gfx;
$gfx->rotate($degrees);
$gfx->formimage($xo, $x, $y, $scale);
}

$pdf_out->saveas($outfile);

[Anonymous]

I have to correct myself, now that I have discovered the FOSS tool
called /unpaper/. Unpaper will straighten out crooked documents
automatically, from the commandline.

[azla...@gmail.com]

For People looking for the answer: google "PDF escape", goto the site, upload your pdf, goto the "Page" tab, then "More" and then "deskew" and you use the line to align your skewed document by 10th's of a degree. It worked great for me!

[drak...@gmail.com]

Finereader software will do the trick

[stev...@gmail.com]

On Tuesday, October 8, 2013 at 6:45:36 PM UTC-4, Fran Jones wrote

>
> Is there freeware that will rotate a PDF page by just a degree
> or three (just like The Gimp rotates an image)?

Yes Auto Straightener from sabsoft.com does a great job of straightening pdf files automatically but does not allow the user to rotate by degree.
Make sure to click on pdf straightener in the program because there is 2 options ...scanned page or scanned pdf

[Alankar Mishra]

Yes, There is a free website available where you can easily Rotate page online with paying so try and please share your experience with us.
Website : https://pdfdoctor.com/rotate-pdf

[Arlen Holder]

On Wed, 5 Aug 2020 06:06:13 -0700 (PDT), Alankar Mishra wrote:

>>> Is there freeware that will rotate a PDF page by just a degree
>>> or three (just like The Gimp rotates an image)?
>> Yes Auto Straightener from sabsoft.com does a great job of straightening pdf files automatically but does not allow the user to rotate by degree.
>> Make sure to click on pdf straightener in the program because there is 2 options ...scanned page or scanned pdf
>
> Yes, There is a free website available where you can easily Rotate page online with paying so try and please share your experience with us.
> Website : https://pdfdoctor.com/rotate-pdf

What I do is super simple which is I use Irfanview for everything above.
1. Irfanview freeware is the twain scanner software
2. Irfanview has the F12 menu (which is the "Paint" menu)
3. On that "Paint" menu is a straightener tool

[Arlen Holder]

On Sun, 2 Aug 2020 18:34:04 -0700 (PDT), stev...@gmail.com wrote:

> Auto Straightener from sabsoft.com does a great job
> of straightening pdf files automatically

When I went to <http://sabsoft.com/AutoStraightener.htm> to get the
software, Malwarebytes blocked it as a trojan.

[Amanda Goodyear]

I just used PDFEScape Free Online Tool.
Once you upload the document, they have a tool called "descew" in the bottom right on their tools bar, where you click and drag in the direction of the rotation, and it asks you for the degree of rotation.
Easy to tweak and save.

[Tom Dunbar]

Try cleaning the rollers under the top left side of the printer.

[Jason Boyle]

On Tuesday, October 8, 2013 at 11:45:36 PM UTC+1, Fran Jones wrote:
> My HP Laserjet 3200m printer/scanner often scans pages off kilter.
> It's noticeable, but slight (maybe a couple of degrees).
>
> If it were a JPEG, I could use The Gimp freeware to straighten
> but it's a PDF. If I must, I could scan to JPEG and then use
> cutePDF freeware or Adobe Acrobat 6.0 Standard payware to then
> print the straightened JPG to PDF - but that seems convoluted.
>

> Is there freeware that will rotate a PDF page by just a degree
> or three (just like The Gimp rotates an image)?

It can be done with the popular free image editor "IrfanView" , the program is simple to use & is quickly opened unlike Gimp. Go to the "Image" menu then select "Custom /Fine rotation" . The angle of rotation can be adjusted with the up/down arrows or enter a value direct into the box. For increments of less than 1 degree you need to enter the value direct into the box as the arrows only adjust by a value of 1 whole degree. Use trial and error with the values to get the rotation your happy with, you can then save as an image or a pdf. It's been my go-to app for all simple image editing for a long time.

[Peter Flynn]

On 03/11/2020 13:30, Jason Boyle wrote:
> On Tuesday, October 8, 2013 at 11:45:36 PM UTC+1, Fran Jones wrote:
>> My HP Laserjet 3200m printer/scanner often scans pages off kilter.
>> It's noticeable, but slight (maybe a couple of degrees).
>>
>> If it were a JPEG, I could use The Gimp freeware to straighten
>> but it's a PDF. If I must, I could scan to JPEG and then use
>> cutePDF freeware or Adobe Acrobat 6.0 Standard payware to then
>> print the straightened JPG to PDF - but that seems convoluted.

Not really. I always do it that way (sing GIMP or similar), because I
often need to use the image in multiple circumstances, so a JPEG image
is much more useful. I can use it in a PDF when I need to (like as an
illustration in a document). I've never really understood why a scanner
would want to create a PDF direct, as it's just a wrapper round the JPEG.

P

[Arlen Holder]

On Tue, 17 Nov 2020 10:32:42 +0000, Peter Flynn wrote:

> I've never really understood why a scanner
> would want to create a PDF direct, as it's just a wrapper round the JPEG.

That's an interesting observation, particularly since I scan 8.5x11 pages.
o I use Irfanview to scan pages through the printer document feeder

I typically save that result as a PDF.

Based on what you said above...

I didn't know that the scanner software creates the JPEG first.
o And only after creating the JPEG does it save to PDF.

I figured the scanner software made the image in its own internal format.
o Which, for Irfanview, could be JPEG for all I know (as I don't know).

Does this imply that scanning & then saving as a JPEG results in better
quality than scanning and then saving the document as a PDF?

[Peter Flynn]

On 20/11/2020 07:34, Arlen Holder wrote:
[...]

> I didn't know that the scanner software creates the JPEG first.

> And only after creating the JPEG does it save to PDF.

As far as I understand it, PDF is just a wrapper around many formats,
including JPEG, PNG, etc. The only things which are "native" are fonts,
text, and vector graphics. But that's just the bits I use; PDF can do
lots of other things.

> I figured the scanner software made the image in its own internal format.

I think the scanner's internal firmware does indeed use its own format
(HP hardware has its own internal language, so do other manufacturers).
But what format the data is in during its transfer from scanner to
software I have no idea.

> Which, for Irfanview, could be JPEG for all I know (as I don't know).

Desktop software probably lets you pick (somewhere) what format to save
the file as. I use XSane which does this:
http://www.silmaril.ie/downloads/xsane-settings.png

> Does this imply that scanning & then saving as a JPEG results in better
> quality than scanning and then saving the document as a PDF?

That would really depend on the software settings. I don't know what
compression PDF applies to embedded JPEGs. My software delivers the raw,
uncompressed data for saving, so the files are huge, but every dot is
there. Typically I then need to edit the file to correct alignment,
colour, and crop, so I then save uncompressed if it's for print
publication (full-resolution, best quality, but still huge), or 50%
compression, which is much smaller, lower resolution, but still
perfectly acceptable for other applications.

The real advantages of a PDF are:

• you avoid the small PDF wrapper overhead
• many computer users don't know what to do with a JPEG (weird when
you consider that's what their camera produces)
• you can put multiple images into a single PDF
• you can create a PDF via WP or DTP software so you can have text
and images combined

If you have a workflow that works, I'd stay with it.

Peter

[Phillip Helbig (undress to reply)]

In article <i1pomq...@mid.individual.net>, Peter Flynn

<pe...@silmaril.ie> writes:

> > Does this imply that scanning & then saving as a JPEG results in better
> > quality than scanning and then saving the document as a PDF?
>
> That would really depend on the software settings. I don't know what
> compression PDF applies to embedded JPEGs. My software delivers the raw,
> uncompressed data for saving, so the files are huge, but every dot is
> there.

JPEG files are usually already compressed. Running a compression
algorithm on them again probably won't compress them any more (and might
even make them bigger).

[Peter Flynn]

On 20/11/2020 12:14, Phillip Helbig (undress to reply) wrote:
[...]

> JPEG files are usually already compressed. Running a compression
> algorithm on them again probably won't compress them any more (and might
> even make them bigger).

Right. As I don't always want JPEG (some clients use TIFF, and some
prefer PNG for web images), I set XSane to send the raw data (probably
PBM or something internally).

Peter