PDF to PNG: Newbie victory and a question

[ThomasHR]

This is my first post. For the experts reading, this will be deadly dull, and I invite you to skip to the last paragraph. Mainly, I'm writing for other newbies who have landed here after a Google search, and they're desperate for answers to the same problems that I had. It took two full days from the time I downloaded ImageMagick to the time I wrote this post, with much desk-pounding during those two days, but this story has a happy ending.

My problem: Convert a 300-dpi, sRGB-colorspace, PDF 1.3-format file into a 300-dpi PNG file with white background and no degradation of image.

BTW, the PDF file was created by using Scribus desk-top publishing software. Gotta love open-source.

First problem: When I tried to do the conversion, IM gave me error messages that mentioned "gs32c." I have gs32c on my computer because I "print" MS Word files to PDF using PDF995.

First solution: I downloaded and loaded 64-bit Ghostscript (gs64c).

Second problem: I still got the error message about 32-bit Ghostscript. ImageMagick didn't know that I now had the 64-bit version on my computer.

Second solution, part 1: I wrote a little BATCH file that I call in CMD.EXE, before I call "convert":
set PATH=C:\Program Files\gs\gs9.14\bin;%PATH%

Second solution, part 2: I edited DELEGATES.XML, globally replacing
& q u o t ; @PSDelegate@ & q u o t ;
with
& q u o t ; gswin64c & q u o t ;

Third problem: I couldn't save DELEGATES.XML after search-and-replace. The file didn't have its "System," "Hidden," or "Read-Only" attributes set, so how could it refuse to save? Windows insisted that the file was "in use." That was news to me.

Third solution: I shut down my computer. I started it up using the Windows Bootable Diagnostic CD (which takes a while to load, btw). Once I saw a menu, I chose "Command line" from the menu. After trying unsuccessfully to call "edlin" and "edit." I successfully tried "Notepad." Once Notepad had opened up, I loaded DELEGATES.XML into it.

(BTW number 1: When I'm working from my Windows Bootable Diagnostic CD, my C drive is now "D.")

(BTW number 2: When DELEGATES.XML loads in the Diagnostic CD version of Notepad, carriage-returns/linefeeds don't display. They're there, and this version of Notepad saves them afterward, they just don't display. All of DELEGATES.XML takes up only five very long lines in Diagnostic-CD Notepad.)

Anyway, I did the global search-and-replace within DELEGATES.XML that I described above, and saved (successfully). I restarted Windows normally.

For the first time, I could call
convert "big image.pdf" "big image.png"
and not get an error message.

Fourth problem: The background layer of the PDF came out transparent in the PNG file. Higher layers came out properly opaque, and this included parts of upper layers that were white. I tried using
-alpha off
but that not only made the transparent parts turn white, but it made the text look awful.

Fourth solution: I created an intermediary PNG file, then flattened it. The code:
convert -density 300 "big image.pdf" temp_png.png
convert temp_png.png -background white -flatten "big image.png"

MY QUESTION FOR THE EXPERTS: I wrote a BATch file that works, but it takes two lines of code, plus it takes time to write to, and read from, the big PNG file temp_png.png. How do I do what I want in one command, and without creating a temporary file? I can't figure out how to do stuff with parentheses. I have IM 6.8.9-Q16 and Windows 7 64-bit.

[fmw42]

try

Code: Select all

convert -density 300 -background white "big image.pdf" result.png

or

Code: Select all

convert -density 300 "big image.pdf" -background white -flatten result.png

Vector files such as pdf, need settings before reading the input pdf. Raster files need to read the input first then add the settings and operators.
-density and -background are settings but -flatten is an operator.

[ThomasHR]

To fmw42--

The first code results in once again the PDF's background layer showing up as transparent in the PNG.

The second code results in a strange result. I normally view graphics using Graphic Workshop (from Alchemy Mindworks), and this second code results in a PNG file that has a normal thumbnail, and looks normal when viewed in "imdisplay," but which appears totally transparent when Graphic Workshop displays it. I've had Graphic Workshop since 2000, and I've never seen GWS behave this way except for a few ImageMagick-generated PNG files.

[fmw42]

Post your input pdf to some free hosting service such as dropbox.com (public folder) and put the URL here. I would need to have your PDF, so I can check it and try the commands.

[snibgo]

IM sometimes generates a superfluous alpha channel, even when all pixels are opaque. It's worth trying "-alpha off" before writing the output file.

[ThomasHR]

To snibgo: I tried your suggestion.

convert -density 300 "big image.pdf" -background white -flatten -alpha off result.png
^^^^ Results in "transparent" PNG file when viewed in Graphic Workshop

convert -density 300 "big image.pdf" -background white -alpha off -flatten result.png
^^^^ Text is degraded

convert -density 300 "big image.pdf" -alpha off -background white -flatten result.png
^^^^ Text is degraded

[snibgo]

convert -density 300 "big image.pdf" -background white -flatten -alpha off result.png
^^^^ Results in "transparent" PNG file when viewed in Graphic Workshop

Very strange. What version of IM, on what platform?

Can you post your PDF somewhere?

[ThomasHR]

convert -density 300 "big image.pdf" -background white -flatten -alpha off result.png
^^^^ Results in "transparent" PNG file when viewed in Graphic Workshop

Very strange. What version of IM, on what platform?

Can you post your PDF somewhere?

I'll post the PDF later today.

I have IM 6.8.9-Q16 and Windows 7 64-bit.

Graphic Workshop Professional is version 3.0a Patch 43, from 2010.

[snibgo]

Can you ...

Code: Select all

identify -verbose result.png >r.txt

... and paste the contents of r.txt here, between [ code ] and [ /code ]?

[ThomasHR]

Can you ...

Code: Select all

identify -verbose result.png >r.txt

... and paste the contents of r.txt here, between [ code ] and [ /code ]?

You mean for the "transparent" result.png?

Code: Select all

Image: result.png
  Format: PNG (Portable Network Graphics)
  Mime type: image/png
  Class: DirectClass
  Geometry: 3924x2775+0+0
  Resolution: 300x300
  Print size: 13.08x9.25
  Units: Undefined
  Type: TrueColor
  Endianess: Undefined
  Colorspace: sRGB
  Depth: 16-bit
  Channel depth:
    red: 16-bit
    green: 16-bit
    blue: 16-bit
  Channel statistics:
    Red:
      min: 0 (0)
      max: 65535 (1)
      mean: 43270 (0.660259)
      standard deviation: 23254.3 (0.354838)
      kurtosis: -1.3956
      skewness: -0.464055
    Green:
      min: 0 (0)
      max: 65535 (1)
      mean: 43712.8 (0.667014)
      standard deviation: 22085.6 (0.337005)
      kurtosis: -1.24999
      skewness: -0.484068
    Blue:
      min: 0 (0)
      max: 65535 (1)
      mean: 38787.5 (0.59186)
      standard deviation: 24381.6 (0.37204)
      kurtosis: -1.56771
      skewness: -0.148037
  Image statistics:
    Overall:
      min: 0 (0)
      max: 65535 (1)
      mean: 41923.5 (0.639711)
      standard deviation: 23259.4 (0.354916)
      kurtosis: -1.40522
      skewness: -0.371079
  Rendering intent: Perceptual
  Gamma: 0.45455
  Chromaticity:
    red primary: (0.64,0.33)
    green primary: (0.3,0.6)
    blue primary: (0.15,0.06)
    white point: (0.3127,0.329)
  Background color: white
  Border color: srgb(223,223,223)
  Matte color: grey74
  Transparent color: black
  Interlace: None
  Intensity: Undefined
  Compose: Over
  Page geometry: 3924x2775+0+0
  Dispose: Undefined
  Iterations: 0
  Compression: Zip
  Orientation: Undefined
  Properties:
    date:create: 2014-08-12T01:20:49-05:00
    date:modify: 2014-08-12T01:20:55-05:00
    pdf:HiResBoundingBox: 941.666x666+0+0
    pdf:Version: PDF-1.3 
    png:bKGD: chunk was found (see Background color, above)
    png:cHRM: chunk was found (see Chromaticity, above)
    png:gAMA: gamma=0.45454544 (See Gamma, above)
    png:IHDR.bit-depth-orig: 16
    png:IHDR.bit_depth: 16
    png:IHDR.color-type-orig: 2
    png:IHDR.color_type: 2 (Truecolor)
    png:IHDR.interlace_method: 0 (Not interlaced)
    png:IHDR.width,height: 3924, 2775
    png:pHYs: x_res=300, y_res=300, units=0
    png:sRGB: intent=0 (Perceptual Intent)
    png:text: 4 tEXt/zTXt/iTXt chunks were found
    signature: 8a9398f71fff9b2ca644f8b6126db493dc7b324d247d569f4e3e5684bd95db96
  Artifacts:
    filename: result.png
    verbose: true
  Tainted: False
  Filesize: 9.592MB
  Number pixels: 10.89M
  Pixels per second: 10.81MB
  User time: 0.796u
  Elapsed time: 0:02.006
  Version: ImageMagick 6.8.9-6 Q16 x64 2014-07-22 http://www.imagemagick.org

[snibgo]

Yes, that's what I meant. I can't see anything in there that would make software think it should be transparent. It does have "pdf:HiResBoundingBox: 941.666x666+0+0". I don't know what that means.

[ThomasHR]

Just for comparison, here are the numbers for "big image.png" (see original post), which meets all specifications, including being viewable by Graphic Workshop Professional:

Code: Select all

Image: big image.png
  Format: PNG (Portable Network Graphics)
  Mime type: image/png
  Class: DirectClass
  Geometry: 3924x2775+0+0
  Resolution: 300x300
  Print size: 13.08x9.25
  Units: Undefined
  Type: TrueColorAlpha
  Endianess: Undefined
  Colorspace: sRGB
  Depth: 8-bit
  Channel depth:
    red: 8-bit
    green: 8-bit
    blue: 8-bit
    alpha: 1-bit
  Channel statistics:
    Red:
      min: 0 (0)
      max: 255 (1)
      mean: 168.366 (0.660258)
      standard deviation: 90.4836 (0.354838)
      kurtosis: -1.3956
      skewness: -0.464054
    Green:
      min: 0 (0)
      max: 255 (1)
      mean: 170.089 (0.667014)
      standard deviation: 85.9362 (0.337005)
      kurtosis: -1.24999
      skewness: -0.484066
    Blue:
      min: 0 (0)
      max: 255 (1)
      mean: 150.924 (0.591859)
      standard deviation: 94.8702 (0.37204)
      kurtosis: -1.56771
      skewness: -0.148033
    Alpha:
      min: 255 (1)
      max: 255 (1)
      mean: 255 (1)
      standard deviation: 0 (0)
      kurtosis: 0
      skewness: 0
  Image statistics:
    Overall:
      min: 0 (0)
      max: 255 (1)
      mean: 122.345 (0.479783)
      standard deviation: 78.3783 (0.307366)
      kurtosis: 1.41301
      skewness: 0.301559
  Rendering intent: Perceptual
  Gamma: 0.45455
  Chromaticity:
    red primary: (0.64,0.33)
    green primary: (0.3,0.6)
    blue primary: (0.15,0.06)
    white point: (0.3127,0.329)
  Background color: white
  Border color: srgba(223,223,223,1)
  Matte color: grey74
  Transparent color: none
  Interlace: None
  Intensity: Undefined
  Compose: Over
  Page geometry: 3924x2775+0+0
  Dispose: Undefined
  Iterations: 0
  Compression: Zip
  Orientation: Undefined
  Properties:
    date:create: 2014-08-12T04:11:47-05:00
    date:modify: 2014-08-12T04:11:52-05:00
    pdf:HiResBoundingBox: 941.666x666+0+0
    pdf:Version: PDF-1.3 
    png:bKGD: chunk was found (see Background color, above)
    png:cHRM: chunk was found (see Chromaticity, above)
    png:gAMA: gamma=0.45454544 (See Gamma, above)
    png:IHDR.bit-depth-orig: 8
    png:IHDR.bit_depth: 8
    png:IHDR.color-type-orig: 6
    png:IHDR.color_type: 6 (RGBA)
    png:IHDR.interlace_method: 0 (Not interlaced)
    png:IHDR.width,height: 3924, 2775
    png:pHYs: x_res=300, y_res=300, units=0
    png:sRGB: intent=0 (Perceptual Intent)
    png:text: 4 tEXt/zTXt/iTXt chunks were found
    signature: 8eca60fa3268f91caac84db3a0049962de603c851712d04739afc8ed5833d7a9
  Artifacts:
    filename: big image.png
    verbose: true
  Tainted: False
  Filesize: 7.844MB
  Number pixels: 10.89M
  Pixels per second: 19.04MB
  User time: 0.577u
  Elapsed time: 0:01.572
  Version: ImageMagick 6.8.9-6 Q16 x64 2014-07-22 http://www.imagemagick.org

[fmw42]

It would be best to post your image so we can experiment with it to try to duplicate your bad results.

[glennrp]

The "verbose" outputs show that your "transparent" PNG is color-type 2, and does not have
a tRNS chunk, so it's opaque. But it has 16-bit samples, so maybe GWS has trouble
reading 16-bit PNGs. The other output is 8-bit RGBA, which could be transparent but
only if any of the pixels have a non-255 alpha sample.

You could try adding "-depth 8" as the final option while writing the "transparent" one,
and see if the 8-bit result gets read properly.

[ThomasHR]

Here's my PDF:

http://tomhrichardson.files.wordpress.c ... bcover.pdf

I'll keep it up for a week (through August 19).