Knowledge Base (Posts about convert)

Convert pdf to jpg

bgstack15 — Wed, 18 Dec 2024 13:58:00 GMT

Quick and dirty, but way better than convert input.pdf output.jpg. I'm assuming there was a resolution setting somewhere for "convert" that I never bothered to research.

pdftoppm -jpeg -r 300 input.pdf output

Specifically omit the extension, because you will get files like output-1.jpg, output-2.jpg, et al.: one for each page.

References

Ripped shamelessly from:

How to convert a PDF into JPG with command line in Linux? - Stack Overflow

Crop screenshots to same window area

bgstack15 — Fri, 11 Oct 2024 12:46:00 GMT

I had problems getting exact screenshots of some work, and was merely able to get full desktop screenshots. So I needed to crop to just the relevant portion. In my case, I wanted an 800x600 picture, from my second monitor (first being the left monitor, and second being the right monitor), starting at 1920+560 pixels by 240 pixels.

for word in *png ; do convert "${word}" -crop 800x600+2480+240 foo/${word} ; done

Really small, and dumb, but I'll need this again someday.

References

Cropping Images using Command Line Tools Only - Ask Ubuntu

Convert .dng to .jpg

bgstack15 — Sat, 18 Jun 2022 13:04:03 GMT

I had a batch of .dng files that should be converted to .jpg format. I tried a few ways, but they did not preserve all the exif metadata. My process documented below does.

Use Irfanview to batch convert all .dng files to .jpg. This will strip a lot of the metadata.
Add in metadata from original file.

time for word in *dng ; do exiftool -all= -tagsfromfile "${word}" -exif:all "${word%%.dng}.jpg" ; done

This command syntax is verbatim from the man page for exiftool.

Apply original timestamp.

for word in *dng ; do touch --no-create --reference "${word}" "${word%%dng}jpg" ; done
Optionall, remove the original files.

rm $( for word in *.jpg ; do echo "${word%%.jpg}.dng" ; done ; )

References

Man pages

exiftool(1p)

Converting FCD to ISO using the original software

bgstack15 — Mon, 14 Feb 2022 14:23:24 GMT

On my main file server, I store an old number of files in a format last used on the pre-NT kernel Windows platform. Nowadays for compact disc images I would use the iso format. But this nifty piece of software from the past, IMSI CD Copier Pro, used its own format, .FCD. It worked great at the time!

$ ls -al *FCD
-rw-rw-r--. 1 bgstack15 bgstack15 209218511 May 19  2001 Math Blaster 3.FCD
-rw-rw-r--. 1 bgstack15 bgstack15 402961457 May 26  2001 Math Blaster 4.FCD
-rw-rw-r--. 1 bgstack15 bgstack15 258842982 May 22  2001 Math Blaster 5.FCD
-rw-rw-r--. 1 bgstack15 bgstack15  55161301 May  4  2002 Math Blaster 6.FCD

With my recent Windows 98 SE virtual machine, I can now load up the CD image files and get the sweet, sweet files for these educational computer games.

I was unable to get any nfs or SMB mounts to work in my vm, so I ended up just using some virtualized CD drives, and building new iso files that contain the FCD files which the VM will then read and re-mount in yet another virtualized CD drive. This sounds like a meme...

My first step was to install CD Copier Pro. I have the installer image as an ISO, and my original license key. I failed to take screenshots for this step, but here it is running on the vm.

Now, to get the .FCD files into the VM with smb or nfs, I just make a new iso file and mount it in the VM. I made the iso file:

time mkisofs -J -rock -V 'MB3fcd' -o /mnt/public/Support/SetupsBig/Windows/mb5-fcd.iso Math\ Blaster\ 3.FCD

And then configure the VM to use that file. That whole path was where I had already configured libvirt to use as a storage pool.

So on the hot-updated CD drive in the vm, you can now see the FCD file. I just draged it into the main area of CD Copier Pro, and now it is available as an option. I then dragged it from this main area, to the H: drive in the left-side area of the window. And now H:\ is available in Windows Explorer.

I set up a directory on C:\ to hold the CD contents. I then copied the contents for each FCD I was interested in.

Once I got all the FCD files' contents to C:, I was ready to mount up the guest OS (I shut it down first). On the kvm host, I ran this command.

$ guestmount -a /var/lib/libvirt/images/win98-01a.qcow2 -m /dev/sda1 /mnt/foo
$ ls -al /mnt/foo/CDROMs
total 80
drwxr-xr-x.  5 root root 16384 Feb  7 10:29 .
drwxr-xr-x.  6 root root 16384 Dec 31  1969 ..
drwxr-xr-x. 11 root root 16384 Feb  7 10:29 MB3rdGr
drwxr-xr-x. 13 root root 16384 Feb  7 10:49 MB4thGr
drwxr-xr-x. 12 root root 16384 Feb  7 11:14 MB5thGr

And now I make a new iso file, from these recently extracted contents.

cd /mnt/foo/CDROMs
time mkisofs -J -rock -V 'MB3rdGr' -o /mnt/public/CDROMs/Games/Math\ Blaster/MB3rdGr.iso MB3rdGr

And now I can go try to run Math Blaster 5 in Wine!

References

Weblinks

https://www.xmodulo.com/mount-qcow2-disk-image-linux.html

My scanpdf solution

bgstack15 — Tue, 25 Jan 2022 14:15:58 GMT

I have not personally needed to forcibly scan a document, i.e., print one out, just to sign it, and then scan it in again. I have my trusty signatures.pdf that I scanned in a few years ago with 10 iterations of my signature. I always just screenshot a signature, and use LibreOffice Draw's ability to set a transparent color of an imported image.

But, after reading an interesting Hacker News discussion, I have whipped up a "scanpdf" solution. 99% of this is ripped straight from the author, mrb.

Here is my file ~/bin/scanpdf.sh, with typo/semantic fixes.

	`#!/bin/bash`
	`# from https://news.ycombinator.com/item?id=30024165#30027344`
	`# Make a pdf look like it was scanned.`

	`if [ $# -ne 2 ]; then`
	`echo "Usage: $0 input output" >&2`
	`exit 1`
	`fi`
	`tmp="$1".scanner-look.tmp`
	`mkdir "$tmp" &&`
	`# without -flatten some PDF convert to a JPG with a black background`
	`convert -density 150 "$1" -colorspace Gray -quality 60 -alpha flatten "$tmp"/p_in.jpg &&`
	`: \|\| exit 1`
	`# each page is randomly shifted in the X and Y plane.`
	`# units seem to depend on angle of rotation in ScaleRotateTranslate?`
	`offset() { echo $(($RANDOM % 1000)); }`
	`for f in "$tmp"/p_in*jpg; do`
	`# each page is randomly rotated by [-0.5 .. 0.5[ degrees`
	`angle=$(python3 -c 'import random; print(random.random()-0.5)')`
	`x=$(offset)`
	`y=$(offset)`
	`convert "$f" \`
	`-blur 0x0.5 \`
	`-distort ScaleRotateTranslate "$x,$y $angle" +repage \`
	`$ +clone +noise Random -fill white -colorize 95% $ \`
	`-compose darken \`
	`-composite \`
	`${f/p_in/p_out}.pdf \|\| exit 1`
	`done`
	`# concatenate all the pages to one PDF`
	`# use "ls -v" to order files correctly (p_out-X.jpg where X is 0 1 2 ... 9 10 11 ...)`
	`pdftk $(ls -v "$tmp"/p_out*.pdf) cat output "$2" &&`
	`rm -rf "$tmp"`

Dependencies include pkdftk(1) and ImageMagick. Also important is to make sure the ImageMagick policy had read and write acess to PDFs. Ensure something similar to this logic is in file /etc/ImageMagick-6/policy.xml.

<policy domain="coder" rights="read | write" pattern="PDF" />

References

Weblinks

Convert b6i to iso

bgstack15 — Fri, 01 Oct 2021 13:14:27 GMT

Afte some Internet searching, I came across cdemu for GNU/Linux that can convert the files the way I want. I use Devuan Ceres (the unstable release), so I wanted to rebuild the packages for my exact libs.

So I used the OBS to build them: subproject cdemu. I had to borrow one hack that the Ubuntu team came up with, to suit the needs of systemd-free Devuan.

For my headless VM where I was running these tasks, I had to start a virtual X server. In a GNU screen:

Xvfb :1

In another screen, run:

export DISPLAY=:1
sudo modprobe vhba
sudo chmod 0666 /dev/vhba_ctl
cdemu-daemon &

The daemon insists on using X11, so I had to do it this way. But now you can use client commands. I loaded my .b6t file and then listed it to make sure it was available to my system.

cdemu load 0 /path/to/file.b6t
cdemu device-mapping

The output should resemble:

$ cdemu device-mapping
Device mapping:
DEV   SCSI CD-ROM     SCSI generic
0     /dev/sr1        /dev/sg2

And now you can mount the contents, and then write a new iso file.

sudo mount -v /dev/sr1 /mnt/cdrom
DISKLABEL="$( sudo file -s /dev/sr1 | awk -F\' '{print $2}' )" ; echo "${DISKLABEL}"
mkisofs -V "${DISKLABEL}" -J -rock -o ~/disc.iso /mnt/cdrom

Convert vhs video tape to mkv file in Linux

bgstack15 — Wed, 17 Jul 2019 13:05:22 GMT

Last updated 2019-09-11

I still own video tapes (VHS) and I have finally gotten around to getting a nice setup for saving them down to hard disk so I can clear out the old video tape collection. This is how I did it. First of all, I knew I needed an RCA- to-something adapter. I found an "AV to USB converter" [amazon.com] that does the trick. When attached to a Ubuntu 16.04 system, it shows up as /dev/video0. No additional steps required! To quote one of my role models [thealaskalinuxuser.com] on the Internet: "Linux -- keep it simple." Now, I discovered over time that the audio input just didn't work. But the video comes in perfectly! I was using vlc [videolan.org] for the input capture. To get the audio to work, I had to find an RCA to mini-phono cable which I already owned. An example product is on monoprice.com [monoprice.com]. I also had to fiddle with the audio input settings to get the input to work right. I had to poke around with pulseaudio, alsa, and vlc to get the audio input to work the way I needed it. For my setup (pictured below), I needed to make sure my input source was mic (the front jack on my desktop computer) and not line, which is probably the input in the back. I chose to set my mic boost (akin to gain) really low, and capture volume really high. I needed to set pulseaudio to allow the stereo input. I have discovered that pulseaudio is fragile, and also never remembers the settings I choose for it, so for every reboot or login or pulseaudio --start command, I have to go readjust the "profile." I am starting to understand why people smarter (and more verbose) than me tend to dislike pulseaudio. With alsa and pulse set up correctly, it is now a matter of getting vlc to accept the audio input while accepting the video input. I had to discover how my system identifies the audio jack.

$ arecord -l
**** List of CAPTURE Hardware Devices ****
card 0: PCH [HDA Intel PCH], device 0: CX20641 Analog [CX20641 Analog]
  Subdevices: 0/1
  Subdevice #0: subdevice #0
card 0: PCH [HDA Intel PCH], device 2: CX20641 Alt Analog [CX20641 Alt Analog]
  Subdevices: 1/1
  Subdevice #0: subdevice #0
card 1: USB20 [AV TO USB2.0], device 0: USB Audio [USB Audio]
  Subdevices: 1/1
  Subdevice #0: subdevice #0

By trial and error, I learned the mic input is card 0, device 2. In alsa terms, that's alsa://0,2. So, to put it all together, I had to set VLC to open a "capture device" of /dev/video0, with audio device of hw:0,2. To get the audio and video to both be output by VLC, I had to go into advanced optiosn and se audio input = 0. VLC always defaults to audio input=-1, so no audio will come through. When I selected the Play button, it presents both audio and video from the VCR! So, with the video on the screen that I want, I pressed the record button (from the toolbar visible when you select View -> Advanced controls). VLC records by default to avi format, which is not my preferred method, even though it is a temporary format. So, I had to set all the settings again, and then select the drop-down arrow beside the Play button, and select "Convert." A new window appears, and I select the file format I want and the output file. With all the options set, I now return to the first options screen and select the "Convert" button. After running through the whole video, I come back and hit stop in VLC. I use HandBrake [handbrake.fr] to convert the file to the format I want. The details of that are beyond the scope of this post. For a bonus, I discovered that killing vlc safely closes the written file, so it is still viewable. So I just run a command in a terminal to kill vlc after the published length of the videotape:

$ sleep $(( 60 * 120 )) ; killall vlc

I could have done something fancier with xdotool to select the "stop" button, but why bother?

Addendum

For some reason, when I capture video (possibly because of the NTSC framerate conversation), the audio gets a little bit out of sync with the video. Thankfully, there is a tool that can change the scale of the audio to video: mkvmerge, which is a part of mkvtoolnix suite. After much experimentation, I am satisfied that a safe number to use to scale the audio perfectly from my VCR for all content is in this command:

mkvmerge -y '1:0,100.060552/100' -o new.mkv old.mkv

That is take stream 1 (zero-indexed, and usually the audio in a simple captured video file) and stretch it with an anchor point of zero to 100.060552% of the original speed. So it slows down the audio, and then the audio syncs perfectly with the video, for a 2-hour movie.

Extract icons from .exe files

bgstack15 — Tue, 24 Jul 2018 13:17:10 GMT

In this cross-platform world, sometimes you just really need an icon from a .exe binary. Thanks to the StackOverflow community, here is the definitive method to do that. Install icoutils.

dnf install icoutils
# OR
sudo apt-get install icoutils

Use resource 14 (icon groups) to extract the icons.

wrestool -x -t 14 mybinary.exe -o outputdir/