Continuing the discussion started here: viewtopic.php?f=5&t=32873&sid=d6f7b94f771abd3435f7a964cc1e95d6#p100516

For recording the screen (gif or video) I use Sharex: https://getsharex.com/

For displaying mouse clicks I use an Autohotkey script: https://autohotkey.com/boards/viewtopic.php?t=19235

I'll be testing a keystroke display script for autohotkey (not my script): https://gist.github.com/tmplinshi/6933557
(not very pretty/graphical, but it seems to work)



Running on Win 8.1-64.



Please post what you are using, system, resources, and a sample!

On Ubuntu / Linux Mint (and other Linux systems), a great combo is:

Peek: https://github.com/phw/peek (can record mp4, webm and gif, example below is a gif)
key-mon: https://launchpad.net/ubuntu/+source/key-mon/1.17-1ubuntu1 (but it might no longer be developed, can't find the most official repository, there are a couple on github - if anyone knows, please post!)

On win, using sharex for screen capturing.
OBS studio will be a next step further.

Haven't edited any screen captures yet, maybe blender can be used for that?


For key capturing tried two softwares.
Used qiexpress in this video.
Don't like the quick flickering, the absence of the key combinations and the lack of mouse pointer.

This one, called PxKeystrokesForScreencasts provides a much better graphical look however it fails to work with inkscape -disables click-selecting etc.

Moini, I'll definitely have to check out Peek, as I'm on Ubuntu Mate.

For the recording in the other thread that kicked this off, I used RecordMyDesktop (actually GTK-RecordMyDesktop), but that only creates OGV files, which I then had to convert to GIF. I just used an online converter, as the file was fairly small.

I'd already restricted the capture to my Inkscape window, but in an effort to get the file size down even smaller I loaded the GIF into The Gimp, and cropped the image to the relevant area of the application.

Xav, Jabier once gave me a script for recordMyDesktop, which I refined later (it talks ), but I'm not using it anymore, as peek is so much better in usability. It's a transparent window that you put above the area that you want to record.
For my old method, I've actually written a blog post (in 2015): http://vektorrascheln.de/posts/2015/Dec ... ng-of.html

Under Win7, I made good experience so far with ScreenToGif (I guess Windows only)...
https://github.com/NickeManarin/ScreenToGif/
You have a nice editor but the GIFs are not the most compact.

A suitable program to compress the GIF further might be Gifsicle/Giflossy
https://github.com/pornel/giflossy

Gifsicle is cool, too I've tested running gifsicle with the peek output, it didn't reduce file size by much (3,5 Mb -> 3,4 Mb), so peek files are already well optimized. For the files produced by recordMyDesktop, it's really useful, though.

tylerdurden wrote:(not very pretty/graphical, but it seems to work)

If you're talking about the large black strip across the bottom, showing the keys being used, I actually like that very much! It's easy to see. You can catch a glimpse of the key from peripheral vision, and not have to look away from the work going on the canvas. Those ones that put the key in a small box in the corner - I end up having to stop, rewind, replay several times, before I understand what's happening. Oh yes, I like that a lot!!

Well, it could be a little bit smaller, and still be effective. But I wouldn't go more than 10 to 20% smaller.

How do you make those separate scripts work with the video or gif capturing app or program? Do you have to hack the video or gif capture code?

brynn wrote:How do you make those separate scripts work with the video or gif capturing app or program? Do you have to hack the video or gif capture code?

You can buy artificial intelligence hackbots from adobe for that.

Raspi wrote:

brynn wrote:How do you make those separate scripts work with the video or gif capturing app or program? Do you have to hack the video or gif capture code?

You can buy artificial intelligence hackbots from adobe for that.

MMmmmm..... I think I'd be surprised if Xav purchased something, rather than using open source tools. Maybe if there were no good open source solution. But it's hard to think all these nice people who are making video tutorials for inkscape are purchasing something like that, to make nice videos.

Actually someone told me, not too long ago, that Adobe doesn't use artificial intelligence..... Hhmm......

Just came across this video editor:

openshot -might be easier to handle than blender.

It doesn't even start for me (on Xfce Linux Mint 18.2), their tutorial windows can't be closed and keep flashing... If it works, openshot is easy to use, though. I liked it a lot when it still did. It's supposed to also be able to edit gif animations, not just videos, in its latest versions - but I can't even test, and bug reports go unanswered.

Another, libre video editor:

kdenlive

(Still, haven't delved into video editing though.)

Xav or Moini, how do you make the scripts for the keys or mouse work together with the video or animated gif. Do you just have to find one that has them already combined?

brynn wrote:Xav or Moini, how do you make the scripts for the keys or mouse work together with the video or animated gif. Do you just have to find one that has them already combined?

I'm not sure what you mean. If you're referring to the key-mon application I used, I just had to start it running and make sure it was within the capture area when I started recording the screen, nothing special required.

Ooohh, I see. Thanks for explaining.

But wouldn't it be nice if there was a video program that would do both?

$$ Gets you whatever you like.

brynn wrote:But wouldn't it be nice if there was a video program that would do both?

Occasionally computer users of old will still talk about "the Unix philosophy", which is that each program should do one thing, and do it well - and, implicitly, that the operating system should provide ways to link those programs together to create something more than the sum of the parts. It really refers to the way in which command line programs can be chained and scripted, but arguably applies in this case as well. By having two separate programs I have some choice and flexibility in which I use for showing keypresses, and which I use for capturing the screen.

In fact the first keypress program I tried took the "black strip across the screen" approach, which I didn't like at all. Similarly, this thread has suggested other screen capture programs I can try - ones that export natively to GIF, for example. If there was a single program that did both, there's a good chance that the keypress program wouldn't look the way I like, or that the screen capture would lack some useful features.

So, personally, I prefer the flexibility of the Unix philosophy - different programs for different tasks, but which will work together easily. It's really the difference between being given a Lego set, or some pre-fixed moulded toys. Give me Lego any day!

Interesting the differing approaches

Never had much problem with screencaptures in linux, probably because over the years I have always used KDE.

This in Kubuntu 16.04. KDE uses 'activities' as well as 'desktops' This my 'activity' setup for screencapture, only inside the box is recorded using ffmpeg command line. Keymon is no problem either. screenshot https://i.imgur.com/7M6gpAv.jpg



That makes a largish avi file with a keyframe every 12th frame, good for editing in (say) Avidemux, audio added, finished work gets saved to xvid and much compressed.

As you can see, I am no Inkscape expert, more used to Gimp. 1 minute example. https://youtu.be/MzbjsxdCdTI