nantipov/longscreen

A time-lapse screen recording utility

longscreen's Issues

Follow to the mouse

Idea notes

1. Keep small capturing screen rectangle following to the mouse pointer.

Notes collecting

1. Screen time lapse recording
2. Dictation mode
   a. hot key
   b. record audio
   c. switch time lapse to a normal recording or shrink the interval between frames
3. Smooth (non-staccatto) compilation
4. Background processing
5. Highlight the mouse pointer
   a. hot key
   b. yellow semi-transparent circle
6. Magnifying glass
   a. hot key
   b. normal recoding speed
7. Multiscreen and multi-region
8. Background service and detached CLI util for commands
9. Recording speed switcher
   a. hot key
   b. multi-position switch: 10x, 5x, 2x, 1x
10. OSD
11. Compile to a small GIF (modest palette)
12. Drop a title
    a. reduce speed
13. Templates / project settings in a JSON file
14. Project home option
15. Audio recording beep sound for further sync
16. Stop (Ctrl-C) handler, stop all clips (wait)
17. Capture specific application
18. Put mouse pointer on compile/export-time
    a. Interpolate mouse coordinates for intermediate frames
19. Custom static ffmpeg build with binding

https://github.com/go-vgo/robotgo
https://github.com/gordonklaus/portaudio/blob/master/examples/record.go
https://github.com/golang-standards/project-layout

https://github.com/tfriedel6/canvas
https://github.com/fogleman/gg
https://github.com/llgcode/draw2d

https://stackoverflow.com/questions/24961127/how-to-create-a-video-from-images-with-ffmpeg
https://trac.ffmpeg.org/wiki/Slideshow

https://github.com/markbates/pkger
Switch to the native feature expected in Go16 (Feb2021)

Dependencies

1. sudo apt-get install portaudio19-dev

project
.. region (1..n)
.... display
.... rectangle

help
list - list regions
region (r, reg) - start a new region
interval (i, int) - change interval (often, rare)
audio (a, aud) - audio notes mode
exit, quit, q

project/
.. region-name\
.... images\
...... image-<seq>.bmp
.... audio-notes\
...... audio-note-<seq>.wav
.... tmp
.. exports\
.... region-name-<counter>.mov
.. project.db

Statements and assumptions

1. Simultaneous recording is possible (multi-screen use-case);
2. Audio is a separate track (not a region / clip)

Questions

1. Code organization

ffmpeg concat
# ffmpeg -f concat -safe 0 -i list.txt -c copy c1.ts
file 'file1.ts'
file 'file2.ts'

ffmpeg -i c1-360.m3u_output.ts -map 0 -c copy  c1-360.m3u_output.mp4
ffmpeg -i c2-720.m3u_output.ts -map 0 -c copy  c2-720.m3u_output.mp4

ffmpeg -i c3-720.m3u_output.ts -map 0 -c copy  c3-720.m3u_output.mp4
ffmpeg -i c4-720.m3u_output.ts -map 0 -c copy  c4-720.m3u_output.mp4

Consider not utilizing database during screen capturing

Notes

1. Perhaps database brings additional complexity and might introduce a locking behavior during screen capturing.
2. Check and rebuild solution.
3. Consider filename as an info carrier.
4. Do we need database at all? Or can we use it only during video export/compile?

Auto-dictation

Notes

1. Long lasting audio clip with detection of audio signal volume.
2. Keep audio buffer unsaved
3. If average volume is high enough
   a. store buffer to the disk;
   b. set screen recording to the "real-time" speed;
   c. leave a flag in database.
4. Once average volume came "inactive" back, consider the rare speed for screen recording.