- Download youtube video
- Download youtube video and subtitles
- Download youtube audio
- Hardcode subtitles into video
- Download a specific format from youtube video
- Convert GIF to MP4
- Unsorted ffmpeg-fu
- Adjust webcam settings
- DSLR Video Webcam
- Virtual webcam from OBS output
- Multiplex video devices
- Teleprompter
- Video production tips (Not Linux related)
youtube-dl URL-VIDEO
Slow downloads? Try
youtube-dl --format bestURL-VIDEO
youtube-dl --write-auto-sub URL-VIDEO
youtube-dl -x --audio-format mp3 URL-VIDEO
ffmpeg -i VIDEO-FILE -vf subtitles=SUBS-FILE OUTPUT-FILE
First check the available formats
youtube-dl -F GKgfCthuiV0 where GKgfCthuiV0 is the youtube code of the video
[youtube] GKgfCthuiV0: Downloading webpage
[youtube] GKgfCthuiV0: Downloading video info webpage
[info] Available formats for GKgfCthuiV0:
format code extension resolution note
249 webm audio only tiny 69k , opus @ 50k (48000Hz), 3.38MiB
250 webm audio only tiny 87k , opus @ 70k (48000Hz), 4.29MiB
140 m4a audio only tiny 130k , m4a_dash container, mp4a.40.2@128k (44100Hz), 6.95MiB
251 webm audio only tiny 157k , opus @160k (48000Hz), 7.91MiB
160 mp4 192x144 144p 13k , avc1.4d400b, 25fps, video only, 436.59KiB
133 mp4 320x240 240p 18k , avc1.4d400d, 25fps, video only, 528.49KiB
278 webm 192x144 144p 27k , webm container, vp9, 25fps, video only, 755.36KiB
394 mp4 192x144 144p 30k , av01.0.00M.08, 25fps, video only, 892.15KiB
395 mp4 320x240 240p 47k , av01.0.00M.08, 25fps, video only, 1.17MiB
242 webm 320x240 240p 48k , vp9, 25fps, video only, 1.13MiB
18 mp4 400x300 240p 147k , avc1.42001E, mp4a.40.2@ 96k (44100Hz), 7.90MiB (best)Select the one you want, in my case 251, which is the highest audio quality
youtube-dl -f 251 GKgfCthuiV0
Optional: transcode the webm file into wav (or any other) format you want
ffmpeg -i file.webm file.wav
ffmpeg -i animated.gif -movflags faststart -pix_fmt yuv420p -vf "scale=trunc(iw/2)*2:trunc(ih/2)*2" video.mp4
Explanation
movflags– This option optimizes the structure of the MP4 file so the browser can load it as quickly as possible.pix_fmt– MP4 videos store pixels in different formats. We include this option to specify a specific format which has maximum compatibility across all browsers.vf– MP4 videos using H.264 need to have a dimensions that are divisible by 2. This option ensures that’s the case.
- Neil's variable bit rate 1080p MP3:
ffmpeg -i input_video -vcodec libx264 -crf 25 -preset medium -vf scale=-2:1080 -acodec libmp3lame -q:a 4 -ar 48000 -ac 2 output_video.mp4 - Neil's no audio:
ffmpeg -i input_video -vcodec libx264 -b:v 1000k -vf scale=-2:1080 -an output_video.mp4 - Use nvidia hw encoder
-vcodec h264_nvenc - Convert single image to a 5s video
ffmpeg -loop 1 -i image.png -t 5 output.mp4or batch convert `` - Concatenate videos
ffmpeg -f concat -safe 0 -i videos.txt -c copy output.mp4Wherevideos.txtis:
file '/path/to/file1'
file '/path/to/file2'
file '/path/to/file3'
Check webcams:
[unix ~]$ v4l2-ctl --list-devices
Integrated Camera: Integrated C (usb-0000:00:14.0-8):
/dev/video0
/dev/video1
/dev/media0Check available settings
[unix ~]$ v4l2-ctl -d /dev/video0 --list-ctrls
brightness 0x00980900 (int) : min=0 max=255 step=1 de
contrast 0x00980901 (int) : min=0 max=255 step=1 de
saturation 0x00980902 (int) : min=0 max=100 step=1 de
hue 0x00980903 (int) : min=-180 max=180 step=1
white_balance_temperature_auto 0x0098090c (bool) : default=1 value=1
gamma 0x00980910 (int) : min=90 max=150 step=1 d
power_line_frequency 0x00980918 (menu) : min=0 max=2 default=1 v
white_balance_temperature 0x0098091a (int) : min=2800 max=6500 step=e
sharpness 0x0098091b (int) : min=0 max=7 step=1 defa
backlight_compensation 0x0098091c (int) : min=0 max=2 step=1 defa
exposure_auto 0x009a0901 (menu) : min=0 max=3 default=3 v
exposure_absolute 0x009a0902 (int) : min=4 max=1250 step=1 d
exposure_auto_priority 0x009a0903 (bool) : default=0 value=0Set a parameter v4l2-ctl -d /dev/video0 --set-ctrl=exposure_auto=3
gphoto2 --auto-detect
output:
Model Port
----------------------------------------------------------
Canon EOS M6 usb:001,007 gphoto2 --summary
output:
Camera summary:
Manufacturer: Canon Inc.
Model: Canon EOS M6
Version: 3-15.0.1.0
Serial Number: 37B3BF29137941D7AA01C88F4B5CCA22
Vendor Extension ID: 0xb (1.0)
Vendor Extension Description:
Capture Formats: JPEG
Display Formats: Association/Directory, Script, DPOF, MS Wave, JPEG, Defined Type, Unknown(b103), Unknown(b982), Unknown(b105), Unknown(bf01)
Device Capabilities:
File Download, File Deletion, File Upload
No Image Capture, No Open Capture, Canon EOS Capture 2
Canon Wifi support
Storage Devices Summary:
store_00010001:
StorageDescription: SD
VolumeLabel:
Storage Type: Removable RAM (memory card)
Filesystemtype: Digital Camera Layout (DCIM)
Access Capability: Read-Write
Maximum Capability: 7939850240 (7572 MB)
Free Space (Bytes): 7939784704 (7571 MB)
Free Space (Images): -1
Device Property Summary:
Event Emulate Mode(0xd045):(readwrite) (type=0x4) Enumeration [1,2,3,4,5,6,7] value: 2
Property 0xd04a:(readwrite) (type=0x2) Enumeration [0,1,2,3] value: 0
Size of Output Data from Camera(0xd02e):(read only) (type=0x6) 524288
Size of Input Data to Camera(0xd02f):(read only) (type=0x6) 524288
Battery Level(0x5001):(read only) (type=0x2) Enumeration [0,1,2,3] value: 3% (3)
Battery Type(0xd002):(read only) (type=0x4) Enumeration [0,1,2,3,4,5] value: Unknown (0)
Battery Mode(0xd003):(read only) (type=0x6) Enumeration [0,1,2,3] value: Normal (1)
UNIX Time(0xd034):(readwrite) (type=0x6) 1640587702
Type of Slideshow(0xd047):(read only) (type=0x4) 0
DPOF Version(0xd046):(read only) (type=0x4) 257
Remote API Version(0xd030):(read only) (type=0x6) 256
Model ID(0xd049):(read only) (type=0x6) 67567616
Camera Model(0xd032):(read only) (type=0xffff) 'Canon EOS M6'
Camera Owner(0xd033): error 201b on query.
Firmware Version(0xd031):(read only) (type=0x6) 16777216
Property 0xd050:(read only) (type=0x2) 0
Property 0xd051: error 201b on query.
Property 0xd052:(read only) (type=0x2) 0
Property 0xd053: error 201b on query.
Property 0xd054: error 201b on query.
Property 0xd057:(read only) (type=0xffff) '478251000023'
Property 0xd402:(read only) (type=0xffff) 'Canon EOS M6'
Property 0xd406:(readwrite) (type=0xffff) 'Windows'
Property 0xd407:(read only) (type=0x6) 1
Property 0xd303:(read only) (type=0x2) 1gphoto2 --stdout --capture-movie | ffmpeg -i - -vcodec rawvideo -pix_fmt yuv420p -threads 0 -f v4l2 /dev/video30
You need a capture card to get the HDMI out signal from the DSLR. I bought a cheap HDMI to USB3 capture card in Amazon that works with the V4L2 module without drivers, just plug and play. The card is unbranded and the chipset inside is:
[66367.259465] usb 2-2: Product: FHD Capture
[66367.259468] usb 2-2: Manufacturer: VXIS Inc
[66367.269584] uvcvideo: Found UVC 1.00 device FHD Capture (1bcf:2c99)The DSLR camera will become now a Webcam. You can even select it in your videoconferences.
It is convenient that you also get a AC battery adapter to make sure the battery won't die during capture.
You can capture video with pretty much anything but I mostly use OBS Studio. It allows you to create scenes, overlays, green screen, etc.
The Canon 7D has a mini HDMI port out. With the Canon firmware 2.0.3 I cannot obtain a clean HDMI out. There is always the focus rectangle there. So I downloaded Magic Lantern Firmware. To control the camera settings I use Entangle which allows me to capture images as well as control the camera settings.
Canon M6 has a micro HDMI port. With the Canon M6 I can obtain a clean HDMI out in manual focus. But controlling the camera settings is just annoying. USB tether does not seem to work and there is a Canon App which is so faulty.
Magic Lantern is not available for the M6. It could run CHDK but at the moment of writing the firmware is still unported.
The camera turns black every 30 minutes which is annoying. To override this I followed the instructions found in CHDK forum
- Prepare the SD card, format it low level in the camera or as FAT32 (not exFAT!). Save this script as
makeCanonSDCard.sh
#!/bin/bash
#Enable powershot basic scripting on a memory card
if [ $# -ne 1 ] ; then
echo
echo "Usage : ./makeCanonSDCard.sh [ device ]"
echo
echo " [ device ] is a fat32 / fat16 partition on the memory card"
echo " example : ./makeCanonSDCard.sh /dev/sdb1"
echo "NOTE: please run as root"
exit 112
fi
echo Boot sector of $1 will be modified. If you are not sure this is what you want then cancel with Ctrl-C
sleep 8
#TAG on boot sector
if ! echo -n SCRIPT | dd bs=1 count=6 seek=496 of=$1 ; then {
echo failed writing to boot sector
exit 113
} fi
if mount | grep /mnt ; then {
umount /mnt
} fi
if ! mount $1 /mnt ; then {
echo failed to mount
exit 114
} fi
#create script request file
echo -n "for DC_scriptdisk" > /mnt/script.req
#Canon M6 script
echo 'DIM palette_buffer_ptr = 0x11d4c
DIM active_palette_buffer = 0x11d44
DIM palette_to_zero = 0
private sub RegisterProcs()
System.Create()
ExecuteEventProcedure("UI.CreatePublic")
end sub
private sub Initialize()
RegisterProcs()
LockMainPower()
adr = *palette_buffer_ptr
adr = adr + (palette_to_zero * 4)
if *adr <> 0 then
adr = *adr + 4
memset(adr, 0, 256 * 4)
end if
end sub
'>/mnt/extend.m
#Done !
echo "Please check /mnt for files extend.m and script.req"
This includes saving the following script as extend.m in the root directory of the card.
DIM palette_buffer_ptr = 0x11d4c
DIM active_palette_buffer = 0x11d44
DIM palette_to_zero = 0
private sub RegisterProcs()
System.Create()
ExecuteEventProcedure("UI.CreatePublic")
end sub
private sub Initialize()
RegisterProcs()
LockMainPower()
adr = *palette_buffer_ptr
adr = adr + (palette_to_zero * 4)
if *adr <> 0 then
adr = *adr + 4
memset(adr, 0, 256 * 4)
end if
end sub
-
Check the sdcard devide with
lsblkand run the script as a superuser. SERIOUS WARNING!! MAKE SURE YOU SPECIFY THE CORRECT SD CARD DEVICE!!! If you don't know what I am talking about don't do it. -
Every time you start the camera you need will see the overlays. Enter PLAYBACK mode and do a single press of SET. This will cause the script execution. Go back to shooting mode, if the overlays are not present, the display will stay ON.
-
Do not format the card in low level or you will have to repeat steps 1 and 2 again.
There is a way. Tricky, not trivial, to pipe the output of OBS studio into a virtual webcam that you can use in a videoconference (for instance). Here's how I do it.
First clone and install v4l2loopback from here https://github.com/umlaeute/v4l2loopback. You can load the kernel module like this:
sudo modprobe v4l2loopback video_nr=9
This will create a virtual video device /dev/video9. You can also create more than one virtual camera and label them:
sudo modprobe v4l2loopback devices=2 video_nr=10,20 card_label=TimerCam,StreamCam
When you want to remove the device use sudo modprobe -r v4l2loopback or sudo rmprobe v4l2loopback. Finally remember to rebuild the kernel module every time you update the kernel!
Note: Now obs will do the modprobe as long as the module is compiled for the kernel (reinstall the module after rebooting in a new kernel)
I installed the AUR package Now OBS includes a obs-v4l2sink which is actually a sink where OBS will pour the virtual webcam. Then in OBS top menu select tools/v4l2sink and choose the video device /dev/video9 that you activated before and YUV12 format.Start Virtual Camera button that does just that.
You will be able now to use a new webcam that will appear (tested in zoom and Firefox).
I asume here that /dev/video0 is the webcam. Create a loopcam as seen before
sudo modprobe v4l2loopback video_nr=10 card_label=COPYWebcam
Now examine the format of your original webcam
ffmpeg -f v4l2 -list_formats all -i /dev/video0 which will display something similar to
[video4linux2,v4l2 @ 0x559d808150c0] Raw : uyvy422 : UYVY 4:2:2 : 1920x1080 1280x720 1024x768 640x480 320x240
[video4linux2,v4l2 @ 0x559d808150c0] Compressed: mjpeg : Motion-JPEG : 1920x1080 1280x720 1024x768 640x480 320x240Now dump the stream of the original webcam into the COPYWebcam device
ffmpeg -i /dev/video0 -f v4l2 -codec:v rawvideo -pix_fmt uyvy422 /dev/video10
This is useful in the case you start a virtual cam in OBS and you want to use the virtual camera in ZOOM and a website. Create 2 virtual cameras
sudo modprobe -r v4l2loopback && sudo modprobe v4l2loopback devices=2 video_nr=10,20 card_label=ZOOMCam,NINJAcam
and start virtual camera in ZOOM. Then list the properties of the first virtual camera
ffmpeg -f v4l2 -list_formats all -i /dev/video10 which will show properties like this one
[video4linux2,v4l2 @ 0x56517d3c70c0] Raw : yuyv422 : YUYV 4:2:2 : 1280x720
And now dump the contents of the first virtual camera into the second virtual camera
ffmpeg -i /dev/video10 -f v4l2 -codec:v rawvideo -pix_fmt yuyv422p /dev/video20
I use Imaginary Teleprompter with a custom glass frame. The problem I face is slight ghosting of the text but nothing too annoying for the price (free). If I keep teleprompting I will consider purchasing a glass or a prompter from https://imaginary.tech/teleprompter/
- Chroma
- Green: For daylight scenes
- Blue: For dark scenes
- RGB Green: For low light
- RGB backgrounds
- Keep Iso low
- Aperture full open
- 180 degree rule
- Better grey than white backdrop
- Teleprompter
- Use street language, not written language
- Use peripheral vision to read
- Use reminder keywords not word by word script