ive been working on this for a while

it lets you measure the time it takes your processor to encode video to different video codecs, it simply times the amount of time it takes to prep and encode a shotcut project file that has 4K video with some text overlay and some awesome music in flac by GUNSHIP

  • bench.x264.py Encoding 3840x2160p24 4K Shotcut Project using qmelt to x264 with Slow Preset and 16 CRF
  • bench.x265.py Encoding 3840x2160p24 4K Shotcut Project using qmelt to x265 with Slow Preset and 16 CRF
  • bench.vp9.py todo


because mp4 is outdated, flac and mkv is the future


this should test cpu/ram and ssd/hdd and the results are in seconds

the below scores are average of 5 runs
please note that these systems are not benchmark systems and are real world

thanks to iWalkingCorpse for 1500X, Vegetables for 1700 3.8Ghz

This Benchmark is based on Shotcut and MLT open source project
which you can find at https://www.shotcut.org/

Crop Real 4K to UltraHD 4096>3840

ffmpeg -i A001_C014_01217G.4k.y4m -vf "crop=3840:2160:128:0" outputgirl.2160p.y4m

Resize to 1080p

ffmpeg -i outputgirl.2160p.y4m -vf scale=1920:-1 outputgirl.1080p.y4m

Resize to 720p

ffmpeg -i outputgirl.2160p.y4m -vf scale=1280:-1 outputgirl.720p.y4m

Convert source to YUU4MPEG2

ffmpeg -pix_fmt yuv420p -i any-source-video.webm output.y4m

Convert a sequence of images to YUU4MPEG

ffmpeg -f image2 -i "A004_C001_0122K7.00%05d.png" -pix_fmt yuv420p ouput.y4m

extract a single frame

ffmpeg -i input.avi -f image2 -ss 14.342 -vframes 1 frame.png

extract a single frame (altenative method)

ffmpeg -i input.y4m -f image2 -vf select="gte(n\, 1)" -vframes 1 frame.png

extract 1 frame after frame 100

ffmpeg -i input.y4m -f image2 -vf select="gte(n\, 100)" -vframes 1 frame.png

Easily deploy an SSD cloud server on @DigitalOcean in 55 seconds. Sign up using my link and receive $10 in credit
Proudly published with Ghost | Theme based on Draugur | Header Image by GUWEIZ@DeviantArt
Creative Commons Licence