e-t172
10th January 2024, 22:17
https://raw.githubusercontent.com/dechamps/videojitter/f6dd99b49d6e34d3bd9c2c6a3b8488a548e30921/videojitter_test/cases/asuswmp_23p_at_240hz/test_output/report.svg (https://github.com/dechamps/videojitter)
This thread is to introduce and discuss videojitter (https://github.com/dechamps/videojitter), my latest addition to the world of video measurement tools.
videojitter is a free and open source tool for accurate measurement of frame presentation timing in a video playback system. It can be used to detect and quantify issues such as:
Frames being displayed too early or too late
Dropped frames
Repeated frames
Periodic patterns such as 3:2 "24p@60Hz"
Sudden changes in presentation cadence
Video clock jitter
Video clock skew
etc.
(See examples) (https://github.com/dechamps/videojitter/blob/master/EXAMPLES.md)
Contrary to purely software-based solutions, videojitter relies on a hardware instrument to physically measure the light output of the display. This means that it is not limited to software video players and can measure the output of any video player, even dedicated hardware players (Smart TVs, set-top boxes, smartphones, etc.), as long as they can play the test video files videojitter provides. You can build your instrument (https://github.com/dechamps/videojitter/blob/master/INSTRUMENT.md) in minutes with only a few dollars worth of parts and a screwdriver.
videojitter treats the entire video playback chain as a black box. Contrary to "white box" approaches (e.g. GPU statistics), the playback system cannot lie in videojitter measurements - videojitter measures the same light that reaches your eyeballs.
For more information, feel free to refer to the documentation (https://github.com/dechamps/videojitter#how-does-it-work).
Hope you like it! Feedback is welcome, either here or as a GitHub issue (https://github.com/dechamps/videojitter/issues) or discussion (https://github.com/dechamps/videojitter/discussions).
This thread is to introduce and discuss videojitter (https://github.com/dechamps/videojitter), my latest addition to the world of video measurement tools.
videojitter is a free and open source tool for accurate measurement of frame presentation timing in a video playback system. It can be used to detect and quantify issues such as:
Frames being displayed too early or too late
Dropped frames
Repeated frames
Periodic patterns such as 3:2 "24p@60Hz"
Sudden changes in presentation cadence
Video clock jitter
Video clock skew
etc.
(See examples) (https://github.com/dechamps/videojitter/blob/master/EXAMPLES.md)
Contrary to purely software-based solutions, videojitter relies on a hardware instrument to physically measure the light output of the display. This means that it is not limited to software video players and can measure the output of any video player, even dedicated hardware players (Smart TVs, set-top boxes, smartphones, etc.), as long as they can play the test video files videojitter provides. You can build your instrument (https://github.com/dechamps/videojitter/blob/master/INSTRUMENT.md) in minutes with only a few dollars worth of parts and a screwdriver.
videojitter treats the entire video playback chain as a black box. Contrary to "white box" approaches (e.g. GPU statistics), the playback system cannot lie in videojitter measurements - videojitter measures the same light that reaches your eyeballs.
For more information, feel free to refer to the documentation (https://github.com/dechamps/videojitter#how-does-it-work).
Hope you like it! Feedback is welcome, either here or as a GitHub issue (https://github.com/dechamps/videojitter/issues) or discussion (https://github.com/dechamps/videojitter/discussions).