I have some video which appears under VLC to have a frame rate of 29.970030 when I edit it in Shotcut it appears to be perfect but when I export the scene I want from it - it has some quite severe jerking in it making it unusable, Shot cut is showing the frame rate as being 29.97 and says the video is interlaced ( i know the interlaced part is true ).
Does Shotcuts frame rate field size need increasing and could this be the cause of the tearing.
The true frame rate of that video is most likely 30000/1001 and 29.970030 is an approximation as well. If you look at the XML of the export job, we detect 29.97 and convert it back to the fraction form (see frame_rate_num and frame_rate_den):
<?xml version='1.0' encoding='utf-8'?>
<mlt root="/home/ddennedy/Desktop/Shotcut/Shotcut.app" out="00:00:30.030" LC_NUMERIC="en_US.UTF-8" title="Shotcut version 18.03.02" parent="producer0" version="6.7.0" in="00:00:00.000">
<profile display_aspect_num="1280" width="1280" sample_aspect_den="1" frame_rate_den="1001" colorspace="601" height="720" progressive="1" sample_aspect_num="1" display_aspect_den="720" description="automatic" frame_rate_num="30000"/>
<consumer mlt_service="avformat" threads="0" real_time="-4" width="1280" frame_rate_den="1001" aspect="1.77778" preset="fast" ar="48000" frame_rate_num="30000" top_field_first="2" vcodec="libx264" crf="21" target="/home/ddennedy/Videos/tests/test.mp4" progressive="1" deinterlace_method="yadif" height="720" bf="2" f="mp4" rescale="bilinear" g="15" movflags="+faststart" acodec="aac" ab="384k"/>
In case your framerate really is 29.970030 and not 30000/1001, it would not be severe and there would be only the very occasional frame drop.
Actually what I am seeing is really odd.
I’m a sci-fi fanatic and was making animated shorts to represent my videos which I have format shifted onto my raid array.
One series called Andromeda when I export through any method leaves jerking in the output video every 1 to 2 seconds whereas if I compare that with the short I made for Startrek The Next Generation it is absolutely perfect.
My only conclusion - THUS FAR ( subject to change ) is that there is something in the Andromeda shot which is mucking up input to Shotcut.
Note the source material for both of these shorts is absolutely perfect when played in VLC.
My next step ( if it is allowed will be to post a small amount of video, no longer than 10 seconds each - so 40 seconds in total so you can see the differences and analyze the files ) but this is dependent on me being able to find a source file cutter other than shot cut so you can see the changes - I thought I might try avidemux with the source copy function and check to see if that works.