All way above doesn't help my. This problem not related with frame processing after reading. I got this problem on version opencv-python 4.11.0.86, but how I understand it can be in any version opencv. This problem is related exclusively to the codec ffmpeg. I just replaced this codec on my env path C:\Users\x\anaconda3\envs*name*\Lib\site-packages\cv2\opencv_videoio_ffmpeg4110_64.dll to a codec from another version opencv on which I checked the functionality. For me I take ffmpeg codec from version opencv-python==4.5.4.60 This one opencv_videoio_ffmpeg454_64.dll And just renamed it. This problem was solved.