Comment 19 for bug 1195050

I decided to try the memory leak tool valgrind. It reported what seemed to be many memory leaks. I wonder if some of those are false leaks.
Here is the command that I ran:
valgrind --tool=memcheck --trace-children=yes --error-limit=no --leak-check=full ./openshot | tee debug
Here is a summary of the final report:
==622== LEAK SUMMARY:
==622== definitely lost: 40,640 bytes in 82 blocks
==622== indirectly lost: 57,858 bytes in 1,797 blocks
==622== possibly lost: 619,101,321 bytes in 86,111 blocks
==622== still reachable: 476,260,399 bytes in 95,483 blocks
==622== suppressed: 0 bytes in 0 blocks
==622== Reachable blocks (those to which a pointer was found) are not shown.
==622== To see them, rerun with: --leak-check=full --show-reachable=yes
==622==
==622== For counts of detected and suppressed errors, rerun with: -v
==622== Use --track-origins=yes to see where uninitialised values come from
==622== ERROR SUMMARY: 186054124 errors from 5088 contexts (suppressed: 0 from 0)
My debug file is basically empty so I ,perhaps, should have run it with >&

Here are a few final lines. I will have to rerun it. It crashed before rendering was complete.
==622== 96,468,992 bytes in 23 blocks are possibly lost in loss record 20,865 of 20,867
==622== at 0x4C290FE: memalign (in /usr/lib64/valgrind/vgpreload_memcheck-amd64-linux.so)
==622== by 0x117214D6: mlt_pool_alloc (mlt_pool.c:125)
==622== by 0x15008FAA: allocate_buffer (producer_avformat.c:1219)
==622== by 0x1500C0DA: producer_get_image (producer_avformat.c:1528)
==622== by 0x117091DA: mlt_frame_get_image (mlt_frame.c:573)
==622== by 0x28F08A3C: deinterlace_yadif (filter_deinterlace.c:139)
==622== by 0x28F08F8E: filter_get_image (filter_deinterlace.c:235)
==622== by 0x117091DA: mlt_frame_get_image (mlt_frame.c:573)
==622== by 0x147D6FAD: get_image (filter_fieldorder.c:34)
==622== by 0x117091DA: mlt_frame_get_image (mlt_frame.c:573)
==622== by 0x147D5B7B: filter_get_image (filter_crop.c:77)
==622== by 0x117091DA: mlt_frame_get_image (mlt_frame.c:573)