2.7.6 crashes reproducibly on its own py script: Cannot set up encoder

[gamez]

OK, let me retry on 64 bit OS
g.

[eumagga0x2a]

OK, let me retry on 64 bit OS

...with the latest 64-bit Avidemux nightly, please.

[gamez]

latest 64-bit Avidemux nightly

Yes sure - I figure it's the 201102_ffa9a6d4b29?

So the good news is that
a) it now correctly saves the border frames' µs as .........244 and never .........256
b) in consequence has no more seeking problems.

There's still a (little) question for the markers in conjunction with stacked segments.

In my tests I was doing

Code Select Expand

adm.addSegment(0, 173056244, 918760000)
adm.addSegment(0, 1681216244, 790080000)
adm.addSegment(0, 3076816244, 1335600000)
adm.addSegment(0, 4939096244, 1209960000)
adm.addSegment(0, 6826136244, 574720000)
with and without

Code Select Expand

adm.markerA = 0
adm.markerB = 4829120000Note that markerB sits at the end of the sum of the segments.

When I set the markers, the resulting file size is 2.390.157.657, when I leave them out it's  2.390.161.909.

Where does the 4K difference come from?

g.

[eumagga0x2a]

a) it now correctly saves the border frames' µs as .........244 and never .........256

I wish I could attribute this to a particular code change.

When I set the markers, the resulting file size is 2.390.157.657, when I leave them out it's  2.390.161.909.

How many frames are written to the output file in the first vs. the second case? I wonder what does

Code Select Expand

print("marker A = " + str(adm.markerA) + " marker B = " + str(adm.markerB) + "\n")
report to the log when added after all the addSegment() calls.

I assume that you have not forgotten to clearSegments() prior to creating your custom segment layout using addSegment().