Large downloads slowing/stopping as they progress

[Jim web]

In article <54c0749291web at audiomisc.co.uk>, Jim web
<web at audiomisc.co.uk>
wrote:
> In article <554C2670.90809 at internode.on.net>, Nick Payne
> <nick.payne at internode.on.net> wrote:
> > I've been downloading parts of the World Snooker Championship
> > coverage. Some of the files are quite large - four hours long and over
> > 4Gb if downloaded as HD. I've noticed (this happens with 2.92 on both
> > Linux and Windows) that as the downloads progress they get slower and
> > slower, and quite a number of the downloads of 120 minutes or more
> > eventually come to a complete halt, without any error message being
> > displayed. If I stop the download at that point and start it again it
> > won't resume successfully (loops continually with the message "Command
> > exit code 1 (raw code = 256)" and I have to start it again from the
> > beginning.


> FWIW I also was trying to get just the last part of the 'final evening'
> programme of the snooker. Was using --start 02:00:00 to avoid having to
> fetch a file for a program that was over 4 hours in total.

I tried again as an experiment this morning. This time with a start at
2:45:00 as the previous try had provided me with the section from 2h to
about 3h. The program was (apparently) about 4h15m long so this should have
got me a file about 1h30m long.

I left this running in a xterminal window as I did something else. When I
checked some time later I was getting a looping error

File contains neither video or audio exit code 1 (raw code = 256)

So I did a ctrl-C and shut the terminal, then looked at the resulting flv.

As came, it will play with VLC, but as with earlier results, showed the
wrong duration and would crash out if I tried to jump to a later section
whilst playing. I guess this may be the norm for the raw flv's?

So (as before) I ran the flv though ffmpeg with codec copies into an mp4.

This says it is 1h18m long, and plays OK, inc being able to jump to a
different point in VLC. Seems to cover up to the end result. It seems
slightly shorter than I'd expected but has what is required.

Makes me continue to think there is something odd about this programme file
which confuses gip. One curious point is that both the earlier attempt and
this one gave a file whose playable duration was about 1h18m, but from
different start points. Maybe a coincidence?

Other programme files I've fetched recently seemed fine.

Jim

-- 
Electronics  http://www.st-and.ac.uk/~www_pa/Scots_Guide/intro/electron.htm
Armstrong Audio  http://www.audiomisc.co.uk/Armstrong/armstrong.html
Audio Misc  http://www.audiomisc.co.uk/index.html