On 16/12/10 08:17, Dave Koelmeyer wrote:
On 16 Dec, 2010,at 07:39 PM, Dave Koelmeyer<[email protected]> wrote:
On 14 Dec, 2010,at 07:36 AM, Eugen Dedu<[email protected]>
wrote:
On 09/12/10 04:42, Dave Koelmeyer wrote:
>
> On 08 Dec, 2010,at 05:05 AM, Eugen Dedu<[email protected]>
wrote:
>>
>> Please wait a bit, until 3.3.0 gets out, to test it and inform us if the
>> freeze still appears.
>
> Hi Eugen,
>
> Will do :) I have shot a video of the behaviour along with a description
here if
> it helps at all:
>
> http://www.davekoelmeyer.co.nz/ekigatheora.html
>
> This should remove any doubt as to how the freezing is triggered.
I have put a Windows executable of pre-3.3.0 Ekiga at
http://eugen.dedufree.fr/ekiga-setup-3.3.0-git-852_g1f9a30d.exe
<http://eugen.dedu.free.fr/ekiga-setup-3.3.0-git-852_g1f9a30d.exe> so that
you test if your freeze still appears (you can build it yourself if you
prefer: take unstable from
http://wiki.ekiga.org/index.php/Building_Ekiga_for_Windows)
Thanks very much Eugen. I have installed your .exe on the system in question
(Windows 7 Enterprise 64 bit).
I don't believe it has resolved the freezing issue, but I have made the
following observations.
The freezing is much, much more easy to induce with low "Maximum video
bitrate" settings in the Ekiga client. All my tweaking involved adjusting the
bitrate, while leaving the "Picture Quality" slider parked at maximum frame
rate (i.e all the way to the right).
1) With "Maximum video bitrate" set to 64Kb/s, the video will freeze with very
slight movement, more often than not it will freeze as soon as the call is
connected.
2) With "Maximum video bitrate" set to 1024Kb/s, I can reproduce the behaviour
exhibited in my original video above, as that was what the settings were when
the video was recorded.
3) With "Maximum video bitrate" set to a high number, e.g 3072Kb/s, then it
becomes very difficult to freeze the video deliberately; lots of flailing
around in front of the camera to induce a freeze (it can still be done
however).
Settings above ~4000Kb/s seem to revert to the behaviour observed when set to
~1000Kb/s however.
Just another note that I can also reproduce this behaviour identically on
Windows XP Professional SP3, using the stable Ekiga 3.2.7.
The problem is that I have not yet tested debugging crashes/freezes on
Windows... There is some info at
http://wiki.ekiga.org/index.php/Debugging_Ekiga#How_to_get_a_stack_backtrace_from_a_crash,
but I think it is obsolete. There is someone who work on this. In the
meantime, please post a bug report with this issue, so that we do not
forget it.
Thanks,
--
Eugen
_______________________________________________
ekiga-list mailing list
[email protected]
http://mail.gnome.org/mailman/listinfo/ekiga-list
