[prev in list] [next in list] [prev in thread] [next in thread]
List: linux-ppc
Subject: Re: RealPlayer stability
From: Warren Nagourney <warren () dirac ! phys ! washington ! edu>
Date: 2000-07-04 18:14:19
[Download RAW message or body]
Peter,
I changed the preference to "...OSS...", quit and restarted the player and
it seemed to work fine for a while. I then (while it was playing) attempted
to look at the "Statistics" menu to see the graph of network performance
(it worked fine a couple of days ago). The graph came up and a couple of
seconds later the player crashed.
Finally, thinking it is a timing problem involving the network, I put some
load on the network from another computer (I have several computers sharing
the same 28.8 modem via a server machine which uses IP net masquerading).
Reading mail from the other computer was fine - as soon as I attempted to
load a complex web page, the player crashed.
By the way, the aliasing I mentioned is not there with all audio sources.
Listening to our local streaming NPR station (KUOW) sounds fine (no
aliasing) but when listening to a clip from a Prairie Home Companion
archive, there are a lot of spurious high frequency artifacts which I took
to be aliasing. Both are playing at 16 Kbps.
Thanks much for the help.
Warren Nagourney
--On Tue, Jul 4, 2000 9:18 AM -0700 Peter Godman <pgodman@real.com> wrote:
> Warren:
>
> I have another idea. If you have the energy please try the following:
>
> Switch the audio preferences to "Support for older OSS" and possibly close
> and restart your player (I think there may be a bug with that pref taking
> effect). Anyway, the frequency with which you're seeing crashes seems
> quite suspicious and I curious as to whether disabling postional-query
> ioctls on the audio device may help the situation.
>
> WRT your esd problem. Please change the pref and restart the player.
> I'll look into why these don't work properly...
>
> Thanks,
> Peter Godman
>
> On Mon, 3 Jul 2000, Warren Nagourney wrote:
>
>> Thank you for the help.
>>
>> I added the line "NetworkingThreading=0" and RealPlayer crashed after
>> about 1 minute (formerly it seemed more stable). Of course, I
>> relaunched RealPlayer after changing the file.
>>
>> Even though ESD is running on my machine, selecting ESD in the
>> preferences results in complaints from RealPlayer ("Cannot open sudio
>> device. Another application may be using it").
>>
>> Originally, I installed it in my home directory. Reinstalling it (as
>> root) in /usr/local seemed to help the stability - it didn't crash
>> again for about half an hour of use (except when I tried to use the
>> panel controls).
>>
>> Now it is crashing again. This all may be coincidental, but these are my
>> experiences.
>>
>> Thanks.
>>
>> Warren Nagourney
>>
>> --On 07/03/00 11:09:19 -0700 Peter Godman <pgodman@real.com> wrote:
>>
>> >> Has anyone gotten real player, recently made available for linuxppc,
>> >> stay up for more than a few minutes?
>> >
>> > I have ;-)
>> >
>> >> I downloaded and installed it last night and my longest un-interrupted
>> >> use
>> >> was about 15 minutes - usually it would suddenly crash every few
>> >> minutes or
>> >> so. I am using a 28.8 connection - increasing the size of the buffer
>> >> and reducing the expected speed (to 19.2 kbps) seemed to help a
>> >> little.
>> >
>> > If you could try something on your system, I'd appreciate it.
>> > Try adding
>> > NetworkThreading=0
>> > to your ~/.RealNetworks_RealMediaSDK_60 file
>> >
>> > This will turn off threaded networking, which is a semi-experimental
>> > feature at present.
>> >
>> >> Also, there are signs of aliasing in the decoded output making it
>> >> sound acceptable only with the treble (on the external speakers)
>> >> turned down all
>> >> the way.
>> >
>> > I believe that the audio device implementation for some machines may
>> > use a braindead resampler (sample repetition) which I believe could
>> > cause the sort of aliasing you experience. Most audio apps for linux
>> > open the device at 44100, resulting in no aliasing. RealPlayer
>> > chooses the audio dev. speed based on datatype.
>> >
>> > One way to work around this is to use ESD, which AFAIK always opens the
>> > device at 44100. Another is to look into whether the audio device
>> > implementation has a quality (fft) resampler.
>> >
>> >> System is linuxppc 2k on beige g3/400 running Windowmaker and latest
>> >> versions of stable kernel and XF86-4.
>> >
>> >> I'm glad they ported it, but it seems unusable at present.
>> >
>> > I'm sorry to hear that. If you want it to work well straight away,
>> > please try turning off threading and using ESD.
>> >
>> > Thanks,
>> > Peter Godman
>> >
>> >> Warren Nagourney
>>
>>
>>
>>
>>
>>
--------------------------------------------------------------------------
Warren Nagourney <warren@dirac.phys.washington.edu> Voice: 206-543-9585
University of Washington 206-543-0143
Physics Dept., Box 351560, Seattle, WA 98195 Fax: 206-685-0635
** Sent via the linuxppc-user mail list. See http://lists.linuxppc.org/
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic