[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Search]
Re: Error: Could not load libibmeci.so
- To: "Robert D. Crawford" <rdc1x@xxxxxxxxxxx>
- Subject: Re: Error: Could not load libibmeci.so
- From: Lukas Loehrer <listaddr1@xxxxxxxxxxx>
- Date: Sat, 27 May 2006 19:25:21 +0200
- Delivered-To: priestdo@xxxxxxxxxxx
- Delivered-To: emacspeak@xxxxxxxxxxx
- In-Reply-To: <8764jrsbir.fsf@xxxxxxxxxxx>
- List-Help: <mailto:emacspeak-request@xxxxxxxxxxx?subject=help>
- List-Post: <mailto:emacspeak@xxxxxxxxxxx>
- List-Subscribe: <mailto:emacspeak-request@xxxxxxxxxxx?subject=subscribe>
- List-Unsubscribe: <mailto:emacspeak-request@xxxxxxxxxxx?subject=unsubscribe>
- Old-Return-Path: <listaddr1@xxxxxxxxxxx>
- Reply-To: listaddr1@xxxxxxxxxxx
- Resent-Date: Sat, 27 May 2006 13:25:24 -0400 (EDT)
- Resent-From: emacspeak@xxxxxxxxxxx
- Resent-Message-ID: <4Qq2JB.A.FI.EuIeEB@xxxxxxxxxxx>
- Resent-Sender: emacspeak-request@xxxxxxxxxxx
Hi Robert,
first the easy stuff. If the beep tool from the beep package does no
longer work, One way to restore the beep is to use the beep script
from the servers directory in the emacspeak CVS. You will need the
tones utility from the siggen package available from:
http://www.comp.leeds.ac.uk/jj/linux/siggen.html
I believe that if you run Tim's utility while an .asoundrc file is
present, it is not surprising that the values shown will correspond to
those in the .asoundrc file.
I have seen the overlapping problem too. For me, it occured
after I have removed the .asoundrc file.
under theese circumstances, a very large buffer size was used. This
would not be a large problem unless dmix is used. The problem only
occured for me after more than one process has tried to access the
sound device simultaniously. I guess that the step from 1 to 2
triggers the activation of dmix.
I suspect that with dmix, it is hard to cancel audio that has already
been mixed into the shared buffer. Discarding the buffer content when
one process issues a snd_pcm_drop() call would also discard the audio
being played by other processes, which is unacceptable. If the shared
buffer is large, this effect becomes noticeable. What I do not
understand is why the cancelling process continues playback
immediately, and does not wait until its own audio has stopped
playing. This is what causes the overlap.
Anyway, what you have to do is choose a buffer size that is small, I
would say at most 100ms. Mine is 4096 frames. Add some printfs to
atcleci.cpp to see what sizes for buffer_size and chunk_size are used.
Note that when testing different settings in the .asoundrc file, the
changes only take effect after all processes have closed the sound
device. This was at least the case for me. I guess that once the dmix
mechanism is started with certain settings, these can not be changed
while some processes still use the old ones. This may be different
with multi-channel sound cards, but in this case, there are probably
better alternatives than the use of dmix anyway.
Best regards, Lukas
Robert D. Crawford writes ("Re: Error: Could not load libibmeci.so"):
> Tim Cross <tcross@xxxxxxxxxxx> writes:
>
> > the stutter problem can probably be sorted by tweaking the parameters
> > in the .asoundrc file.
>
> Is there any easily understood documentation on what the values should
> be,, more importantly, what they should not be, and the like? It is not
> clear to me what all the settings do. I looked at what is returned by
> your utility and what the settings are in the .asoundrc file, and most
> of them are the same. As I mentioned before, I changed the one that was
> not the same, but it seemed to make no difference.
>
> I also notice that there are values returned that are not listed in the
> asoundrc file. Should I add these? Seems that what is being done now
> is to use defaults. It is useable now, and I notice only a few issues:
>
> 1. The overlapping text being spoken. If one bit of text is being
> spoken and another starts before the first finishes, there is
> overlap creating unintelligible output.
>
> 2. Sometimes there is a bit of scratch to the output. This is a _very_
> minor thing. I only mention it for completeness.
>
> 3. I no longer have my beep when going over blank lines. I figure this
> is an issue that can easily be resolved with a bit of reading, I
> just haven't taken the time.
>
> > However, also try resetting the system settings with C-e d R. After
> > starting emacspeak and the startup message has completed, hit C-e d R
> > and see if that stops the stutter.
>
> This did not work for me, but I will keep it in mind for the future.
>
> Thanks,
>
> rdc
>
> --
> ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> Robert D. Crawford rdc1x@xxxxxxxxxxx
>
> -----------------------------------------------------------------------------
> To unsubscribe from the emacspeak list or change your address on the
> emacspeak list send mail to "emacspeak-request@xxxxxxxxxxx" with a
> subject of "unsubscribe" or "help"
>
-----------------------------------------------------------------------------
To unsubscribe from the emacspeak list or change your address on the
emacspeak list send mail to "emacspeak-request@xxxxxxxxxxx" with a
subject of "unsubscribe" or "help"
Emacspeak Files |
Subscribe |
Unsubscribe | Search