[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: SourceForge CVS



Hi Tim,

A utility for checking out card parameters will certainly be useful to
many peaple here.

I believe that the apparently superfluous tests in alsa_configure are
to do automatic parameter selection by default and if that fails, the
user can initialize some of the variables to a non-zero value and the
automatic setting will be skipped. The use of the
snd_pcm_get_period_time function in the place it is currently used is
most likely a bug however. 

I also have one machine where the period size is by default 16, which
is far too low for a buffer size for the eci callback
(chunk_size). Choosing chunk_size as a multiple of the cards
period size and about a quarter of the alsa buffer size has given me
good results. As we ar using interlieaved mode, I believe that the
actual value of the period size is not that important. 

By the way, I believe that most crashes I see are due to bugs in the ViaVoice
speech synthesis code. I have been recording the communication between
emacspeak and outloud to see what could trigger the crashes but have
not yet been able to draw any meaningful conclusions.

Best regards, Lukas

 > Hi Lukas, > >
Thanks for that, you saved me a bit of time. In fact, after a little >
thought and a bit of judicious scanning of the alsa docs (I'm new to >
alsa), plus a little very simple utility program I wrote to dump out >
my cards hw parameters, I've managed to get things working really well
> in just over an hours work over lunch.  > > Here is what I've
discovered: > > The snd_pcm_hw_params_get_buffer_time will work if you
have saved your > earlier settings prior to calling it. I believe what
is happening is > that once you have selected the mode, format, rate
etc, then the call > will work. I guess by the time you have selected
the other parameters, > the possible buffer size choices have been
constrained enough for the > function to identify a specific size.  >
> I guess we could do one of two things. Put a call to set the >
parameters prior to calling this function or test the return code from
> snd_pcm_hw_params_get_buffer_time and if its less than 0 then call >
snd_pcm_hw_params_get_buffer_time_max. The problem with the second >
solution, as you pointed out, is that you may end up with an overly >
large buffer and get underruns etc. However, I did notice the result >
from calling the function after having set and saved the earlier >
values seems very small (16), especially when compared tot he max
value.  > > Something I did notice when looking at the code is there
are a couple > of if statements which don't appear to be really doing
anything - I'm > guessing they are an artifact from when Raman was
developing the code.  > For example > > if (buffer_time == 0 &&
buffer_frames == 0) { > > both buffer_time and buffer_frames are
initialized to 0 and this is > the first time they are referenced, so
the test is always true. I've > not removed them yet, but unless I'm
missing something, there doesn't > seem to be any reason for them to
be there.  > > I wrote a very simple utility which dumps out my cards
hardware > settings. Using the values from this, I created a new
.asoundrc file > which appears to have fixed all my underrun
problems. I'm going to > clean it up a bit and send it to the list
just in case someone else > finds it useful.  > > One very good thing
about the new atcleci.cpp code is that it seems *a > lot* more
stable. In fact, rather than the server dying every few > minutes, it
has not died for me in nearly an hour so far. I'm thinking > this
could be due to your suggested additions to add the functions to >
free up resources correctly - the crashes could have been due to >
resource limitations. At any rate, its working remarkably well now.  >
> Another interesting discovery I made is that the ViaVoice eci.ini
file > appears to be slightly different in current versions than the
older > one. I decided to generate a new one using the initgen utility
just to > check there had been no changes in later versions of
ViaVoice. The one > generated by initgen has some extra header fields
not in the original > i.e.  > > [1.1] >
Path=/opt/IBM/ibmtts/lib/eng50.so > Version=5.0 > Concatenative=1.0 >
CallbackFlag=0x3f > ...  > ...  > > regards, > > Tim > >
-----------------------------------------------------------------------------
> To unsubscribe from the emacspeak list or change your address on the
> emacspeak list send mail to "emacspeak-request@cs.vassar.edu" 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@cs.vassar.edu" with a
subject of "unsubscribe" or "help"


Emacspeak Files | Subscribe | Unsubscribe | Search