Re: audio driver

Tech Tip: Click here to run a free scan for Windows Errors and optimize PC performance

hi paul and dean thank u very much.

one more doubt,
is ther any need to change the mdd?
suppose we need 2 channels ;ie two input(one at a time)
then we need to change in pdd only?
we are just taking the mdd as library only.Isn't it?
My understanding is that we need to change the pdd only.

Dean Ramsier wrote:
Actually, most audio drivers still use the legacy model. A UAM driver is
nothing more than a fancy name for a driver that implements more/different
functionality than the standard MDD. Both models still provide the same
upper level interface. Most audio hardware doesn't support multiple output
streams, mixing etc so the provided MDD/PDD model is sufficient.

Multiple streams typically means multiple output streams. Trying to play
more than one sound simultaneously for example. The MDD/PDD driver only
allows supports one output stream, so the second attempt to open one would
fail. CE solves this by providing a software mixer in the kernel that
intercepts the opens and mixes them together before sending them to the
single output channel supported by the driver. The driver only sees one
open, so it is happy. The applications all get to play their sounds, so
they are happy.

The only downside is that software mixing is an expensive task in terms of
performance. If your audio hardware was able to support multiple streams
and mix them *in hardware* then you could expose that support in the audio
driver. You couldn't use the standard MDD without modification because it
doesn't support multiple opens. In that case you modify the MDD or go to a
completely custom implementation (and call it a UAM driver).
--
Dean Ramsier - eMVP
BSQUARE Corporation


"Henrik Viklund" <henrik.viklund@xxxxxxxxx> wrote in message
news:1164190010.866420.140860@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
princyfrancis@xxxxxxxxx skrev:
so u r saying,i can use PDD MDD model for supporting "multiple streams"
of data.
I'm saying that the information you refer to discuss the previous
driver limitations (CE3 and before?). Back then the audio architecure
was basically monolithic stream interface driver consisting among other
things of a MDD and a PDD layer, where the MDD layer only could handle
one open stream per driver instance. The rudimentary audio architecture
offered back then left a lot to ask and with the introduction of CE4
the UAM was introduced offering a much improved architecture and wave
API with support for hardware mixers, audio compression management etc.
Among other things the MDD architecture was improved to handle multiple
open streams per driver if the underlying harware is up to the task.

actually what is this "multiple streams"
It refers to the ability for a stream interface driver (in this case
the waveform audio driver) to handle multiple open streams correctly
(like one serial port driver can handle multiple serial ports
simultaneously). In this case it basically means multiple audio
channels, like a recording channel and a playback channel.

Henrik Viklund
http://www.addlogic.se


Henrik Viklund wrote:
Note that the particular paragraph you're quoting referes to the
*previous* architecture and sample drivers.

Henrik Viklund
http://www.addlogic.se

princyfrancis@xxxxxxxxx skrev:

thank u very much Dean.
i was totally confused by the usage"waveform audio ".
ok i have one more doubt.I PDD MDD model it is said that "Using MDD
and PDD, the model had the following limitations:
1.No support for multiple streams
2.No multiple devices on one driver
3.No reliable support for looping
4.Poor support for streaming "

what is mean by the above 4 limitations?

what is mean by"multiple streams"?
so can't we give 2 inputs( ie one from mic and one from some receive
data pin)?
is it not possible?



Dean Ramsier wrote:
The hardware that does the digital to analog conversion uses
waveform audio
(PCM). You can play other formats, but higher level software
translates it
into a PCM stream before it reaches the audio driver.

--
Dean Ramsier - eMVP
BSQUARE Corporation


<princyfrancis@xxxxxxxxx> wrote in message
news:1164102463.676769.191880@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
hi all
i am developng an audio driver in wince5.0 based on pxa270
processor
.why there is in platform builder ,mensioned only about waveform
audio.whether it can play only files with .wav entension?

i am not getting ,why you are always useing the term "waveform
audio"?

why does this audio driver is given the name "waveform
audio".plz
rely me.
what is the significance of the term "wave".
plz reply me soon.i am in urgent need of it.
thanks in advance.
regards princy



.

Relevant Pages