[hpsdr] Janus-Ozy Sound Card Driver [WDM]

[Philip Covington]

On Sat, May 31, 2008 at 4:48 PM, Richard Faust <dick_faust at hotmail.com> wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
>
> Greetings All,
>
> The ASIO comment is interesting and will help to illustrate several points in my original post.
> I have a Steinberg SDK for AISO 2.2, but have not as yet deployed.  I have used AISO4ALL [which I am reasonably sure was developed from Steinberg's SDK] and that program requires a WDM driver !  Point - ASIO and WDM are what some might loosely refer to as specifications, they define requirements to successfully communicate data.
>
> Another point & my opinion is the net.framework that M$ uses on Windoze is what may be causing many problems associated with SDR and DSP.  This is a program interface that is primarilary designed > for business database communications applications.  The program and its parts are huge and directly relate to the doze in windoze.  This is why I do not wish to use a net.framework development chain.

This is pure nonsense.  DttSP, which does the DSP in PowerSDR, is
written in C - no managed code there.  PortAudio which is used in
PowerSDR is written in native code and not .NET managed code.  If
"many" of the problems associated with SDR and DSP are caused by the
.NET framework as you claim, then how do you account that the DSP
functions that are experiencing those "problems" are done in
non-managed C code?  There are many time when managed code outperforms
native code especially when .NET's JIT compiler is able to
optimizations that are not possible when the code is precompiled to
machine code before hand - there is much written about this with
benchmarks if you google around.

I will never understand people's fascination with making OZY + JANUS
look like a sound card - in my "opinion" it is a giant step backwards
as far as SDR use and making it NOT look like a sound card was one of
the reasons in developing OZY - JANUS in the first place!  Much of the
wide variability in the performance of the SDR-1000 was because of the
user's choice of sound card.

Phil N8VB

 1212287759.0