Excellently written Jim. Guess one could conclude that in the Orion, the 25
ms key in to RF out delay points out some deficiencies in either design or
firmware implementation?
Thanks for the good explanation!
Tom - W4BQF
> On Fri, 13 Aug 2004 10:28:03 -1000, Ken Brown wrote:
>
> >I agree with you on a lot of the points you have made.
>
> DSP can be a double-edged sword. You can do all sorts of wild and
wonderful
> things with it, but you can easily "outrun your headlights" if you aren't
careful. The
> biggest issue is, indeed, latency -- the time that it takes DSP to do it's
thing. I
> work in pro audio, where we use DSP extensively. In our world, 5 ms is a
big
> deal, and it takes close to that to do A/D, simple processing, and a D/A
to feed
> loudspeakers.
>
> On the other hand, when you're sending and receiving CW and trying to
> accomplish QSK, 5 ms is not a lot, and shouldn't degrade performance a lot
if
> you are reasonably 1) careful and 2) skillful. That is, the processing
that you are
> doing on transmit can go in sequence with the processing on receive, so as
long
> as the 5ms is simply an overall delay (as opposed to being added to the
> transitions from keydown to keyup).
>
> Likewise, the quality of the DSP algorithms has a lot to do with how well
it works.
> That seems to boil down to the DSP power available and the
skill/experience of
> the programmers. There are clearly learning curves at play, and later
generation
> DSP is clearly more satisfactory than earlier stuff, both in pro audio and
ham
> radio.
>
> The bottom line is that I don't think it's correct to say that just
because there is
> DSP that QSK has to suffer. It MAY, but I don't think it MUST. It really
does
> depend on the implementation.
>
> Jim Brown K9YC
>
>
> _______________________________________________
> TenTec mailing list
> TenTec@contesting.com
> http://lists.contesting.com/mailman/listinfo/tentec
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|