--- In FT817@e..., baranger1@a... wrote:
Wow!
I know from all of my pro audio gear that one of the bigest problems
with DSP gear is latency.
125ms delay is very,very poor.
The SGC processor has some effective algorithms in it and no doubt the
long delay is due to the computation needed. It includes de-noising,
of course, and for voice, formant filtering, whjihc is rathe
reffective. Ther are voice, CW and data presents, and it is also
possible to adjust the stopband edges normally with front panel
controls. All in all, I have no complaint with how it works. And as I
mentioend, I may be able to stop it processing during trnamsissions by
hooking it up to the PTT line.
Now, in the SGC-2000 CLear Speech option (the same circuit) I believe
the DSP is bypoassed for 1. CW and 2. tuning. Obviously this is more
difficult when using a unit externally. Incidentally, it's not
available separately any more, and the lack of FCC Part 15
certification may be part of that. Also, other effective audio DPS
boxes are being sold now that are less costly to build (new
technolgy) and -- more important - cost us less to buy.
I have several Radio Shack DSP boxes, which are an interesting design.
I was at Tandy (doing EMI) when these were conceived and designed, and
the initial design had an 8-bit processor, which almost made
de-noising moot, as they had a higher noise floor than the receiver.
De-noising was never implemented in the RS DSP. But at theprice it is
a useful accessory anyway. I bought two of them for about $40 each
(half price sale) and the last one for $15!
Cortland
