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

RE: [linrad] Big surprise!



Hello Leif and all,

At 23:59 02/09/13 +0200, SM5BSZ wrote:

[partly snipped]

>Hmmm, remember SM5FRH has a VERY BIG antenna..........
>
>However, as you surely have noticed extremely few stations are so weak
>that you can not copy easily at much lower levels. Use the peak S-meter
>to find the maximum during a 1 minute sequence. This number stays nearly
>constant for each station (unless there is an antenna positioning or tx
>error). You will find that practically all stations can be copied at
>8dB lower S/N (get N from the S-meter on the noise floor)
>
>8dB below SM5FRH is a normal 4-yagi station:-)

I know that SM5FRH is one of two big guns :-)
It's good to know how big gun detects that many stations from moon.
What I try to do is to select week one in the band (in recorded file) as 
possible.

>> My 500MHz Pentium III with 384MB ram shows rather high CPU load.
>> I observed around 90% by 'ps aux' command in remote login shell.
>> I wonder if this is the case in your system too?
>Have you tried the different FFT versions ? With Pentium III you can
>use MMX and SIMD instructions that run a good deal faster than ordinary 
>floating point.

Do you mean 'FFT version' in parameter setup?
I've been using version 2 for both forward and 2nd forward.
I have tried to increase the first one up to 5 but no significant change.

>You need them to process 2 x 96 kHz bandwidth.

Sorry, I don't understand this.

Well, I found the CPU load remains almost same (heavy) as it's running
during set up window.
In other words, about CPU load of 80% doesn't change in running phase
nor setup phase.
This load slowly increases from about 70% at the startup
up to more than 90% finally.

Will you please let me know of your reported CPU load of 'ps aux' command?

Thank you very much for your help.

Kohjin - JR1EDE


LINRADDARNIL
p