Message-ID: <955F0C15833A459FA2557314E925534C@AGB>
From: "Graham" <g8fzk@g8fzk.fsnet.co.uk>
To: <rsgb_lf_group@blacksheep.org>
References: <4B0570B4.8040209@telia.com> <fc7eccec0911190929q187dbf39u96bdd4024d437810@mail.gmail.com>
In-Reply-To: <fc7eccec0911190929q187dbf39u96bdd4024d437810@mail.gmail.com>
Date: Thu, 19 Nov 2009 18:35:51 -0000
MIME-Version: 1.0
Importance: Normal
Subject: Re: LF: WSPR 2.0 is now available
Content-Type: text/plain;
	format=flowed;
	charset="iso-8859-1";
	reply-type=original
Content-Transfer-Encoding: 7bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org


All

I asked Joe  K1JT , this question a short time ago, below is the reply 

G .. 

> Q : Is there a  means to
> compensate  for the  sound  card  sampling rate? ,

No, but I think an even better feature.  The sample rate in WSPR 2.0 
will always be 48000 Hz.  This will prevent a problem that occurs in a 
few systems that insist on sampling at 44100 Hz and then resampling (in 
the driver) to the 12000 Hz requested by WSPR 1.2 ... and doing a *very* 
poor job of the resampling.

--------------------------------------------------
From: "Andy Talbot" <andy.g4jnt@googlemail.com>
Sent: Thursday, November 19, 2009 5:29 PM
To: <rsgb_lf_group@blacksheep.org>
Subject: Re: LF: WSPR 2.0 is now available

> 
> It has a calibration procedure built in, but as far as I can make out,
> is aimed at solving frequency offset / tuning errors.   I may have
> missed something, but there seems to be no facility to calibrate  for
> soundcard errors in the same way as WSJT does, which can affect symbol
> timing .  As we found in discussion on theis reflector a year or so
> ago.    Sample rate timing error is enough to destroy WSPR decoding
> even when signals are strong.
> 
> May be wrong, and will go and re read the appendices associated with
> calibration, but no mention is made, specifically, of soundcard
> sampling rate error - only refrences to frequency error.
> FOrtunately, as WSPR is designed around 12000Hz sampling, this is
> rarely out by more than a few ppm in most modern cards.   See my
> Column in the latest RadCom for some soundcard sampling rate
> measurements - and how bad some are
> 
> 
> Andy
> www.g4jnt.com
> 
> This email has been scanned for damaging side-effects by the health
> and safety police
> 
> 
> 
> 2009/11/19 Johan Bodin <jh.bodin@telia.com>:
>>
>>  http://physics.princeton.edu/pulsar/K1JT/
>>
>>
>> 73
>> Johan SM6LKM
>>
>>
>>
> 
>