Message-ID: <003e01c0c800$0d3eab60$0700000a@parissn2>
From: "Stewart Nelson" <sn@scgroup.com>
To: rsgb_lf_group@blacksheep.org
References: <3.0.1.16.20010418112245.218f3f10@mail.cc.kuleuven.ac.be>
Subject: Re: LF: WOLF test results, soundcard trouble
Date: Wed, 18 Apr 2001 14:06:49 +0200
Organization: SC Group
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org
Sender: <majordom@post.thorcom.com>

Hi all,

Thanks for the interesting posts on frequency errors in sound cards and
drivers.

It appears that WOLF should use 11025 Hz, rather than 8000 Hz, as its
standard sample rate.  Upsampling or downsampling by a factor of 4
should pose no problems for drivers, and I would guess that hardware
sample rate generators would also be accurate, limited only by the
oscillator's calibration error and drift.

Version 0.53 can handle 11025 Hz files on receive, but there is currently
no way to set that value in the .wav header on Tx.  I'll add that feature
if it will be useful.  Let me know what actual Tx and Rx sample rates
are seen from your card at 11025 Hz nominal.

Rik's idea of bypassing the card is another good solution.
One alternative that already exists is the VE2IQ interface.  See
http://cafe.rapidus.net/bill1/bbs.htm and download SDPCX.ZIP .
This is a home brew Sigma-Delta A/D which drives a serial port.
Many PSK and CCW programs can use it, and it has also been used with WOLF.
You can buy kits from Bill for US$54, or build your own from scratch.
Sample rate is a precise 7200 Hz.  Resolution is only 8 bits,
but IMO that is plenty, unless the audio signal is dominated
by a strong interfering carrier.  If most of the audio power is
random noise and/or static crashes, then even 1 bit is pretty good!
Steve VK2ZTO has shown impressive results sampling this way.

If you really need high resolution and linearity, there are several USB
audio systems on the market.  They offer up to 20 bits at 96 kHz, costing
~$150.  I've never tested one, but they are popular in the "semi-pro"
marketplace, so I would guess that linearity and noise is pretty good.

73,

Stewart KK7KA