Message-ID: <4B16B6A7.9000805@usa.net>
Date: Wed, 02 Dec 2009 19:49:11 +0100
From: Alberto di Bene <dibene@usa.net>
User-Agent: Thunderbird 2.0.0.23 (Windows/20090812)
MIME-Version: 1.0
To: rsgb_lf_group@blacksheep.org
References: <200912021848460547.0592D856@smtp2.numeo.fr>
In-Reply-To: <200912021848460547.0592D856@smtp2.numeo.fr>
Z-USANET-MsgId: XID000NLBsxo9040X39
Subject: Re: LF: Sound card sampling rate
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Sender: owner-rsgb_lf_group@blacksheep.org
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org


John RABSON wrote:

> I am running a USB-connected Creative Labs sound blaster model SB0270 with Argo V1 build 134 on an XP SP3 system.
 > How do I find out the sampling rate, please?

The sampling rate of a sound card up to Windows 98 used to be set at the hardware level by the program that
used the card. The program called a Windows API telling it that it should set the card, let's say, at 11025 Hz
as sampling frequency. If the card was capable of that rate, ok, otherwise Windows came back with an error
to the requesting program.

Windows XP changed this scenario. If the card is not capable at the hardware level to support a specific
sampling rate, then Windows silently does a under-the-hood software resampling, so that to please the
requesting program, no matter what the requested sampling rate is.

Many manufacturers of cheap sound cards or sound chips took advantage of this, designing sound cards that
were capable of 48 kHz only, then relying on Windows to provide the other rates.
Needles to say, the resampling done by Windows is far from optimal... that's why the manufacturers of
sound cards intended for serious musical applications (studio recording, etc.) do clearly specify in the
specs of the card that the various sampling rates are 'hardware sampling rates', no Windows involved...

That said, and coming to Argo, when you are in full band view mode, Argo sets the sampling frequency at 11025 Hz,
while in one of the QRSS modes it is set to 5125 Hz. In CW mode it is set to 22050 Hz.

When I will find the time to rejuvenate Argo, I will change that. I will ask Windows to sample at 48 kHz, and
this will give me the real hardware sampling rate, then I will use the fractional ratio resampling code that
I wrote for Winrad, which works better than Windows resampling, to resample down to the needed sampling rate.

But, as said, this will happen when I will find the time....

73  Alberto  I2PHD