Re: [softrock40] Re: High Definition 2.0 audio from Dell
I seem to have the same problem nulling the image with my Dell D620 using
Rocky. It has a Sigma Tel Audio soundcard.. How did you determine that it
had a sample offset at 192k?
----- Original Message -----
From: "Dave" <dmertz@...>
Sent: Friday, February 29, 2008 10:18 AM
Subject: [softrock40] Re: High Definition 2.0 audio from Dell
>> I need a new laptop. Dell seems to be saying that I can get one with
>> a "High Definition 2.0" soundcard. It seems also to say it has a
>> stereo line input. Does anyone know the specs on this card? (Khz and
> I'm not sure if it's the same, but my E1405, purchased about 8 months
> ago, has a SigmaTel High Definition Audio CODEC. It has a combo
> mic/stereo line iput jack. While the thing seems to support sampling
> at 192K, the spectrum looks like it's only 96K and I can't find any
> sound card setting to improve it. Also, there seems to be a constant
> one-sample-at-192K offset between Left and Right, making it nearly
> impossible to null the image. I ended up getting a Sound Blaster
> Live! 24 External to use with this machine. (Still a bit quirky
> because of Vista.)
> Dave, WA3OFF
> I seem to have the same problem nulling the image with my Dell D620My laptop machine has Vista and Rocky doesn't see any sound card, so
> using Rocky. It has a Sigma Tel Audio soundcard.. How did you
> determine that it had a sample offset at 192k?
I can't comment on whether or not Rocky is capable of using it at
192K with another OS like XP. Using KGKSDR Transceiver 1.1 and
ASIO4ALL I was able to operate it at 48K (ASIO4ALL limited to 48K)
and do a manual calibration that resulted in a steep phase curve.
Calculations showed the phase offset to be equivalent to one sample
time at 192K. I was able to see the same fixed time offset using
KGKSDR 1.0 Decoder 96K (does not use ASIO4ALL). Unfortunately,
KGKSDR won't auto null with such high phase offsets.
Winrad seems to work fine at 192K (although the sound card seems
silent beyond 96K) and I was able to get reasonable nulling with a one
sample offset at 192K.