Re: input/output sound level calibration

On Tue, 07 Mar 2006 14:12:48 +0100, Celer <card.lemoine@xxxxxxxxxx>
wrote:

Hello,

I'd like to know some objective thinks about the sound PC cards in general,
mine in particular (PCI64 with ens1370 chip), or a kind of what could be
considered as a standard one. That is the various max (although acceptable)
audio signal levels along the converting chain, but at least inputs (mic,
line) and output of the card(s). Is that any doc on the point ?
Also a mean for calibrating the chain (input -> ADC -> driver -> appli ->
driver -> DAC -> output) without sophisticated tools would be welcomed.

Thank you in advance.
Alain


Alas, there are no standards whatsoever, and Windows
has no driver mechanism to allow a card to provide
calibration information. The mixer controls (part of
the driver for the card) only work in "steps", but
the step size is not available and in fact is often not
constant throughout the range of a control.

I have been working on this issue for my upcoming
Daqarta for Windows. The basic idea is that with
a loopback, the software generates a signal and
adjusts the mixer while it monitors the input. This
allows it to determine the relative size of each step
in both the input and output mixer. Then to get an
absolute calibration, you need to provide a signal
with a known amplitude, and the software can then
tie everything together. This is easy if you have
test equipment, but the whole point of Daqarta is
to allow the computer and sound card to act as
a measurement and signal generator, so I'm working
on ways to provide the known input signal using
some inexpensive approach. This is complicated
by the fact that sound cards don't respond to DC,
so that limits the sort of things you can use.

My freeware DaqGen signal generator has a
way to calibrate the relative steps of the output
attenuators without any test equipment.at all.
The idea is to create the same signal on the left
and right channels, then move the left channel
down one mixer step. The right channel is now
louder, so you reduce the digital version of the
wave going to the right channel (while keeping its
mixer unchanged) until it matches the left exactly.
There are some tricks to doing this, but basically you
can get a pretty good match by inverting the polarity
of one channel and combining them together. When
the amplitudes are equal, they cancel completely.
Note however, that this only allows calibration of
relative step size, not absolute output.

Best regards,







Bob Masta
dqatechATdaqartaDOTcom

D A Q A R T A
Data AcQuisition And Real-Time Analysis
www.daqarta.com
Home of DaqGen, the FREEWARE signal generator
.

Relevant Pages

  • Re: input/output sound level calibration
    ... Nevertheless without the inner knowledge of the sound card we are a bit ... adjusts the mixer while it monitors the input. ... a measurement and signal generator, ... The right channel is now ...
    (comp.sys.ibm.pc.soundcard.tech)
  • Kismet and wi0
    ... between Kismet and the Orinoco Gold card. ... MPSAFE network stack disabled, ... # to set the channel the interface listens on. ... # filter_dump - Packets filtered at the dump level are tracked, displayed, ...
    (freebsd-questions)
  • Re: ath using hostap sets MTU to 2290 / channel 0 no longer works
    ... I have a system running pfSense, which is built on top of FreeBSD ... as it appears their code actually tries to set the MTU to 1500. ... I have an ath card I'm working with that sets its MTU to 1500 in hostap, ... show up unless they manually specify a channel. ...
    (freebsd-net)
  • Re: Device Driver Design Issues
    ... For example a serial card typically have ... Using a ioctl is fine ... > I have one card called VoicerPhone that has only one channel but I can ... > My approach in Windows (I alread have a device driver in Windows) is ...
    (comp.os.linux.development.system)
  • Re: kismet session
    ... does is put the card in monitor mode and then fire up pcap. ... I think kismet tries to decrypt the packets before logging ... As far as the channel thing - I've noticed a few strange things that might ... Check out our Advanced Hacking course, ...
    (Pen-Test)