I wanted to post this because it's an obscure problem and may have
happened to other people. I have been using an AKG CK80 phantom-powered
microphone for some time in a PA system with no problems. Recently we
started recording the audio using a PC, but the recordings from the AKG
have a low-level whistle at about 4KHz. All our other sources do not show
this problem (SM58, piano, DI etc)

I've found out what the problem is: the AKG is generating a low level
signal at about 43.8KHz which is aliasing in the soundcard down to audio.
It is likely that the AKG has a switching DC/DC converter running at
43.8KHz.

Actually IMO it's not the AKG which is at fault- really the soundcard ADC
should have a good enough antialias filter to eliminate everything above
half the sample rate, but in this case it doesn't. To solve the problem it
is necessary to add extra lowpass filtering at 43.8KHz.

Hope this posting will help if anyone else is getting this problem...

Robert Bristow wrote:
I wanted to post this because it's an obscure problem and may have
happened to other people. I have been using an AKG CK80 phantom-powered
microphone for some time in a PA system with no problems. Recently we
started recording the audio using a PC, but the recordings from the AKG
have a low-level whistle at about 4KHz. All our other sources do not show
this problem (SM58, piano, DI etc)

I've found out what the problem is: the AKG is generating a low level
signal at about 43.8KHz which is aliasing in the soundcard down to audio.
It is likely that the AKG has a switching DC/DC converter running at
43.8KHz.

Actually IMO it's not the AKG which is at fault- really the soundcard ADC
should have a good enough antialias filter to eliminate everything above
half the sample rate, but in this case it doesn't. To solve the problem it
is necessary to add extra lowpass filtering at 43.8KHz.

This isn't all that obscure or unusual, and is common for a lot of mikes
that use DC-DC converters to have some ultrasonic leakage. The A-T 4051
is definitely worse than the C460 which is worse than the C480. The B&K
2034 supply produces junk around 21 KHz.

What preamp and soundcard are you using? This is the sign of a severely
screwed-up soundcard or preamp, though. This is a common enough issue with
mikes that competently designed preamps and converters should have no problem
with it.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

"Scott Dorsey" wrote in message

Robert Bristow wrote:

I wanted to post this because it's an obscure problem
and may have happened to other people. I have been using
an AKG CK80 phantom-powered microphone for some time in
a PA system with no problems. Recently we started
recording the audio using a PC, but the recordings from
the AKG have a low-level whistle at about 4KHz. All our
other sources do not show this problem (SM58, piano, DI
etc)

I've found out what the problem is: the AKG is
generating a low level signal at about 43.8KHz which is
aliasing in the soundcard down to audio. It is likely
that the AKG has a switching DC/DC converter running at
43.8KHz.

If your audio interface is this flakey, I'm a little surprised that you
aren't also picking up similar junk from your display's power supply whether
CRT or LCD.

Actually IMO it's not the AKG which is at fault- really
the soundcard ADC should have a good enough antialias
filter to eliminate everything above half the sample
rate, but in this case it doesn't. To solve the problem
it is necessary to add extra lowpass filtering at
43.8KHz.

This isn't all that obscure or unusual, and is common for
a lot of mikes that use DC-DC converters to have some
ultrasonic leakage. The A-T 4051
is definitely worse than the C460 which is worse than the
C480. The B&K 2034 supply produces junk around 21 KHz.

What preamp and soundcard are you using? This is the
sign of a severely screwed-up soundcard or preamp,
though.

Agreed. I remember trying to muscle some ultrasonics through the line input
of an Turtle Beach Pinnacle back in the late 90's. I applied the TTL
square-wave output of my analog signal generator set to 100 KHz. It was like
nothing was happening.

This is a common enough issue with mikes that
competently designed preamps and converters should have
no problem with it.

There have been both Echo and M-Audio audio interfaces that had these
problems. For the M-Audio cards a new driver sufficed.