advise please on filtering audio

Hi, I'm just a common enjoyer so bear with me on any advise you give.

I record fm and although I never hear any crackling, the recording does
exhibit crackling. I record in 44.1hz 16b 2c.

Do you have a recommendation on filtering this crackling.  I will try to
explain what I am seeing.

above 15hz you can see the bubbling effect on the spectrum.  Of course
this is inaudible but apparently it is part of the crackling down thru
the audible portion but you can't readily see it within the true audio.

Apparently this is what everyone calls aliasing.

My take is, I should lowpass above the 15khz on recording.  correct.

And, how do I get rid of it after recorded, or is it possible.  I use
goldwave for my editing.



-- 
Phil

Phil wrote:

> Hi, I'm just a common enjoyer so bear with me on any advise you give.
> 
> I record fm and although I never hear any crackling, the recording does
> exhibit crackling. I record in 44.1hz 16b 2c.
> 
> Do you have a recommendation on filtering this crackling.  I will try to
> explain what I am seeing.
> 
> above 15hz you can see the bubbling effect on the spectrum.  Of course
> this is inaudible but apparently it is part of the crackling down thru
> the audible portion but you can't readily see it within the true audio.
> 
> Apparently this is what everyone calls aliasing.
> 
> My take is, I should lowpass above the 15khz on recording.  correct.
> 
> And, how do I get rid of it after recorded, or is it possible.  I use
> goldwave for my editing.

Phil,

If your program material has components as high as half the sample rate,
they need to be removed before sampling. Many codecs do that internally,
and FM broadcasts don't include such frequencies. It's possible that
aliasing is not your problem. Have you tried recording with the volume
set a little lower?

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

I know fm is lowpassed at 15khz and it is apparent from the recordings.
But I do suspect the occasional crackling is electrical interference
from the building, which I can't prevent.

the noticeable shape of the crackling is seen above the 15khz because
that is the quiet portion.

It looks like even sequential bubbles maybe 2 or 3 right up the
frequency range to about 22kz.


Jerry Avins <jya@ieee.org> posted in
news:4040bbbb$0$3105$61fed72c@news.rcn.com

> Phil wrote:
>
> Phil,
>
> If your program material has components as high as half the sample
> rate, they need to be removed before sampling. Many codecs do that
> internally, and FM broadcasts don't include such frequencies. It's
> possible that aliasing is not your problem. Have you tried recording
> with the volume set a little lower?
>
> Jerry

-- 
Phil

Phil wrote:

> I know fm is lowpassed at 15khz and it is apparent from the recordings.
> But I do suspect the occasional crackling is electrical interference
> from the building, which I can't prevent.
> 
> the noticeable shape of the crackling is seen above the 15khz because
> that is the quiet portion.
> 
> It looks like even sequential bubbles maybe 2 or 3 right up the
> frequency range to about 22kz.

   ..

What is a bubble? Received noise may not be properly band limited for
your sampler, but most codecs have an anti-alias filter at the input. In
the other hand, if too large a signal overloads the input, several
different faults can develop, most of them generating high frequencies.
Did you try recording at a reduced level?.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

Jerry Avins wrote:

> (snip)


> If your program material has components as high as half the sample rate,
> they need to be removed before sampling. Many codecs do that internally,
> and FM broadcasts don't include such frequencies. It's possible that
> aliasing is not your problem. Have you tried recording with the volume
> set a little lower?

FM broadcasts have a 19kHz pilot, 38kHz stereo subcarrier,
and possibly a 67kHz FM SCA subcarrier, all above 15kHz.

Usually these should be well filtered by a receiver, but
they are part of the FM broadcast signal.

-- glen

glen herrmannsfeldt wrote:

> Jerry Avins wrote:
> 
>> (snip)
> 
> 
> 
>> If your program material has components as high as half the sample rate,
>> they need to be removed before sampling. Many codecs do that internally,
>> and FM broadcasts don't include such frequencies. It's possible that
>> aliasing is not your problem. Have you tried recording with the volume
>> set a little lower?
> 
> 
> FM broadcasts have a 19kHz pilot, 38kHz stereo subcarrier,
> and possibly a 67kHz FM SCA subcarrier, all above 15kHz.
> 
> Usually these should be well filtered by a receiver, but
> they are part of the FM broadcast signal.
> 
> -- glen

Those are part of the detector output, but shouldn't be part of the
audio. Nevertheless, what ought to be and what is often differ. I once
thought I had found an oscillating power amplifier. Even during periods
of silence, there were several watts of output well above the audio
range. My first hint of the truth was the problem's existing only when
listening to AM. It turned out to be the 456 KHz IF. The audiophile
customer had clipped out a capacitor to avoid diagonal clipping.

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

Sorry, I meant to say bubble like on a spectrum graph.

Another way to describe it is like a scope output of a simple ac to dc
bridge rectifier only more rounded then sinesoidal.

It sounds like crackling.  It is intermitant, probably from someone's
appliance in the building.


Jerry Avins <jya@ieee.org> posted in
news:4040da5c$0$3069$61fed72c@news.rcn.com

> Phil wrote:
>
>> I know fm is lowpassed at 15khz and it is apparent from the
>> recordings. But I do suspect the occasional crackling is electrical
>> interference from the building, which I can't prevent.
>>
>> the noticeable shape of the crackling is seen above the 15khz because
>> that is the quiet portion.
>>
>> It looks like even sequential bubbles maybe 2 or 3 right up the
>> frequency range to about 22kz.
>
>    ..
>
> What is a bubble? Received noise may not be properly band limited for
> your sampler, but most codecs have an anti-alias filter at the input.
> In the other hand, if too large a signal overloads the input, several
> different faults can develop, most of them generating high
> frequencies. Did you try recording at a reduced level?.
>
> Jerry

-- 
Phil

Phil wrote:

> Sorry, I meant to say bubble like on a spectrum graph.

Sorry, that doesn't come through to me.

> Another way to describe it is like a scope output of a simple ac to dc
> bridge rectifier only more rounded then sinesoidal.

Rectifier without filter capacitor, I assume. In other words, a rounded
pulse.

> It sounds like crackling.  It is intermitant, probably from someone's
> appliance in the building.

Do you hear the crackling in the received music, or only in the
recording? If recording only, what tells you it's external
interference?

   ...

Jerry
-- 
Engineering is the art of making what you want from things you can get.
�����������������������������������������������������������������������

Jerry Avins wrote:

> Phil wrote:
> 
>> Sorry, I meant to say bubble like on a spectrum graph.
> 
> Sorry, that doesn't come through to me.

In the spectrogram, a sonic event in a quiet spectral region appears 
as a differently-colored blob. That's probably what Phil meant.


Martin

-- 
Quidquid latine dictum sit, altum viditur.

If you can clearly hear the crackling, it likely has components well below
15kHz.  The reason they may only show up above 15k is that there is nothing
else there so it is most obvious.  But there are probably components in the
lower frequencies as well.

You can test your method by getting a copy demo copy of CoolEdit and doing
the filtering.

"Phil" <stuff@basketball.net> wrote in message
news:UP10c.2393$ea6.665369762@newssvr11.news.prodigy.com...
> Hi, I'm just a common enjoyer so bear with me on any advise you give.
>
> I record fm and although I never hear any crackling, the recording does
> exhibit crackling. I record in 44.1hz 16b 2c.
>
> Do you have a recommendation on filtering this crackling.  I will try to
> explain what I am seeing.
>
> above 15hz you can see the bubbling effect on the spectrum.  Of course
> this is inaudible but apparently it is part of the crackling down thru
> the audible portion but you can't readily see it within the true audio.
>
> Apparently this is what everyone calls aliasing.
>
> My take is, I should lowpass above the 15khz on recording.  correct.
>
> And, how do I get rid of it after recorded, or is it possible.  I use
> goldwave for my editing.
>
>
>
> -- 
> Phil
>
>