I'm starting to transfer my record collection to MP3 format, and am having
a problem with hum. The hum is not there while the record is playing, but
when I play back the resulting file it is there. As a test, I tried it
with everything disconnected (nothing plugged into the line in), and it
still happens, so it must be something about the recording process.

Frequency analysis (using CoolEdit) of such "silent" tracks shows a strong
peak at 4.1 kHz, with smaller (harmonic?) peaks at 8.2, 12.3 and 16.4, as
well as another large one out around 17.5. The right channel has
additional peaks at 7.5 and 13.5 (and the 8.2 and 12.3 peaks are smaller),
while the left has an additional peak at 9. To give an idea of the
magnitude, at 4 kHz the background is around -102dB, and at 4.1 the peak
goes up to as high as -81dB.

I read some of the earlier thread about grounding, but I don't think that's
my problem since the hum is not there normally, only when playing back
recorded material. I'm guessing that it must be to do with the A/D
conversion during recording.

So, I am wondering whether there is anything I can do about this (e.g. is
there a better sample rate than 44.1 kHz to record at, or some other
variable I don't know enough about to consider) or if it is caused by the
(crappy?) sound card built into the motherboard (it's a C-Media
CMI8738/C3DX, in case anyone is familiar with it). I have another computer
with a Sound Blaster Live in it, would it be worth trying to record on it
instead? (That system is otherwise woefully underpowered compared to the
one with the built-in sound card.)

--
Greg Schmidt
Trawna Publications http://www.trawna.com/

On Sat, 14 Aug 2004 03:33:45 -0400, Greg Schmidt
wrote:

So, I am wondering whether there is anything I can do about this (e.g. is
there a better sample rate than 44.1 kHz to record at, or some other
variable I don't know enough about to consider) or if it is caused by the
(crappy?) sound card built into the motherboard (it's a C-Media
CMI8738/C3DX, in case anyone is familiar with it). I have another computer
with a Sound Blaster Live in it, would it be worth trying to record on it
instead? (That system is otherwise woefully underpowered compared to the
one with the built-in sound card.)


I think you have to point the finger at your crappy sound "card".

You could disable onboard sound and put a SB in this machine. You
should find a second-hand Live on ebay very cheaply. Or, if you want
to dip a toe into semi-pro audio quality, look at the M-Audio
Audiophile 2496, currently being sold new at a very attractive price.

Why not borrow the Live out of the other computer and experiment?

In article ,
Greg Schmidt wrote:

I'm starting to transfer my record collection to MP3 format, and am having
a problem with hum. The hum is not there while the record is playing, but
when I play back the resulting file it is there. As a test, I tried it
with everything disconnected (nothing plugged into the line in), and it
still happens, so it must be something about the recording process.

Frequency analysis (using CoolEdit) of such "silent" tracks shows a strong
peak at 4.1 kHz, with smaller (harmonic?) peaks at 8.2, 12.3 and 16.4, as
well as another large one out around 17.5. The right channel has
additional peaks at 7.5 and 13.5 (and the 8.2 and 12.3 peaks are smaller),
while the left has an additional peak at 9. To give an idea of the
magnitude, at 4 kHz the background is around -102dB, and at 4.1 the peak
goes up to as high as -81dB.

I read some of the earlier thread about grounding, but I don't think that's
my problem since the hum is not there normally, only when playing back
recorded material. I'm guessing that it must be to do with the A/D
conversion during recording.

So, I am wondering whether there is anything I can do about this (e.g. is
there a better sample rate than 44.1 kHz to record at, or some other
variable I don't know enough about to consider) or if it is caused by the
(crappy?) sound card built into the motherboard (it's a C-Media
CMI8738/C3DX, in case anyone is familiar with it). I have another computer
with a Sound Blaster Live in it, would it be worth trying to record on it
instead? (That system is otherwise woefully underpowered compared to the
one with the built-in sound card.)

It may be a design problem inside your computer that can't be fixed.
Try a different input device. Maybe borrow one for testing.

"Greg Schmidt" wrote in message

]
I'm starting to transfer my record collection to MP3 format, and am
having a problem with hum. The hum is not there while the record is
playing, but when I play back the resulting file it is there.


As a
test, I tried it with everything disconnected (nothing plugged into
the line in), and it still happens, so it must be something about the
recording process.

I've never seen such a thing.

Frequency analysis (using CoolEdit) of such "silent" tracks shows a
strong peak at 4.1 kHz, with smaller (harmonic?) peaks at 8.2, 12.3
and 16.4, as well as another large one out around 17.5. The right
channel has additional peaks at 7.5 and 13.5 (and the 8.2 and 12.3
peaks are smaller), while the left has an additional peak at 9. To
give an idea of the magnitude, at 4 kHz the background is around
-102dB, and at 4.1 the peak goes up to as high as -81dB.

4 KHz isn't a hum, subjectively it's a high whistle. Are you sure the thing
that is bothering you is a hum? Hum in your case would be about 50 Hz.

On Sat, 14 Aug 2004 07:03:53 -0400, Arny Krueger wrote:

"Greg Schmidt" wrote in message

]
I'm starting to transfer my record collection to MP3 format, and am
having a problem with hum. The hum is not there while the record is
playing, but when I play back the resulting file it is there.

As a
test, I tried it with everything disconnected (nothing plugged into
the line in), and it still happens, so it must be something about the
recording process.

I've never seen such a thing.

Googling for 4.1 kHz shows that it seems to be a common frequency for audio
DSP components. I doubt that this is a coincidence! I guess C-Media used
sub-standard components.

Frequency analysis (using CoolEdit) of such "silent" tracks shows a
strong peak at 4.1 kHz, with smaller (harmonic?) peaks at 8.2, 12.3
and 16.4, as well as another large one out around 17.5. The right
channel has additional peaks at 7.5 and 13.5 (and the 8.2 and 12.3
peaks are smaller), while the left has an additional peak at 9. To
give an idea of the magnitude, at 4 kHz the background is around
-102dB, and at 4.1 the peak goes up to as high as -81dB.

4 KHz isn't a hum, subjectively it's a high whistle. Are you sure the thing
that is bothering you is a hum? Hum in your case would be about 50 Hz.

Yeah, high whistle is probably a better term. For some reason, I thought
"hum" was kind of a standard word for all such semi-constant background
noise.

Thanks to everyone for their assistance. It appears that my suspicion of
the sound "card" was well-warranted. I'll see what I can do to use
something better, and report back with my findings.

--
Greg Schmidt
Trawna Publications http://www.trawna.com/

"Greg Schmidt" wrote in message

On Sat, 14 Aug 2004 07:03:53 -0400, Arny Krueger wrote:

"Greg Schmidt" wrote in message

]
I'm starting to transfer my record collection to MP3 format, and am
having a problem with hum. The hum is not there while the record is
playing, but when I play back the resulting file it is there.

As a
test, I tried it with everything disconnected (nothing plugged into
the line in), and it still happens, so it must be something about
the recording process.

I've never seen such a thing.

Googling for 4.1 kHz shows that it seems to be a common frequency for
audio DSP components. I doubt that this is a coincidence! I guess
C-Media used sub-standard components.

Actually, I'm quite familiar with CMedia products. While there was a time
that they were unbelievable junk, they've improved to the point where good
implmentations of their latest chips can be relatively free of audible
defects.

However, they make chips, not finished audio gear, and the manufacturer who
assembles their chips into products could manage to **** in the soup.

Arny Krueger wrote:

Actually, I'm quite familiar with CMedia products. While there was a time
that they were unbelievable junk, they've improved to the point where good
implmentations of their latest chips can be relatively free of audible
defects.

However, they make chips, not finished audio gear, and the manufacturer who
assembles their chips into products could manage to **** in the soup.

So, Arnie, saw "War Of The Roses" recently?

You know a film is dark when Danny deVito is a lawyer *and* the good guy.

ObTastelessAudio: imagining the Foley for the kitchen scene.


Francois.

Arny Krueger wrote:

"Greg Schmidt" wrote in message


Googling for 4.1 kHz shows that it seems to be a common frequency for
audio DSP components. I doubt that this is a coincidence! I guess
C-Media used sub-standard components.

Actually, I'm quite familiar with CMedia products. While there was a time
that they were unbelievable junk, they've improved to the point where good
implmentations of their latest chips can be relatively free of audible
defects.

However, they make chips, not finished audio gear, and the manufacturer who
assembles their chips into products could manage to **** in the soup.

I think this mobo suffers from both afflictions ;-)

It has a C Media implementation.

There's audible 'birdies' for lack of a better desciption in the background
constantly on the line outs that varies with graphics card operation and I
personally doubt that the linearity is much good. Checking out Neutrik's
Minirator emulation application, I get awful distortion on the sinewave out
varying with signal level.

Only upside is that PC Chips had a digital I/O option which I have that
bypasses the crippled analogue section.


Graham

Googling for 4.1 kHz shows that it seems to be a common frequency for
audio
DSP components. I doubt that this is a coincidence! I guess C-Media used
sub-standard components.

More likely a bad layout. For all the agonizing among audiophiles over
technologies, poor layout has as much to do with ultimate sound quality than
just about anything else.

Karl Uppiano wrote:

Googling for 4.1 kHz shows that it seems to be a common frequency for
audio
DSP components. I doubt that this is a coincidence! I guess C-Media used
sub-standard components.

More likely a bad layout. For all the agonizing among audiophiles over
technologies, poor layout has as much to do with ultimate sound quality than
just about anything else.

Attempting a good pcb layout for audio on a mobo is pretty much a thankless task
doomed to indifferent performance.

Graham

"Pooh Bear" wrote in message
...


Karl Uppiano wrote:

Googling for 4.1 kHz shows that it seems to be a common frequency for
audio
DSP components. I doubt that this is a coincidence! I guess C-Media
used
sub-standard components.

More likely a bad layout. For all the agonizing among audiophiles over
technologies, poor layout has as much to do with ultimate sound quality
than
just about anything else.

Attempting a good pcb layout for audio on a mobo is pretty much a
thankless task
doomed to indifferent performance.

Graham



Putting audio on a mobo might be an *example* of a poor layout :-)

"Karl Uppiano" wrote in message
news
"Pooh Bear" wrote in message
...


Karl Uppiano wrote:

Googling for 4.1 kHz shows that it seems to be a common frequency
for audio
DSP components. I doubt that this is a coincidence! I guess
C-Media used
sub-standard components.

More likely a bad layout. For all the agonizing among audiophiles
over technologies, poor layout has as much to do with ultimate
sound quality than
just about anything else.

Attempting a good pcb layout for audio on a mobo is pretty much a
thankless task
doomed to indifferent performance.


Putting audio on a mobo might be an *example* of a poor layout :-)

It can and has been done *right*. I would call *right*, having dynamic range
that is limited by the chip being used to implement the audio interface.

Audio interfaces on motherboards have worked their way up over the years.
The first ones had about 30 dB dynamic range. Then they jumped up to about
50 dB, and now top out around 80 dB. Realtek claims that they have an
on-board chip with 90 dB dynamic range. I'll believe it when I see it, even
though their 80 dB chip is really pretty good.

Arny Krueger wrote:

"Karl Uppiano" wrote in message
news
"Pooh Bear" wrote in message
...


Karl Uppiano wrote:

Googling for 4.1 kHz shows that it seems to be a common frequency
for audio
DSP components. I doubt that this is a coincidence! I guess
C-Media used
sub-standard components.

More likely a bad layout. For all the agonizing among audiophiles
over technologies, poor layout has as much to do with ultimate
sound quality than
just about anything else.

Attempting a good pcb layout for audio on a mobo is pretty much a
thankless task
doomed to indifferent performance.

Putting audio on a mobo might be an *example* of a poor layout :-)

It can and has been done *right*. I would call *right*, having dynamic range
that is limited by the chip being used to implement the audio interface.

The chip isn't the problem at all.

You need to implement a good interface to the 'real world' . That's where the
problems arise.

Audio interfaces on motherboards have worked their way up over the years.
The first ones had about 30 dB dynamic range. Then they jumped up to about
50 dB, and now top out around 80 dB. Realtek claims that they have an
on-board chip with 90 dB dynamic range. I'll believe it when I see it, even
though their 80 dB chip is really pretty good.

It's all about avoiding 'ground noise' from other components on the mobo.

I'm sure it can be acheived fairly competently but I won't be holding my breath
!


Graahm

"Pooh Bear" wrote in message


Arny Krueger wrote:

"Karl Uppiano" wrote in message
news

Putting audio on a mobo might be an *example* of a poor layout :-)

It can and has been done *right*. I would call *right*, having
dynamic range that is limited by the chip being used to implement
the audio interface.

The chip isn't the problem at all.

OK.

You need to implement a good interface to the 'real world' . That's
where the problems arise.

It appears to be a solved problem, at least from time to time.

Audio interfaces on motherboards have worked their way up over the
years. The first ones had about 30 dB dynamic range. Then they
jumped up to about 50 dB, and now top out around 80 dB. Realtek
claims that they have an on-board chip with 90 dB dynamic range.
I'll believe it when I see it, even though their 80 dB chip is
really pretty good.

It's all about avoiding 'ground noise' from other components on the
mobo.

Agreed.

I'm sure it can be acheived fairly competently but I won't be holding
my breath !

As you please. However, the worst thing about a motherboard is probably all
the 5 volt square waves, a problem that is shared with a circuit card inside
a high end DAC.

"Arny Krueger" wrote in message
...
"Pooh Bear" wrote in message


Arny Krueger wrote:

"Karl Uppiano" wrote in message
news

Putting audio on a mobo might be an *example* of a poor layout :-)

It can and has been done *right*. I would call *right*, having
dynamic range that is limited by the chip being used to implement
the audio interface.

The chip isn't the problem at all.

OK.

You need to implement a good interface to the 'real world' . That's
where the problems arise.

It appears to be a solved problem, at least from time to time.

Audio interfaces on motherboards have worked their way up over the
years. The first ones had about 30 dB dynamic range. Then they
jumped up to about 50 dB, and now top out around 80 dB. Realtek
claims that they have an on-board chip with 90 dB dynamic range.
I'll believe it when I see it, even though their 80 dB chip is
really pretty good.

It's all about avoiding 'ground noise' from other components on the
mobo.

Agreed.

I'm sure it can be acheived fairly competently but I won't be holding
my breath !

As you please. However, the worst thing about a motherboard is probably
all
the 5 volt square waves, a problem that is shared with a circuit card
inside
a high end DAC.


As I said in my original post about layout, it's all about layout. I'm quite
certain it's harder, but not impossible to get a good audio layout on a
mobo. It's much easier on a dedicated sound card, although bad power/ground
distribution on the mobo can jinx that, too.

"Karl Uppiano" wrote in message

"Arny Krueger" wrote in message
...
"Pooh Bear" wrote in message


Arny Krueger wrote:

"Karl Uppiano" wrote in message
news

Putting audio on a mobo might be an *example* of a poor layout :-)

It can and has been done *right*. I would call *right*, having
dynamic range that is limited by the chip being used to implement
the audio interface.

The chip isn't the problem at all.

OK.

You need to implement a good interface to the 'real world' . That's
where the problems arise.

It appears to be a solved problem, at least from time to time.

Audio interfaces on motherboards have worked their way up over the
years. The first ones had about 30 dB dynamic range. Then they
jumped up to about 50 dB, and now top out around 80 dB. Realtek
claims that they have an on-board chip with 90 dB dynamic range.
I'll believe it when I see it, even though their 80 dB chip is
really pretty good.

It's all about avoiding 'ground noise' from other components on the
mobo.

Agreed.

I'm sure it can be acheived fairly competently but I won't be
holding my breath !

As you please. However, the worst thing about a motherboard is
probably all
the 5 volt square waves, a problem that is shared with a circuit card
inside
a high end DAC.


As I said in my original post about layout, it's all about layout.
I'm quite certain it's harder, but not impossible to get a good audio
layout on a mobo. It's much easier on a dedicated sound card,
although bad power/ground distribution on the mobo can jinx that, too.

Then we are in perfect agreement.

Have a nice weekend, what's left of it!

Karl Uppiano wrote:

As I said in my original post about layout, it's all about layout. I'm quite
certain it's harder, but not impossible to get a good audio layout on a
mobo.

Yup.

It's much easier on a dedicated sound card, although bad power/ground
distribution on the mobo can jinx that, too.

I would second that.

Graham

Arny Krueger wrote:

"Pooh Bear" wrote in message


I'm sure it can be acheived fairly competently but I won't be holding
my breath !

As you please. However, the worst thing about a motherboard is probably all
the 5 volt square waves, a problem that is shared with a circuit card inside
a high end DAC.

When there's a high level of VHF / UHF interference, I wouldn't like to have to
deal with that.

In practice, I've found that stand-alone DSP cards can be made where the
'digital noise'can be made insignificant.

Clean grounds mean everything though - hence mobo integration offers issues that
are problematical.

I could probably host a seminar about grounding for DSP !


Graham

"Pooh Bear" wrote in message
...


Arny Krueger wrote:

"Pooh Bear" wrote in message


I'm sure it can be acheived fairly competently but I won't be holding
my breath !

As you please. However, the worst thing about a motherboard is probably
all
the 5 volt square waves, a problem that is shared with a circuit card
inside
a high end DAC.

When there's a high level of VHF / UHF interference, I wouldn't like to
have to
deal with that.

In practice, I've found that stand-alone DSP cards can be made where the
'digital noise'can be made insignificant.

Clean grounds mean everything though - hence mobo integration offers
issues that
are problematical.

I could probably host a seminar about grounding for DSP !


Graham



They don't mention audio (yet) but you gotta see this. A Feng Shui type
motherboard.

LOL !

http://www.bbspot.com/News/2004/07/f...therboard.html


Mark Z.

"Mark D. Zacharias" wrote in message



They don't mention audio (yet) but you gotta see this. A Feng Shui
type motherboard.

LOL !

http://www.bbspot.com/News/2004/07/f...therboard.html

The case must be a real trip with the PCI slots like that.

"Greg Schmidt" wrote in message
...
I'm starting to transfer my record collection to MP3 format, and am having
a problem with hum. The hum is not there while the record is playing, but
when I play back the resulting file it is there. As a test, I tried it
with everything disconnected (nothing plugged into the line in), and it
still happens, so it must be something about the recording process.

Frequency analysis (using CoolEdit) of such "silent" tracks shows a strong
peak at 4.1 kHz, with smaller (harmonic?) peaks at 8.2, 12.3 and 16.4, as
well as another large one out around 17.5. The right channel has
additional peaks at 7.5 and 13.5 (and the 8.2 and 12.3 peaks are smaller),
while the left has an additional peak at 9. To give an idea of the
magnitude, at 4 kHz the background is around -102dB, and at 4.1 the peak
goes up to as high as -81dB.

I read some of the earlier thread about grounding, but I don't think
that's
my problem since the hum is not there normally, only when playing back
recorded material. I'm guessing that it must be to do with the A/D
conversion during recording.

So, I am wondering whether there is anything I can do about this (e.g. is
there a better sample rate than 44.1 kHz to record at, or some other
variable I don't know enough about to consider) or if it is caused by the
(crappy?) sound card built into the motherboard (it's a C-Media
CMI8738/C3DX, in case anyone is familiar with it). I have another
computer
with a Sound Blaster Live in it, would it be worth trying to record on it
instead? (That system is otherwise woefully underpowered compared to the
one with the built-in sound card.)

--
Greg Schmidt
Trawna Publications http://www.trawna.com/

Open your recording settings and be sure other devices are disabled such as
the microphone.
John

On Sun, 15 Aug 2004 23:44:14 GMT, jriegle wrote:

Open your recording settings and be sure other devices are disabled such as
the microphone.

That is an excellent suggestion. In fact, I meant to write in my original
post that I had done that, but, my short-term memory being what it is (or,
rather, isn't), by the time I was done writing I had forgotten to include
it. However, that is an excellent suggestion. :-)

--
Greg Schmidt
Trawna Publications http://www.trawna.com/

On Sat, 14 Aug 2004 14:08:54 -0400, Greg Schmidt wrote:

Thanks to everyone for their assistance. It appears that my suspicion of
the sound "card" was well-warranted. I'll see what I can do to use
something better, and report back with my findings.

I've done some limited tests on the system with the SB Live and the early
results appear much better. There is now only a single peak of noise,
about 15dB high, at 5.5 kHz. The audible effects of this are virtually
nil, certainly much less than the effects of the former 20+dB high peak at
4.1. Are these results likely to change substantially if I were to move
the card to the other PC?

I can't really do without the Live in this PC, and the rest of the system
makes it poorly suited for my recording project, so I'll look into getting
a second Live (or better, maybe an Audigy or the M-Audio recommended by
Laurence) for the other PC.

I'd like to keep my expenditure under $100Cdn. I see the Live 5.1 on eBay
for around $25, Audigy LS for about $50, and the M-Audio is around $80.
Recommendations?

--
Greg Schmidt
Trawna Publications http://www.trawna.com/

On Mon, 16 Aug 2004 23:33:55 -0400, Greg Schmidt
wrote:

I'd like to keep my expenditure under $100Cdn. I see the Live 5.1 on eBay
for around $25, Audigy LS for about $50, and the M-Audio is around $80.
Recommendations?

If you need the "features", and don't intend to do multi-track
recording at 44.1KHz, consider the Audigy. For basic functions and
quality, get the Audiophile.

"Laurence Payne" wrote in message
...

I'd like to keep my expenditure under $100Cdn. I see the Live 5.1 on
eBay
for around $25, Audigy LS for about $50, and the M-Audio is around $80.
Recommendations?

If you need the "features", and don't intend to do multi-track
recording at 44.1KHz, consider the Audigy. For basic functions and
quality, get the Audiophile.

From my experience the Audigy LS is not much better than the SB Live. I'd
get the Audiophile or a Turtle Beach Santa Cruz if you want something better
than the SB Live, but less expensive than the M-Audio.

TonyP.