Equalisation for PC mic input/line input

[David Peters (UK)]

QUESTION:
Do the PC mic and line inputs use the same equalisation (on the
frequency spectrum)?


BACKGROUND:
I have some noisy voice tapes from an old analogue dictation machine.
I'm reading the recordings from the "ear" socket of my dictation
machine into my PC and then coverting the recordings to MP3.

Later I will get some software to clean up the noise on the MP3s.

I didn't expect it but my PC allows me to set a decent recording
level whether I record through the mic input or the line input.

Until I clean up the sound, the noise from the original recording
makes it hard to tell if I'm getting a better result from the mic
input or the line input.

I was wondering if there was a different equalisation used by the PC
for the mic and line inputs. If so then I would make sure I used


ANOTHER QUESTION:
What is the input level at which the mic and line inputs are rated?
I had thought mic inputs were about 3 or 3 mV and line inputs were
200 mV.

-------

NOTE:
My PC motherboard chipset is VIA KT266A + VT8235.
A PC reporting utility says it detects a VT8233/A AC97 Enhanced Audio
Controller.

 

[Mike Walsh]

QUESTION:
Do the PC mic and line inputs use the same equalisation (on the
frequency spectrum)?

There is not equalization with line input, i.e. it uses flat frequency response. I am not sure about microphone input; the biggest difference seems to be they operate at a lower signal level.

BACKGROUND:
I have some noisy voice tapes from an old analogue dictation machine.
I'm reading the recordings from the "ear" socket of my dictation
machine into my PC and then coverting the recordings to MP3.

Later I will get some software to clean up the noise on the MP3s.

I didn't expect it but my PC allows me to set a decent recording
level whether I record through the mic input or the line input.

Until I clean up the sound, the noise from the original recording
makes it hard to tell if I'm getting a better result from the mic
input or the line input.

You should use the line input with the headphone output, as both operate at relatively high signal level with flat frequency response.

I was wondering if there was a different equalisation used by the PC
for the mic and line inputs. If so then I would make sure I used

ANOTHER QUESTION:
What is the input level at which the mic and line inputs are rated?
I had thought mic inputs were about 3 or 3 mV and line inputs were
200 mV.

Back in the days of analog recording 0 db was 1 volt. Since this was analog the 0 db level could be and was often exceed. With the advent of digital CDs the 0 db level became the maximum level, which can not be exceeded because of the digital format, and is supposed to be 2 volts. Since these are maximum levels the average will much lower.
Microphone levels are lower and vary widely.

 

[Serge Auckland]

There is not equalization with line input, i.e. it uses flat frequency
response. I am not sure about microphone input; the biggest difference
seems to be they operate at a lower signal level.

Microphone inputs are also flat, i.e. with no equalisation. As you say, they
operate at a much lower level, typically a few millivolts.

You should use the line input with the headphone output, as both operate
at relatively high signal level with flat frequency response.
Correct.

Back in the days of analog recording 0 db was 1 volt. Since this was
analog the 0 db level could be and was often exceed. With the advent of
digital CDs the 0 db level became the maximum level, which can not be
exceeded because of the digital format, and is supposed to be 2 volts.
Since these are maximum levels the average will much lower.
Microphone levels are lower and vary widely.

Not quite. Firstly, a dB is a relative level, not an absolute, so without
stating the reference, a figure of "xdB" is meaningless. Originally, 0dB was
referenced to a power of 1mW into a load of 600 ohms, and was referred to as
0dBm. Later, the same voltage level, but unloaded, that is, without
reference to a 600 ohm load became 0dBu (that is, unloaded) Note that the
voltage level is the same in both cases (0.775v, or 1mW into 600 ohm) There
was a strange semi-standard evolved of referring to 1V rather than 0.775v
and that was 0dBv.

Digital outputs are referred to maximum digital output (when all the bits
are 1) and that is called 0dBFS (0dB Full Scale). It has NO analogue
equivalent, as analogue can keep getting bigger without limit, digital can't
get any bigger than when all the bits are 1. In Digital-Analogue conversion,
a number of different conversion levels have become more-or-less standard.
The EBU (European Broadcasting Union) have defined 0dBFS digital to mean
+18dBu analogue after conversion. The USA prefers that 0dBFS = +24dBu
because that provides 20dB headroom above 0VU. A few dissidents prefer
+25dBu as that's 1dB better than +24...........

CD players have evolved a standard output of 2v analogue for 0dBFS, but as
far as I'm aware, there is no official standard for this.

S.

 

[Kevin Seal]

(snip)

The EBU (European Broadcasting Union) have defined 0dBFS digital to mean
+18dBu analogue after conversion.

Interesting.
Can you let me have a reference to the technical paper for that.
Cheers,

 

[Serge Auckland]

Interesting.
Can you let me have a reference to the technical paper for that.
Cheers,

EBU R68-2000. I'm emailing you a copy directly.

S.

 

[Kevin Seal]

EBU R68-2000. I'm emailing you a copy directly.

Received, thanks.

With 0dBFS as =18dBU, that would mean OVU (+4dBU) would be -14dBFS. Most
people I know line-up their Pro Tools rigs for -18dBFS for 0VU hence
OdBFS is going to be +22dBU.
Isn't it a lovely world!

 

[Serge Auckland]

Received, thanks.

With 0dBFS as =18dBU, that would mean OVU (+4dBU) would be -14dBFS. Most
people I know line-up their Pro Tools rigs for -18dBFS for 0VU hence OdBFS
is going to be +22dBU.
Isn't it a lovely world!

Standard are great, that's why we have so many of them!

S.

 

[kony]

QUESTION:
Do the PC mic and line inputs use the same equalisation (on the
frequency spectrum)?


BACKGROUND:
I have some noisy voice tapes from an old analogue dictation machine.
I'm reading the recordings from the "ear" socket of my dictation
machine into my PC and then coverting the recordings to MP3.

Later I will get some software to clean up the noise on the MP3s.

I didn't expect it but my PC allows me to set a decent recording
level whether I record through the mic input or the line input.

Until I clean up the sound, the noise from the original recording
makes it hard to tell if I'm getting a better result from the mic
input or the line input.

I was wondering if there was a different equalisation used by the PC
for the mic and line inputs. If so then I would make sure I used


ANOTHER QUESTION:
What is the input level at which the mic and line inputs are rated?
I had thought mic inputs were about 3 or 3 mV and line inputs were
200 mV.

-------

NOTE:
My PC motherboard chipset is VIA KT266A + VT8235.
A PC reporting utility says it detects a VT8233/A AC97 Enhanced Audio
Controller.

The short answer is that for best results you should:

- use a quality tape deck with line-out, not the earphone
jack of a dictation system

- use the line-in on a fair quality sound card, not
integrated motherboard audio with a really cheap codec.

 

[Rich Wilson]

Standard are great, that's why we have so many of them!

Decibels, to me, seem to be overused, particularly with digital audio. And
particularly because silence is negative infinity decibels, which isn't a
lot of good if you're writing a computer program that can only cope with
real numbers. What's wrong with plain old 0% to 100%?!
(Rhetorical question, don't feel obliged to answer...)

 

[David Peters (UK)]

Not quite. Firstly, a dB is a relative level, not an absolute,
so without stating the reference, a figure of "xdB" is
meaningless. Originally, 0dB was referenced to a power of 1mW
into a load of 600 ohms, and was referred to as 0dBm. Later, the
same voltage level, but unloaded, that is, without reference to
a 600 ohm load became 0dBu (that is, unloaded) Note that the
voltage level is the same in both cases (0.775v, or 1mW into 600
ohm) There was a strange semi-standard evolved of referring to
1V rather than 0.775v and that was 0dBv.

Digital outputs are referred to maximum digital output (when all
the bits are 1) and that is called 0dBFS (0dB Full Scale). It
has NO analogue equivalent, as analogue can keep getting bigger
without limit, digital can't get any bigger than when all the
bits are 1. In Digital-Analogue conversion, a number of
different conversion levels have become more-or-less standard.
The EBU (European Broadcasting Union) have defined 0dBFS digital
to mean +18dBu analogue after conversion. The USA prefers that
0dBFS = +24dBu because that provides 20dB headroom above 0VU. A
few dissidents prefer +25dBu as that's 1dB better than
+24...........

CD players have evolved a standard output of 2v analogue for
0dBFS, but as far as I'm aware, there is no official standard
for this.

I can't say I understand all of what you write but the parts I do
understand are very useful to me. Thank you for posting.

Are there any web sites or documents which explain this sort of thing
for a beginner: rigorously but not going too fast.

 

[Serge Auckland]

Lots snipped>

I can't say I understand all of what you write but the parts I do
understand are very useful to me. Thank you for posting.

Are there any web sites or documents which explain this sort of thing
for a beginner: rigorously but not going too fast.

I don't know of any specific websites that explain all the ins and outs of
analogue and digital audio. I've learned all this during my professional
life in audio. There used to be a great magazine called Studio Sound, which
had technical articles explaining the basics in rigorous but understandable
form. Sadly SS has been extinct for several years, but you may find copies
in larger public libraries.

You may also want to look at Jim Lesurf's web sites -
Electronics http://www.st-and.ac.uk/~www_pa/Scots_Guide/intro/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html

He's got a lot of useful information, and what's more, it's correct!

Canford Audio have useful information in their catalogue, you may want to
contact them and see if they'll put you on the mailing list for the
catalogue. www.canford.co.uk

S.

 

[Kevin Seal]

Decibels, to me, seem to be overused, particularly with digital audio. And
particularly because silence is negative infinity decibels, which isn't a
lot of good if you're writing a computer program that can only cope with
real numbers. What's wrong with plain old 0% to 100%?!
(Rhetorical question, don't feel obliged to answer...)

Are you trying to put us out of a job?

 

[Stewart Pinkerton]

Are you trying to put us out of a job?

Not much danger of that, if he's such a bad programmer!

BTW, what's wrong with 0-100% is that our hearing is logarithmic, so
deciBels give a much better idea of how things sound. A 10dB increase
in SPL sounds twice as loud, but takes ten times the power.

 

[Dave Plowman (News)]

Digital outputs are referred to maximum digital output (when all the
bits are 1) and that is called 0dBFS (0dB Full Scale). It has NO
analogue equivalent, as analogue can keep getting bigger without limit,
digital can't get any bigger than when all the bits are 1. In
Digital-Analogue conversion, a number of different conversion levels
have become more-or-less standard. The EBU (European Broadcasting
Union) have defined 0dBFS digital to mean +18dBu analogue after
conversion. The USA prefers that 0dBFS = +24dBu because that provides
20dB headroom above 0VU. A few dissidents prefer +25dBu as that's 1dB
better than +24...........

It's quite interesting to look at levels off Freeview. I lined up the
workshop receiver to read PPM 4 on a rare occasion when there was a test
card and line up tone available. And as expected TV progs peak to no more
than PPM 6. But some of the radio ones wrap the PPMs round the end stops.
;-)

My best Freeview receiver out of several is a Sony VTX-D800U and when I
changed it from the previous freebie Sagem which kept crashing I
immediately noticed the audio level was low. Switch the set from the same
channel on analogue to Freeview via a SCART and the difference was too
much. Correspondence with Sony showed that they thought the TV
broadcasters would peak to 0dBFS on FreeView instead of using the normal
EBU line up of peak being -10 dBFS.

 

[don]

dbm is as stated a reference of two like power values to a 1mW reference
however the impedence does not need to be 600 ohms, it can be any value of
ohms as long as both power values are based on the same impedence

dbu is not unloaded but db(micro) it is as above but with a reference of
1microWatt it is not actually a u but the greek character mu

dbv would be a ratio based on two voltage levels and a reference of 1 volt

power db caclulations are 10 log Pout/Pin
Voltage db calculations are 20 log Vout/Vin

dbFS is "decibels full scale". It is an abbreviation for decibel amplitude
levels in digital systems which have a maximum available level (like PCM
encoding). 0 dBFS is assigned to the maximum possible level. There is still
the potential for ambiguity, since some use the RMS value of a full-scale
square wave for 0 dBFS, and some use a sine wave.

this is treated the same as voltag calculations because it is based on the
signal to noise ratio.

 

[Don Pearce]

dbm is as stated a reference of two like power values to a 1mW reference
however the impedence does not need to be 600 ohms, it can be any value of
ohms as long as both power values are based on the same impedence

No, the impedance does not need to be the same, and there are not two
power values, but one - specified as dB with respect to one milliwatt.
Impedance does not appear anywhere in this figure.

dbu is not unloaded but db(micro) it is as above but with a reference of
1microWatt it is not actually a u but the greek character mu

dBu is indeed dB (unloaded). It is a relic of 600 ohm line audio
systems and is the voltage that would have produced 0dBm in 600 ohms,
but since we now run into high impedances instead, must be specified
otherwise - hence dBu.

dbv would be a ratio based on two voltage levels and a reference of 1 volt

dBV, actually, not dBv.

power db caclulations are 10 log Pout/Pin
Voltage db calculations are 20 log Vout/Vin

dbFS is "decibels full scale". It is an abbreviation for decibel amplitude
levels in digital systems which have a maximum available level (like PCM
encoding). 0 dBFS is assigned to the maximum possible level. There is still
the potential for ambiguity, since some use the RMS value of a full-scale
square wave for 0 dBFS, and some use a sine wave.

No, no ambiguity, dB below full scale does not depend on wave shape,
merely how many digital levels remain unused.

this is treated the same as voltag calculations because it is based on the
signal to noise ratio.

It has nothing to do with signal to noise ratio - it is all happening
at the other end of the scale.

d

Pearce Consulting
http://www.pearce.uk.com

 

[Serge Auckland]

Don,

Do you have a reference for this statement? In 34 years in Pro-Audio I have
never heard it expressed in this way, always referred to 1mW into 600 ohms.

dBu is indeed dB (unloaded). It is a relic of 600 ohm line audio
systems and is the voltage that would have produced 0dBm in 600 ohms,
but since we now run into high impedances instead, must be specified
otherwise - hence dBu.

dBV, actually, not dBv.


It has nothing to do with signal to noise ratio - it is all happening
at the other end of the scale.

I have no diea what the above is referring to: dBFS has nothing to do with
signal-to-noise, it is just how many dB below full-scale.

S.

 

[Don Pearce]

Don,

Do you have a reference for this statement? In 34 years in Pro-Audio I have
never heard it expressed in this way, always referred to 1mW into 600 ohms.

No, it is just one milliwatt - no ohms needed. You have only come
across it in relation to 600 ohms because you have been worked in
audio, and that is all you have been exposed to. If you ever worked in
RF, you would have found exactly the same power, referred to in
exactly the same way in 50, 62.5 and 74 ohms. The power is the same in
all of these - and 600 ohms too.

I have no diea what the above is referring to: dBFS has nothing to do with
signal-to-noise, it is just how many dB below full-scale.

S.

Exactly.

d

Pearce Consulting
http://www.pearce.uk.com

 

[Glenn Booth]

Hi,

No, it is just one milliwatt - no ohms needed. You have only come
across it in relation to 600 ohms because you have been worked in
audio, and that is all you have been exposed to. If you ever worked in
RF, you would have found exactly the same power, referred to in
exactly the same way in 50, 62.5 and 74 ohms. The power is the same in
all of these - and 600 ohms too.

Agreed. We had the same conversation on uk.r.a back in 2003. I've
quoted a bit of it here as Serge might be interested:

I said:

I don't disagree that the reference must always be given, but for
measurements of power, such as those that reference dBm (dB referenced
to 1mW) surely the impedance is totally redundant? (Unless what one is
really trying to describe is voltage, but using a power ratio to do
so). The 50R says nothing that I can see about the power, it only
allows one to relate the voltage that will be dropped across that
particular impedance/resistance with that dB worth of power being
dissipated.

To which a certain Mr. Lesurf said this:

However, bear in mind two points:

1) That in most cases (in RF at least) the quoted systems will be based
upon assuming the system is impedance matched and then give the power that
will be delivered to the source. Hence quoting the impedance tells the user
that this is the required matched impedance for optimum power transfer.

2) That in most cases the receiver will tend to be designed to work over a
given voltage range due to finite voltage rails, etc. Hence the impedance
is useful for establishing the voltage levels that must be expected.

It is therefore useful to confirm the assumed impedances. In RF/microwave
we have the annoyance that 50 Ohm is common for system and lab work, but
other impedances like 75 Ohm, etc, crop up for specific purposes/areas.

In principle, though: yes, once you've quoted the signal power in dBm
you've established the power available. You could then use a transformer to
alter the impedance (and hence signal voltage) if so desired.

At the time, I hadn't considered case (1) that Jim pointed out, as I know

less than nothing about RF.

Who says Usenet doesn't go around in circles?

Regards,

Glenn.

 

[Serge Auckland]

Hi,



Agreed. We had the same conversation on uk.r.a back in 2003. I've
quoted a bit of it here as Serge might be interested:

I said:

To which a certain Mr. Lesurf said this:




At the time, I hadn't considered case (1) that Jim pointed out, as I know

less than nothing about RF.

Who says Usenet doesn't go around in circles?

Regards,

Glenn.

Thanks to Don and Glenn. I've learnt something. Looking through my old
college texts books, I can see you're right. What comes of a narrow
upbringing.

S.