Soundcard/preamp noise levels

[George W.]

I'm using Cool Edit Pro 2.1. My hardware is an M-Audio Audiophile 2496
soundcard connected to a FMR Audio RNP preamp. The cables are regular
RCA type with 1/4" adapters on the preamp end. I record acoustic
guitar and vocals. I was previously using a Behringer 802 mixer to the
card. The improvements in recording with the RNP are pretty obvious.

I'm confused about level readings in CEP and was wondering if someone
could explain them. With nothing connected to the soundcard I get
levels of about -84 db, much better than a previous card. I assume the
noise is due to electronics within the PC. With the RNP turned on and
the gain on zero the levels stay about the same. With the gain turned
up all the way to 60 db (an extremely unreasonable setting, I know)
the CEP meters go up to -48 db. At a useful, real-world gain setting
of 35db the noise level drops to about -75db.

The actual db levels with the Behringer were close to the same but the
recordings done with the RNP are much cleaner. I suppose what I'm
asking is whether these numbers seem to be in the ballpark for the
equipment I'm using or if there's something I can do to improve it. Or
if it matters at all.........

Thanks.

[Arny Krueger]

"George W." > wrote in message


> I'm using Cool Edit Pro 2.1. My hardware is an M-Audio Audiophile 2496
> soundcard connected to a FMR Audio RNP preamp. The cables are regular
> RCA type with 1/4" adapters on the preamp end. I record acoustic
> guitar and vocals. I was previously using a Behringer 802 mixer to the
> card. The improvements in recording with the RNP are pretty obvious.

> I'm confused about level readings in CEP and was wondering if someone
> could explain them. With nothing connected to the soundcard I get
> levels of about -84 db, much better than a previous card. I assume the
> noise is due to electronics within the PC. With the RNP turned on and
> the gain on zero the levels stay about the same. With the gain turned
> up all the way to 60 db (an extremely unreasonable setting, I know)
> the CEP meters go up to -48 db. At a useful, real-world gain setting
> of 35db the noise level drops to about -75db.

> The actual db levels with the Behringer were close to the same but the
> recordings done with the RNP are much cleaner. I suppose what I'm
> asking is whether these numbers seem to be in the ballpark for the
> equipment I'm using or if there's something I can do to improve it. Or
> if it matters at all.........

I presume that you are using readings taken from the record level meter in
CEP. That's one way to do audio measurements with CEP, but take a look at
what happens when you highlight part of your recording and click on Analyze,
Statistics. For one thing, the statistics are for a specific part of your
recording. For another thing you can do this any time over the life of the
recording. Unlike what you just did, it's highly repeatable. Finally, you've
got a far larger variety of potentially meaningful numbers. For a real
thrill, try Analyze, Frequency Analysis.

That all said, noise measurements still aren't worth squat without some kind
of fixed reference level that's meaningful in the real world. In the case of
mic preamp measurements, the fixed reference level is often a
highly-attenuated signal from a signal generator of some kind, set up in a
certain way. Think of it as a simulation of a reference microphone picking
up a reference sound.

Both the RNP and the little Behr have one or more gain controls that among
other things move the noise floor around pretty much at will. The usual way
preamp noise floors are measured involves setting gains for a consistent
level with a signal generator, and then terminating the preamp's input with
a certain value resistor, and taking a measurement over a defined
measurement bandwidth.

Noise measurements also aren't worth squat unless referenced to some kind of
relevant measurement bandwidth. This is because the narrower the band, the
smaller the noise voltage A common measurement band is 3 dB down at 20 and
20 KHz, and flat in-between with first order Butterworth filtering. Another
very popular one is called A weighting, which has a more complex frequency
response curve that is something like that of the ear at low levels. You can
do this all in CEP, but it takes a little work.

You can do this all with CEP, your sound card and a few resistors. Ask and
you shall be answered.

Bottom line is that your current set of measurements seem to fail to follow
the general rules I just pointed out. So, they are basically mystery meat
and AFAIK they always will be.

If you're really interested in noise measurements you are going to have to
do a little homework and also do your actual measurements over again with
some fairly cheap bells and whistles attached.

[George W.]

On Sat, 22 Nov 2003 13:11:08 -0500, Arny Krueger wrote:

>"George W." > wrote in message

>
>> I'm using Cool Edit Pro 2.1. My hardware is an M-Audio Audiophile 2496
>> soundcard connected to a FMR Audio RNP preamp. The cables are regular
>> RCA type with 1/4" adapters on the preamp end. I record acoustic
>> guitar and vocals. I was previously using a Behringer 802 mixer to the
>> card. The improvements in recording with the RNP are pretty obvious.
>
>> I'm confused about level readings in CEP and was wondering if someone
>> could explain them. With nothing connected to the soundcard I get
>> levels of about -84 db, much better than a previous card. I assume the
>> noise is due to electronics within the PC. With the RNP turned on and
>> the gain on zero the levels stay about the same. With the gain turned
>> up all the way to 60 db (an extremely unreasonable setting, I know)
>> the CEP meters go up to -48 db. At a useful, real-world gain setting
>> of 35db the noise level drops to about -75db.
>
>> The actual db levels with the Behringer were close to the same but the
>> recordings done with the RNP are much cleaner. I suppose what I'm
>> asking is whether these numbers seem to be in the ballpark for the
>> equipment I'm using or if there's something I can do to improve it. Or
>> if it matters at all.........
>
>I presume that you are using readings taken from the record level meter in
>CEP. That's one way to do audio measurements with CEP, but take a look at
>what happens when you highlight part of your recording and click on Analyze,
>Statistics. For one thing, the statistics are for a specific part of your
>recording. For another thing you can do this any time over the life of the
>recording. Unlike what you just did, it's highly repeatable. Finally, you've
>got a far larger variety of potentially meaningful numbers. For a real
>thrill, try Analyze, Frequency Analysis.
>
>That all said, noise measurements still aren't worth squat without some kind
>of fixed reference level that's meaningful in the real world. In the case of
>mic preamp measurements, the fixed reference level is often a
>highly-attenuated signal from a signal generator of some kind, set up in a
>certain way. Think of it as a simulation of a reference microphone picking
>up a reference sound.
>
>Both the RNP and the little Behr have one or more gain controls that among
>other things move the noise floor around pretty much at will. The usual way
>preamp noise floors are measured involves setting gains for a consistent
>level with a signal generator, and then terminating the preamp's input with
>a certain value resistor, and taking a measurement over a defined
>measurement bandwidth.
>
>Noise measurements also aren't worth squat unless referenced to some kind of
>relevant measurement bandwidth. This is because the narrower the band, the
>smaller the noise voltage A common measurement band is 3 dB down at 20 and
>20 KHz, and flat in-between with first order Butterworth filtering. Another
>very popular one is called A weighting, which has a more complex frequency
>response curve that is something like that of the ear at low levels. You can
>do this all in CEP, but it takes a little work.
>
>You can do this all with CEP, your sound card and a few resistors. Ask and
>you shall be answered.
>
>Bottom line is that your current set of measurements seem to fail to follow
>the general rules I just pointed out. So, they are basically mystery meat
>and AFAIK they always will be.
>
>If you're really interested in noise measurements you are going to have to
>do a little homework and also do your actual measurements over again with
>some fairly cheap bells and whistles attached.

Thanks for the explanation. I'm not sure if any of these numbers mean
much in real-world terms. While the noise levels between the Behringer
and the RNP seem pretty close the resulting recordings with the RNP
are quite an improvement.....The 802 recordings sounded fine to me in
the past but now sound "boxy" by comparison when recording acoustic
guitar with an MXL 603. I've still got a couple of weeks to evaluate
the RNP and I'd just like to be sure it's working as it should be.

G.

[Mike Rivers]

In article > writes:

> I'm confused about level readings in CEP and was wondering if someone
> could explain them. With nothing connected to the soundcard I get
> levels of about -84 db, much better than a previous card. I assume the
> noise is due to electronics within the PC.

Actually probably noise due to the A/D converters and analog circuitry
at the input stage, but who's counting?

> With the RNP turned on and
> the gain on zero the levels stay about the same. With the gain turned
> up all the way to 60 db (an extremely unreasonable setting, I know)
> the CEP meters go up to -48 db. At a useful, real-world gain setting
> of 35db the noise level drops to about -75db.

Is the input of the preamp connected to anything? It's hard to make
noise measurements that are truly meaningful and accurate when you're
doing it this casually. I usually use an XLR connector with a 150 ohm
resistor between pins 2 and 3 as a "dummy" input. Now this isn't
really good, because the noise of the resistor is greater than the
noise of a decent transformer in a microphone, and it doesn't really
simulate a "150 ohm microphone", but I figure that it's closer to the
real world than a short circuited input. I'm about to try to extract
some thoughts on a measurement procedure that can be performed by
curious people like you, that will give some meaningful results, or at
least results comparable with those performed by another curious
person.

> The actual db levels with the Behringer were close to the same but the
> recordings done with the RNP are much cleaner.

It could be the way that you're handling the gain structure. The
Behringer, if it's built like a typical mixer, has an input gain trim
control as well as an output level control. You may not have optimized
the gains for lowest noise. But there's more to sounding clean than
just noise, and I would expect that an RNP would sound better than a
Behringer mixer. Isn't that why you bought it?

I think you're in the right ballpark. Real world at-home measurements
are rarely as good as the marketing specifications. Sometimes it's
because of measurement conditions, more often it's because you think
you're measuring the same thing that you're seeing numbers for on a
spec sheet, but you're actually measuring something else.




--
I'm really Mike Rivers - )
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo

[Terry King]

>
> That all said, noise measurements still aren't worth squat without some kind
> of fixed reference level that's meaningful in the real world. In the case of
> mic preamp measurements, the fixed reference level is often a
> highly-attenuated signal from a signal generator of some kind, set up in a
> certain way. Think of it as a simulation of a reference microphone picking
> up a reference sound.
>
Arne, Could you suggest a specific resistive attenuator to go from an
unbalanced audio generator (like my HP) to a microphone preamp intended
for 'low impedance balanced' microphones??

Thanks!
--
Regards, Terry King ...In The Woods In Vermont

"The one who dies with the most parts LOSES! What do you need??"

[George W.]

On 22 Nov 2003 16:13:27 -0500, Mike Rivers wrote:

>
>In article > writes:
>
>> I'm confused about level readings in CEP and was wondering if someone
>> could explain them. With nothing connected to the soundcard I get
>> levels of about -84 db, much better than a previous card. I assume the
>> noise is due to electronics within the PC.
>
>Actually probably noise due to the A/D converters and analog circuitry
>at the input stage, but who's counting?
>
>> With the RNP turned on and
>> the gain on zero the levels stay about the same. With the gain turned
>> up all the way to 60 db (an extremely unreasonable setting, I know)
>> the CEP meters go up to -48 db. At a useful, real-world gain setting
>> of 35db the noise level drops to about -75db.
>
>Is the input of the preamp connected to anything? It's hard to make
>noise measurements that are truly meaningful and accurate when you're
>doing it this casually. I usually use an XLR connector with a 150 ohm
>resistor between pins 2 and 3 as a "dummy" input. Now this isn't
>really good, because the noise of the resistor is greater than the
>noise of a decent transformer in a microphone, and it doesn't really
>simulate a "150 ohm microphone", but I figure that it's closer to the
>real world than a short circuited input. I'm about to try to extract
>some thoughts on a measurement procedure that can be performed by
>curious people like you, that will give some meaningful results, or at
>least results comparable with those performed by another curious
>person.
>
>> The actual db levels with the Behringer were close to the same but the
>> recordings done with the RNP are much cleaner.
>
>It could be the way that you're handling the gain structure. The
>Behringer, if it's built like a typical mixer, has an input gain trim
>control as well as an output level control. You may not have optimized
>the gains for lowest noise. But there's more to sounding clean than
>just noise, and I would expect that an RNP would sound better than a
>Behringer mixer. Isn't that why you bought it?
>
>I think you're in the right ballpark. Real world at-home measurements
>are rarely as good as the marketing specifications. Sometimes it's
>because of measurement conditions, more often it's because you think
>you're measuring the same thing that you're seeing numbers for on a
>spec sheet, but you're actually measuring something else.

Thanks Mike. I do tend to agonize over these things, wondering if I
did the right thing. I know $500 is cheap as these things go but in my
world it's a pretty good bite. I'm so used to going through defective
items it's always odd to buy something that works out of the box. In
this case I don't really have a point of reference. In any case it's a
marked improvement over what I was using before and I guess that
matters more than numbers and measurement.

Someone mentioned in another post that you really have to learn to
hear what's good and bad about a piece of equipment, especially one
where improvements are more subtle than dramatic. I've recorded some
short acoustic guitar tracks with the Behringer and the RNP, with a
variety of mics and guitars. The RNP tracks are all easily
distinguishable from the Behringer. All better to some degree
(depending on the mic) though I find it hard to describe the
difference. Even though I'm probably too old for this stuff I'm still
learning, and I appreciate the input from everyone.

G.

[Arny Krueger]

"George W." > wrote in message


> Someone mentioned in another post that you really have to learn to
> hear what's good and bad about a piece of equipment, especially one
> where improvements are more subtle than dramatic. I've recorded some
> short acoustic guitar tracks with the Behringer and the RNP, with a
> variety of mics and guitars. The RNP tracks are all easily
> distinguishable from the Behringer. All better to some degree
> (depending on the mic) though I find it hard to describe the
> difference. Even though I'm probably too old for this stuff I'm still
> learning, and I appreciate the input from everyone.

I really don't know too much about the technical differences between a RNP
and the little Behr 802. Mic preamp spec sheets generally suck. I can only
recall seeing one mic preamp spec sheet that I felt represented the product
reasonably well. The RNP tech information I've seen is a lot more wordy, but
seems to have no more hard details than the vacuous spec sheet for the
little Behr.

My past experience suggests that were the comparison to be done rigorously,
differences that seem more obvious under casual testing conditions would
become less obvious if the test were repeated with more vigor.

But, I would be a lot less surprised if mic preamps turned out to sound
significantly different than power amps. There seem to be more variables.

[Mike Rivers]

In article > writes:

> I know $500 is cheap as these things go but in my
> world it's a pretty good bite. I'm so used to going through defective
> items it's always odd to buy something that works out of the box.

Gosh, you must buy a lot of old junk on eBay. <G> I'm impressed with
how much good sounding stuff for not a lot of money is out there
today. $500 is a fair price for a good 2-channel preamp. You can spend
7 times that amount, but you can't get much for half that amount.

> I don't really have a point of reference. In any case it's a
> marked improvement over what I was using before and I guess that
> matters more than numbers and measurement.

Yup. People don't listen to numbers. And more important, the listeners
won't know what gear you used to make your recordings. If they don't
like the music, they won't care, and if the like the music, they can
forgive a couple of dB of hiss in the background. Chances are they
won't be able to hear it over the traffic anyway.





--
I'm really Mike Rivers - )
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me here: double-m-eleven-double-zero at yahoo

[P Stamler]

Listen to what Mike R. said: if you were testing the preamps with nothing
plugged into them, your numbers basically don't mean much. I also use a dummy
plug with a 150-ohm resistor; it's not really that bad a representation.

All that said, the RNP isn't the quietest preamp on the block -- the designer
said that was one of his deliberate choices, sacrificing some noise performance
for better sound on higher-level signals. But it's quiet enough for most
applications involving condenser mikes.

Peace,
Paul