Sign in with
Sign up | Sign in
Your question

M-Audio 2496 problem (possibly)

Tags:
Last response: in Home Audio
Share
Anonymous
December 30, 2004 10:23:06 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

I've been using Audition + the 2496 to do some vinyl-to-digital
transfers, and noticed that the left channel level usually seemed a
bit lower than the right, no matter what LP I was transferring.
I attributed this to some problem with my turntable setup, and
set about checking and adjusting cartridge overhang,
angle, tracking force etc -- which only reminded me why I was so
happy to switch to CD years ago. Anyway, having done that the problem
was still there, so I set about chekcing other components of the chain.
I was very surprised to find that, with all cables disconnected
from the soundcard, if I recorded as normal (i.e., Audition set to
record from the 2496's analog L/R inputs), the ambient 'noise'
tus captured *also* shows the left channel level to be ~1 dB lower
on average than right. It's also quite visible by watchign the
static and dynamic peak indicators while recording -- the peak always
tends to end up about 1 dB higher for the right than the left.
This is not the case if I select the 2496 S/PDIF
input as source, btw -- there the noise floor is not only lower,
as expected, but the peak and everage values for L and R do not differ
very much at all, nor do they favor one channel over another.

How and where else might I check for the source of this difference,
before assuming that there's a problem with the card? It's installed
in a bog-standard Dell PC running WinXP Home. My other question would
be, what am I actually recording when there's nothing connected to
the inputs? Is it the 'sound' of the card, the sound of the PC, or
some combination?
Anonymous
December 30, 2004 10:23:07 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

"Steven Sullivan" wrote ...
> I've been using Audition + the 2496 to do some vinyl-to-digital
> transfers, and noticed that the left channel level usually seemed a
> bit lower than the right, no matter what LP I was transferring.
> I attributed this to some problem with my turntable setup, and
> set about checking and adjusting cartridge overhang,
> angle, tracking force etc -- which only reminded me why I was so
> happy to switch to CD years ago. Anyway, having done that the problem
> was still there, so I set about chekcing other components of the
> chain.
> I was very surprised to find that, with all cables disconnected
> from the soundcard, if I recorded as normal (i.e., Audition set to
> record from the 2496's analog L/R inputs), the ambient 'noise'
> tus captured *also* shows the left channel level to be ~1 dB lower
> on average than right. It's also quite visible by watchign the
> static and dynamic peak indicators while recording -- the peak always
> tends to end up about 1 dB higher for the right than the left.

Have you tried feeding a MONO signal into both inputs (using a
Y-adapter, for example)? It seems possible that the analog gain
in your 2496 is out of balance. I'd guess that 1dB unbalance exeeds
the manufacturer's spec. Can you return it? Or you could just
compensate somewhere else, either hardware (like trimming the
RIAA preamp?), or software (do you do any processing, cleanup,
normalization, etc.?)

> This is not the case if I select the 2496 S/PDIF
> input as source, btw -- there the noise floor is not only lower,
> as expected, but the peak and everage values for L and R do not differ
> very much at all, nor do they favor one channel over another.

I am surprised that you see ANYTHING (noise or otherwise)
with thet SPDIF input. If there are no bits going in, there is nothing
to interperet as audio (even noise).

> How and where else might I check for the source of this difference,
> before assuming that there's a problem with the card?

I would feed a mono signal into both inputs with a Y-adapter to
prove that the two channels are behaving differently with exactly
the same input signal.

> It's installed in a bog-standard Dell PC running WinXP Home.
> My other question would be, what am I actually recording when
> there's nothing connected to the inputs? Is it the 'sound' of the
> card,
> the sound of the PC, or some combination?

It is likely the sound of the analog parts of the card (before and
including the A/D converter). Might also be picking up noise
from inside the computer case (depending on many things).

OTOH, with digital input (SPDIF), I'd expect to see NOTHING
(pure, digital-zero "silence") unless you have some SPDIF source
connnected.
Anonymous
December 30, 2004 10:26:43 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

I should add that I also tried removing and re-inserting it
the soundcard from its expansion slot, on the theory that there
was some crud on the contacts -- but that didn't change
anything. I haven't tried using a differend slot.
Related resources
Anonymous
December 31, 2004 8:26:42 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech and rec.audio.pro, on Thu, 30 Dec 2004 11:49:07
-0800, "Richard Crowley" <rcrowley7@xprt.net> wrote:

>"Steven Sullivan" wrote ...
>> I've been using Audition + the 2496 to do some vinyl-to-digital
>> transfers, and noticed that the left channel level usually seemed a
>> bit lower than the right, no matter what LP I was transferring.
>> I attributed this to some problem with my turntable setup, and
>> set about checking and adjusting cartridge overhang,
>> angle, tracking force etc -- which only reminded me why I was so
>> happy to switch to CD years ago. Anyway, having done that the problem
>> was still there, so I set about chekcing other components of the
>> chain.

What components are those? Specifically, what cartridge are you
using, and what are you using for the RIAA phono preamp? Many
cartridges have a spec for output match between channels, and it can
be 0.5 dB or so, depending on model.

>> I was very surprised to find that, with all cables disconnected
>> from the soundcard, if I recorded as normal (i.e., Audition set to
>> record from the 2496's analog L/R inputs), the ambient 'noise'
>> tus captured *also* shows the left channel level to be ~1 dB lower
>> on average than right. It's also quite visible by watchign the
>> static and dynamic peak indicators while recording -- the peak always
>> tends to end up about 1 dB higher for the right than the left.
>
>Have you tried feeding a MONO signal into both inputs (using a
>Y-adapter, for example)? It seems possible that the analog gain
>in your 2496 is out of balance. I'd guess that 1dB unbalance exeeds
>the manufacturer's spec. Can you return it? Or you could just
>compensate somewhere else, either hardware (like trimming the
>RIAA preamp?),

Also, similarly feed a mono signal into the RIAA preamp with its
outputs connected to the soundcard normally, and see what the
imbalance is (might as well try this at several frequencies, to check
if both channels have the same EQ). If the imbalance is the same as it
is going directly into the soundcard, then the soundcard is the
problem. If not, then the RIAA preamp is a problem (but perhaps not
the only problem).


-----
http://mindspring.com/~benbradley
Anonymous
December 31, 2004 3:34:21 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

Thu, 30 Dec 2004 19:23:06 +0000 (UTC), Steven Sullivan
<ssully@panix.com> pisze:

>I was very surprised to find that, with all cables disconnected
>from the soundcard, if I recorded as normal (i.e., Audition set to
>record from the 2496's analog L/R inputs), the ambient 'noise'
>tus captured *also* shows the left channel level to be ~1 dB lower
>on average than right. It's also quite visible by watchign the
>static and dynamic peak indicators while recording -- the peak always
>tends to end up about 1 dB higher for the right than the left.

Did you check settings on M-Audio mixer? You can have line-in out of
balans. I think this mixer works in digital domain, so it can
attenuate ADC noise level.


--
Pozdrowienia

Andrzej Popowski
Anonymous
December 31, 2004 9:16:29 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech Andrzej Popowski <popej@friko.onet.pl> wrote:
> Thu, 30 Dec 2004 19:23:06 +0000 (UTC), Steven Sullivan
> <ssully@panix.com> pisze:

> >I was very surprised to find that, with all cables disconnected
> >from the soundcard, if I recorded as normal (i.e., Audition set to
> >record from the 2496's analog L/R inputs), the ambient 'noise'
> >tus captured *also* shows the left channel level to be ~1 dB lower
> >on average than right. It's also quite visible by watchign the
> >static and dynamic peak indicators while recording -- the peak always
> >tends to end up about 1 dB higher for the right than the left.

> Did you check settings on M-Audio mixer? You can have line-in out of
> balans. I think this mixer works in digital domain, so it can
> attenuate ADC noise level.

Yes, I made sure all the mixer faders were really at 0.0 dB when I first
noticed the problem.





--

-S
If you're a nut and knock on enough doors, eventually someone will open one,
look at you and say, Messiah, we have waited for your arrival.
Anonymous
January 1, 2005 10:19:17 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

Steven Sullivan wrote:
> I've been using Audition + the 2496 to do some vinyl-to-digital
> transfers, and noticed that the left channel level usually seemed a
> bit lower than the right, no matter what LP I was transferring.
> I attributed this to some problem with my turntable setup, and
> set about checking and adjusting cartridge overhang,
> angle, tracking force etc -- which only reminded me why I was so
> happy to switch to CD years ago. Anyway, having done that the problem
> was still there, so I set about chekcing other components of the chain.
> I was very surprised to find that, with all cables disconnected
> from the soundcard, if I recorded as normal (i.e., Audition set to
> record from the 2496's analog L/R inputs), the ambient 'noise'
> tus captured *also* shows the left channel level to be ~1 dB lower
> on average than right. It's also quite visible by watchign the
> static and dynamic peak indicators while recording -- the peak always
> tends to end up about 1 dB higher for the right than the left.
> This is not the case if I select the 2496 S/PDIF
> input as source, btw -- there the noise floor is not only lower,
> as expected, but the peak and everage values for L and R do not differ
> very much at all, nor do they favor one channel over another.
>
> How and where else might I check for the source of this difference,
> before assuming that there's a problem with the card? It's installed
> in a bog-standard Dell PC running WinXP Home. My other question would
> be, what am I actually recording when there's nothing connected to
> the inputs? Is it the 'sound' of the card, the sound of the PC, or
> some combination?
>
>
>
>
>
>
>
See what happens when you record a mono signal, as suggested before.
Also, try swapping the channels, connect the left to the right input and
vice versa. If everything is equal, you should now see the left side 1
db higher than the right. If not, then I would have to conclude the the
M-Audio card is suspect. Not in a major way, though. Allot of my analog
stereo recording tends to be off by 1-1.5 db, but when I do a test
recording in mono my left and right levels are perfectly even.

CD

CD
Anonymous
January 2, 2005 4:59:54 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech Codifus <codifus@optonline.net> wrote:
> Steven Sullivan wrote:
> > I've been using Audition + the 2496 to do some vinyl-to-digital
> > transfers, and noticed that the left channel level usually seemed a
> > bit lower than the right, no matter what LP I was transferring.
> > I attributed this to some problem with my turntable setup, and
> > set about checking and adjusting cartridge overhang,
> > angle, tracking force etc -- which only reminded me why I was so
> > happy to switch to CD years ago. Anyway, having done that the problem
> > was still there, so I set about chekcing other components of the chain.
> > I was very surprised to find that, with all cables disconnected
> > from the soundcard, if I recorded as normal (i.e., Audition set to
> > record from the 2496's analog L/R inputs), the ambient 'noise'
> > tus captured *also* shows the left channel level to be ~1 dB lower
> > on average than right. It's also quite visible by watchign the
> > static and dynamic peak indicators while recording -- the peak always
> > tends to end up about 1 dB higher for the right than the left.
> > This is not the case if I select the 2496 S/PDIF
> > input as source, btw -- there the noise floor is not only lower,
> > as expected, but the peak and everage values for L and R do not differ
> > very much at all, nor do they favor one channel over another.
> >
> > How and where else might I check for the source of this difference,
> > before assuming that there's a problem with the card? It's installed
> > in a bog-standard Dell PC running WinXP Home. My other question would
> > be, what am I actually recording when there's nothing connected to
> > the inputs? Is it the 'sound' of the card, the sound of the PC, or
> > some combination?
> >
> >
> >
> >
> >
> >
> >
> See what happens when you record a mono signal, as suggested before.
> Also, try swapping the channels, connect the left to the right input and
> vice versa. If everything is equal, you should now see the left side 1
> db higher than the right. If not, then I would have to conclude the the
> M-Audio card is suspect. Not in a major way, though. Allot of my analog
> stereo recording tends to be off by 1-1.5 db, but when I do a test
> recording in mono my left and right levels are perfectly even.


> CD


If I didn't make it clear -- I see this imbalance even when there's
*nothing connected to the soundcard inputs*. So swapping channels should
make no difference.



--




-S If you're a nut and knock on enough doors, eventually someone will open
one, look at you and say, Messiah, we have waited for your arrival.
Anonymous
January 2, 2005 1:05:27 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In article <cr7kiq$fkk$1@reader1.panix.com> ssully@panix.com writes:

> > > I've been using Audition + the 2496 to do some vinyl-to-digital
> > > transfers, and noticed that the left channel level usually seemed a
> > > bit lower than the right

> > See what happens when you record a mono signal, as suggested before.

> If I didn't make it clear -- I see this imbalance even when there's
> *nothing connected to the soundcard inputs*. So swapping channels should
> make no difference.

You made it clear - there's 1 dB difference between channels.

There are a number of freeware programs out there which will create a
WAV file of a single frequency. In fact I think that Audition does
this. Make a 1 kHz file about five minutes long, burn it on a CD, and
use your CD player as a tone generator. That will give you a
calibration signal. Connect it to your sound card inputs and read the
meters. If one is 1 dB off from the other, open up the control panel
and adjust the gain so that the channels are balanced. That's all
there is to it.


--
I'm really Mike Rivers (mrivers@d-and-d.com)
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
Anonymous
January 2, 2005 9:29:36 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

2 Jan 2005 10:05:27 -0500, mrivers@d-and-d.com (Mike Rivers) pisze:

>If one is 1 dB off from the other, open up the control panel
>and adjust the gain so that the channels are balanced. That's all
>there is to it.

OK, that is practical solution, but IMHO there is an error in hardware
or in software setup, that we are unable to trace down.

1dB is about 12% difference of voltage gain. Analog input of DAP2496
consist of 3 opamp 5532 and ADC chip AKM4528VF. I would not expect
more than 1% gain error from that kind of setup, unless something is
wrong.

To Steven: try this software to test your card performance:
http://audio.rightmark.org/downloads/rmaa54.exe

All you need is to connect a wire from line-out to line-in and run
RMAA. It will measure frequency response, dynamic range , distortion
and crosstalk of your card. Check please if there are no other
differences between channels.


--
Pozdrowienia

Andrzej Popowski
Anonymous
January 3, 2005 5:26:21 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech Mike Rivers <mrivers@d-and-d.com> wrote:

> In article <cr7kiq$fkk$1@reader1.panix.com> ssully@panix.com writes:

> > > > I've been using Audition + the 2496 to do some vinyl-to-digital
> > > > transfers, and noticed that the left channel level usually seemed a
> > > > bit lower than the right

> > > See what happens when you record a mono signal, as suggested before.

> > If I didn't make it clear -- I see this imbalance even when there's
> > *nothing connected to the soundcard inputs*. So swapping channels should
> > make no difference.

> You made it clear - there's 1 dB difference between channels.

> There are a number of freeware programs out there which will create a
> WAV file of a single frequency. In fact I think that Audition does
> this. Make a 1 kHz file about five minutes long, burn it on a CD, and
> use your CD player as a tone generator. That will give you a
> calibration signal. Connect it to your sound card inputs and read the
> meters. If one is 1 dB off from the other, open up the control panel
> and adjust the gain so that the channels are balanced. That's all
> there is to it.

No need to -- I did the y-adapter mono test instead, using an audio test
CD as source. Used the pink noise L/R and spot frequency tracks as tests.
And what do you know -- the L and R channel difference
disappeared (channel difference was on the order of 0.05 dB at most).
It came back to ~1.0 dB when the cables were disconnected. So it looks like
there is a channel difference when there's nothing connected, and the
channel difference I saw when I was transferring vinyl was either the LPs,
the TT or the RIAA preamp.

As such I'm not going to fuss about it any longer, I'll just compensate
in the digital realm if need be. Thanks to everyone for their input.
Anonymous
January 3, 2005 5:27:17 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech Andrzej Popowski <popej@friko.onet.pl> wrote:
> 2 Jan 2005 10:05:27 -0500, mrivers@d-and-d.com (Mike Rivers) pisze:

> >If one is 1 dB off from the other, open up the control panel
> >and adjust the gain so that the channels are balanced. That's all
> >there is to it.

> OK, that is practical solution, but IMHO there is an error in hardware
> or in software setup, that we are unable to trace down.

> 1dB is about 12% difference of voltage gain. Analog input of DAP2496
> consist of 3 opamp 5532 and ADC chip AKM4528VF. I would not expect
> more than 1% gain error from that kind of setup, unless something is
> wrong.

> To Steven: try this software to test your card performance:
> http://audio.rightmark.org/downloads/rmaa54.exe

HAving jsut said I won't fuss with this matter any more, I *will* however
try this. ;> Thanks.


--

-S
If you're a nut and knock on enough doors, eventually someone will open one,
look at you and say, Messiah, we have waited for your arrival.
Anonymous
January 3, 2005 5:41:53 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech Andrzej Popowski <popej@friko.onet.pl> wrote:
> http://audio.rightmark.org/downloads/rmaa54.exe

> All you need is to connect a wire from line-out to line-in and run
> RMAA. It will measure frequency response, dynamic range , distortion
> and crosstalk of your card. Check please if there are no other
> differences between channels.


results of 32bit/48kHz (my dstandard trasner parameters)loopback test


Frequency response (from 40 Hz to 15 kHz), dB: +0.02, -0.08 Excellent
Noise level, dB (A): -98.6 Excellent
Dynamic range, dB (A): 93.0 Very good
THD, %: 0.0047 Very good
IMD, %: 0.012 Very good
Stereo crosstalk, dB: -101.3 Excellent
IMD at 10 kHz, %: 0.011 Very good


So it looks again like the card is fine by these measures.


--


-S
If you're a nut and knock on enough doors, eventually someone will open one,
look at you and say, Messiah, we have waited for your arrival.
Anonymous
January 3, 2005 5:50:54 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

and here are individual L/R breakdowns.
Are the IMD and THD discrepancies significant?

Noise level Left Right
RMS power, dB: -94.8 -94.7
RMS power (A-weighted), dB: -98.6 -98.4
Peak level, dB FS: -79.2 -77.0
DC offset, %: -0.00 -0.00

Dynamic range
Dynamic range, dB: +91.9 +91.9
Dynamic range (A-weighted), dB: +93.1 +93.0
DC offset, %: -0.00 -0.00

THD + Noise (at -3 dB FS)
THD, %: 0.1226 0.0047
THD + Noise, %: 0.7154 0.0110
THD + Noise (A-weighted), %: 0.6257 0.0090

Intermodulation distortion
IMD + Noise, %: 0.5387 0.0118
IMD + Noise (A-weighted), %: 0.4722 0.0099

Crosstalk
Parameter L <- R L -> R
Crosstalk at 100 Hz, dB: -100 -102
Crosstalk at 1 kHz, dB: -100 -99
Crosstalk at 10 kHz, dB: -95 -95

IMD (swept tones)
IMD + Noise at 5 kHz, %: 0.0175 0.0116
IMD + Noise at 10 kHz, %: 0.0381 0.0112
IMD + Noise at 15 kHz, %: 0.0891 0.0112




--

-S
If you're a nut and knock on enough doors, eventually someone will open one,
look at you and say, Messiah, we have waited for your arrival.
Anonymous
January 3, 2005 11:01:42 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In article <craagd$em6$1@reader1.panix.com> ssully@panix.com writes:

> -- I did the y-adapter mono test instead, using an audio test
> CD as source. Used the pink noise L/R and spot frequency tracks as tests.
> And what do you know -- the L and R channel difference
> disappeared (channel difference was on the order of 0.05 dB at most).
> It came back to ~1.0 dB when the cables were disconnected.

So when you disconnect the cables, what are you testing that you could
see the 1 dB difference? Have you tried playing a mono record into it?
Or putting your test tone into the phono preamp with it connected to
the sound card? (be sure to knock the level down 40 dB or so!)

> . . . . and the
> channel difference I saw when I was transferring vinyl was either the LPs,
> the TT or the RIAA preamp.

That's what it sounds like to me. But I don't understand the part
about "with nothing connected." Oh, well, as long as you're happy.



--
I'm really Mike Rivers (mrivers@d-and-d.com)
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
Anonymous
January 3, 2005 3:07:11 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

Mon, 3 Jan 2005 02:50:54 +0000 (UTC), Steven Sullivan
<ssully@panix.com> pisze:

>THD + Noise (at -3 dB FS)
>THD, %: 0.1226 0.0047
>THD + Noise, %: 0.7154 0.0110
>THD + Noise (A-weighted), %: 0.6257 0.0090
>
>Intermodulation distortion
>IMD + Noise, %: 0.5387 0.0118
>IMD + Noise (A-weighted), %: 0.4722 0.0099

There is something wrong with left channel. Right is OK, these are
values one can expect form DAP2496, but values for left channel are
comparable to low cost audio, like onboard AC chip.

You can cross-connect wires, to check if distortions come from input
or from output part of card. In either case I think your card is out
of order - it works, but it does not meet required performance.


--
Pozdrowienia

Andrzej Popowski
Anonymous
January 3, 2005 6:41:10 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

Steven Sullivan wrote:
> and here are individual L/R breakdowns.
> Are the IMD and THD discrepancies significant?
>
> Noise level Left Right
> RMS power, dB: -94.8 -94.7
> RMS power (A-weighted), dB: -98.6 -98.4
> Peak level, dB FS: -79.2 -77.0
> DC offset, %: -0.00 -0.00
>
> Dynamic range
> Dynamic range, dB: +91.9 +91.9
> Dynamic range (A-weighted), dB: +93.1 +93.0
> DC offset, %: -0.00 -0.00
>
> THD + Noise (at -3 dB FS)
> THD, %: 0.1226 0.0047
> THD + Noise, %: 0.7154 0.0110
> THD + Noise (A-weighted), %: 0.6257 0.0090
>
> Intermodulation distortion
> IMD + Noise, %: 0.5387 0.0118
> IMD + Noise (A-weighted), %: 0.4722 0.0099
>
> Crosstalk
> Parameter L <- R L -> R
> Crosstalk at 100 Hz, dB: -100 -102
> Crosstalk at 1 kHz, dB: -100 -99
> Crosstalk at 10 kHz, dB: -95 -95
>
> IMD (swept tones)
> IMD + Noise at 5 kHz, %: 0.0175 0.0116
> IMD + Noise at 10 kHz, %: 0.0381 0.0112
> IMD + Noise at 15 kHz, %: 0.0891 0.0112
>
>
>
>
Your soundcard is definitely out of whack somehow. Look how your
distoriton at high frequencies is 3 to 8 times higher on the left
channel as it is on the right.

CD
Anonymous
January 4, 2005 12:05:48 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech Andrzej Popowski <popej@friko.onet.pl> wrote:
> Mon, 3 Jan 2005 02:50:54 +0000 (UTC), Steven Sullivan
> <ssully@panix.com> pisze:

> >THD + Noise (at -3 dB FS)
> >THD, %: 0.1226 0.0047
> >THD + Noise, %: 0.7154 0.0110
> >THD + Noise (A-weighted), %: 0.6257 0.0090
> >
> >Intermodulation distortion
> >IMD + Noise, %: 0.5387 0.0118
> >IMD + Noise (A-weighted), %: 0.4722 0.0099

> There is something wrong with left channel. Right is OK, these are
> values one can expect form DAP2496, but values for left channel are
> comparable to low cost audio, like onboard AC chip.

OK, that accords with my observation that the left channel was wonky.

> You can cross-connect wires, to check if distortions come from input
> or from output part of card. In either case I think your card is out
> of order - it works, but it does not meet required performance.

Good to know, thanks! By cross connect do you mean, R out to L in and
L out to R in?



--

-S
If you're a nut and knock on enough doors, eventually someone will open one,
look at you and say, Messiah, we have waited for your arrival.
Anonymous
January 4, 2005 12:17:43 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech Mike Rivers <mrivers@d-and-d.com> wrote:

> In article <craagd$em6$1@reader1.panix.com> ssully@panix.com writes:

> > -- I did the y-adapter mono test instead, using an audio test
> > CD as source. Used the pink noise L/R and spot frequency tracks as tests.
> > And what do you know -- the L and R channel difference
> > disappeared (channel difference was on the order of 0.05 dB at most).
> > It came back to ~1.0 dB when the cables were disconnected.

> So when you disconnect the cables, what are you testing that you could
> see the 1 dB difference?

Beats me, but *something* is registering on the recording software level meters,
down in the region of -85 or so dB, when I monitor the analog input
path. Have you ever tried this with your card?

> Have you tried playing a mono record into it?
> Or putting your test tone into the phono preamp with it connected to
> the sound card? (be sure to knock the level down 40 dB or so!)

I don't have any mono records, but as noted above I did split the
output of one channel of a stereo CD, into two with a Y-cable,
at the card input, and that gave me even levels on both channels.
That seemed to indicate that the problem was really in the LP chain.
But it now also seems that there's a real problem in the left channel
of the card (see my mightmark IMD results).

> > . . . . and the
> > channel difference I saw when I was transferring vinyl was either the LPs,
> > the TT or the RIAA preamp.

> That's what it sounds like to me. But I don't understand the part
> about "with nothing connected." Oh, well, as long as you're happy.

'with nothing connected' mean: no cables connected to the soundcard.
Then set Audition (for example) to 'record' from the analog inputs.
Which I presumed meant just recording the card's analog path
while it's 'idling' -- not responding to actual input. Or maybe it's recording
'ambient noise' of some sort. Perhaps Arny or someone can say what's
actually going on in 'nothing connected' configuration.

Intersteingly , the noise levels I see from 'nothing connected' are
higher than for a loopback test, where cables connect the card analog
output to its analog input during 'record'.





--

-S
If you're a nut and knock on enough doors, eventually someone will open one,
look at you and say, Messiah, we have waited for your arrival.
Anonymous
January 4, 2005 12:39:22 AM

Archived from groups: rec.audio.pro (More info?)

On Mon, 3 Jan 2005 21:17:43 +0000 (UTC), Steven Sullivan
<ssully@panix.com> wrote:

>I don't have any mono records, but as noted above I did split the
>output of one channel of a stereo CD, into two with a Y-cable,
>at the card input, and that gave me even levels on both channels.
>That seemed to indicate that the problem was really in the LP chain.

1 dB matching in a phono cartridge would be excellent
performance in the real world. Not to worry.

>'with nothing connected' mean: no cables connected to the soundcard.

>Interestingly , the noise levels I see from 'nothing connected' are
>higher than for a loopback test, where cables connect the card analog
>output to its analog input during 'record'.

Open-input noise tests are pretty much meaningless. Shorting
plugs are good enough for telling you something useful.

Good fortune,

Chris Hornbeck
"They'd meet at the Tout Va Bien, a cafe just off the highway."
-JLG, _Bande a part_, 1964
Anonymous
January 4, 2005 2:42:45 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

Mon, 3 Jan 2005 21:05:48 +0000 (UTC), Steven Sullivan
<ssully@panix.com> pisze:

>Good to know, thanks! By cross connect do you mean, R out to L in and
>L out to R in?

Yes. If THD results remain the same, then it means that input circuit
is responsible for distortion.


--
Pozdrowienia

Andrzej Popowski
Anonymous
January 4, 2005 11:17:31 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In article <crccpn$8uj$1@reader1.panix.com> ssully@panix.com writes:

> 'with nothing connected' mean: no cables connected to the soundcard.
> Then set Audition (for example) to 'record' from the analog inputs.
> Which I presumed meant just recording the card's analog path
> while it's 'idling' -- not responding to actual input. Or maybe it's recording
> 'ambient noise' of some sort. Perhaps Arny or someone can say what's
> actually going on in 'nothing connected' configuration.

Well, with the input open-circuited, it's indeed responding to ambient
noise (which may be generated internal to the sound card). I would
expect something like that, and I wouldn't necessarily expect it to be
identical on both channels. If you want to make a true "nothing going
in" test, make up a couple of shorting plugs - an RCA plug with a
jumper wire connected betwen the tip and shield. Plug those in and
then look at the quiescent noise.

It could be that one channel is a tad noisier than the other, and
that's why the THD+M figure is higher. THD isn't just harmonics of the
input signal, it's everything but the input signal, and that includes
self-noise and stray noise pickup.

> Intersteingly , the noise levels I see from 'nothing connected' are
> higher than for a loopback test, where cables connect the card analog
> output to its analog input during 'record'.

That's not surprising either since with "nothing connected" there's no
shielding on the inputs.

--
I'm really Mike Rivers (mrivers@d-and-d.com)
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
Anonymous
January 4, 2005 4:00:27 PM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech Andrzej Popowski <popej@friko.onet.pl> wrote:
> Mon, 3 Jan 2005 21:05:48 +0000 (UTC), Steven Sullivan
> <ssully@panix.com> pisze:

> >Good to know, thanks! By cross connect do you mean, R out to L in and
> >L out to R in?

> Yes. If THD results remain the same, then it means that input circuit
> is responsible for distortion.


It has occurred to me that I might be doing something wrong in setup --
I have these choices when selecting output to the analog jacks

WavOut
Monitor Mixer
S/PDIF
S/PDIF reversed channels
H/W 1/2 In (analog in)

I've been using WavOut because the other inputs produced
warnings of 'level too low' in Rightmark.
Is that correct? Selecting H/W 1/2 In doesn't produce
usable results (due to low levels), and the test aborts.
Only WavOut allows Rightmark to works to completion.

Anyway, I *also* found that in the *Windows* volume control panel,
the fader for WAV/MP3 was set to about halfway. I adjusted this
to full output. I don't know if this was the reason for the
previous results, but when I reran the Rightmark test
(with normal L/R analog loopback connection, input set to
WavOut as before), distortion levels looked much better
for the left channel than before.


RMS power, dB: -94.8 -94.6
RMS power (A-weighted), dB: -98.7 -98.3
Peak level, dB FS: -79.2 -79.2
DC offset, %: 0 0

Dynamic range, dB: 92.1 91.9
Dynamic range (A-weighted), dB: 93.2 93.1
DC offset, %: 0 0

THD, %: 0.0048 0.0047
THD + Noise, %: 0.0113 0.0111
THD + Noise (A-weighted), %: 0.0091 0.0091


IMD + Noise, %: 0.0117 0.0117
IMD + Noise (A-weighted), %: 0.0099 0.01

Crosstalk at 100 Hz, dB: -105 -102
Crosstalk at 1 kHz, dB: -100 -100
Crosstalk at 10 kHz, dB: -95 -95

IMD + Noise at 5 kHz, %: 0.0113 0.0114
IMD + Noise at 10 kHz, %: 0.011 0.0112
IMD + Noise at 15 kHz, %: 0.0112 0.0112


rerunnign the test with crossed connections,
produces essentially the same result (except for
bad crosstalk, as expected)

RMS power, dB: -94.7 -94.6
RMS power (A-weighted), dB: -98.6 -98.3
Peak level, dB FS: -77 -77
DC offset, %: 0 0

Dynamic range, dB: 92.1 91.9
Dynamic range (A-weighted), dB: 93.2 93
DC offset, %: 0 0

THD, %: 0.0049 0.0047
THD + Noise, %: 0.0112 0.0112
THD + Noise (A-weighted), %: 0.0091 0.0091


IMD + Noise, %: 0.0117 0.0117
IMD + Noise (A-weighted), %: 0.0099 0.01

Crosstalk at 100 Hz, dB: -5 -5
Crosstalk at 1 kHz, dB: -5 -5
Crosstalk at 10 kHz, dB: -5 -5

IMD + Noise at 5 kHz, %: 0.0115 0.0114
IMD + Noise at 10 kHz, %: 0.0111 0.0113
IMD + Noise at 15 kHz, %: 0.0111 0.0111


so all looks to be well after all...?
Anonymous
January 5, 2005 3:45:08 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

Tue, 4 Jan 2005 13:00:27 +0000 (UTC), Steven Sullivan
<ssully@panix.com> pisze:

>so all looks to be well after all...?

Looks like you solved the problem :-)


--
Pozdrowienia

Andrzej Popowski
Anonymous
January 6, 2005 12:19:46 AM

Archived from groups: rec.audio.tech,rec.audio.pro (More info?)

In rec.audio.tech Andrzej Popowski <popej@friko.onet.pl> wrote:
> Tue, 4 Jan 2005 13:00:27 +0000 (UTC), Steven Sullivan
> <ssully@panix.com> pisze:

> >so all looks to be well after all...?

> Looks like you solved the problem :-)

I wish it was always that easy ;>

Thanks to everyone for their patience and assistance.


--

-S
If you're a nut and knock on enough doors, eventually someone will open one,
look at you and say, Messiah, we have waited for your arrival.
!