[Paul Stamler]

Hi folks:

I had the computer rigged for testing stuff, so I figured I'd run some
comparisons between a Yamaha GQ1031BII graphic equalizer and a Behringer
Ultragraph Pro (their analog EQ). (I'd borrowed both for a live-sound gig
the next day.) One of the tests I did was 19kHz + 19.5kHz, mixed at equal
levels and fed into the unit under test at +10dBu equivalent level (same P-P
level as a +10dBu sine wave), or 6dB above nominal +4dBu, not an outrageous
level. All sliders were set flat, and the level controls were set for unity
gain. I fed the equalizers from, and to, a CardDeluxe sound card, which has
a 10k input impedance, and I recorded the signals in Adobe Audition. Once
they were in the DAW, I normalized each signal to -0.2dBFS. Then I looked at
the spectra with Audition's spectrum analyzer. The Yamaha's spectrum is here
(pardon the only OK visual resolution; click on the picture for a better
look):

http://twin-x.com/groupdiy/displayimage.php?pos=-2675

And the Behringer's spectrum is he

http://twin-x.com/groupdiy/displayimage.php?pos=-2674

Rather a graphic difference, if you pardon the expression.

They performed differently on 400Hz harmonic distortion tests too (the
Yamaha's harmonics disappeared above the 6th, while the Behr's continued out
to the limit of the audible spectrum), but the pictures weren't as pretty.

I've maintained for a while that IM tests like this are a useful part of the
test battery (not at all an original observation -- Norman Crowhurst made
the point when I was five years old and Eisenhower was president). This is
one of the more striking contrasts I've seen.

You can guess which EQ I used on the house speakers.

Peace,
Paul

[George's Pro Sound Company]

"Paul Stamler" wrote in message
news
Hi folks:

I had the computer rigged for testing stuff, so I figured I'd run some
comparisons between a Yamaha GQ1031BII graphic equalizer and a Behringer
Ultragraph Pro (their analog EQ). (I'd borrowed both for a live-sound gig
the next day.)

Paul, one should not need those test to know the behringer analouge eq sound
like ass
your ears should have been more than adaquate
now behringer digital eq is quite a nice unit, deq2496
George

[Arny Krueger]

"Paul Stamler" wrote in message
news
I had the computer rigged for testing stuff, so I figured
I'd run some comparisons between a Yamaha GQ1031BII
graphic equalizer and a Behringer Ultragraph Pro (their
analog EQ). (I'd borrowed both for a live-sound gig the
next day.) One of the tests I did was 19kHz + 19.5kHz,
mixed at equal levels and fed into the unit under test at
+10dBu equivalent level (same P-P level as a +10dBu sine
wave), or 6dB above nominal +4dBu, not an outrageous
level. All sliders were set flat, and the level controls
were set for unity gain. I fed the equalizers from, and
to, a CardDeluxe sound card, which has a 10k input
impedance, and I recorded the signals in Adobe Audition.

Many of us are familiar with that exact configuration for testing. It can
really work well, as your subsequent results show.

Once they were in the DAW, I normalized each signal to
-0.2dBFS. Then I looked at the spectra with Audition's
spectrum analyzer. The Yamaha's spectrum is here (pardon
the only OK visual resolution; click on the picture for a
better look):

http://twin-x.com/groupdiy/displayimage.php?pos=-2675

And the Behringer's spectrum is he

http://twin-x.com/groupdiy/displayimage.php?pos=-2674

Frankly, neither of the products had performance that thrilled me.

Rather a graphic difference, if you pardon the expression.

Graphic eqs are hard to do right.

They performed differently on 400Hz harmonic distortion
tests too (the Yamaha's harmonics disappeared above the
6th, while the Behr's continued out to the limit of the
audible spectrum), but the pictures weren't as pretty.

Nonlinearity is what we are worried about. THD and IM are just abstractions
that we use to characterize nonlinearity.

I've maintained for a while that IM tests like this are a
useful part of the test battery (not at all an original
observation -- Norman Crowhurst made the point when I was
five years old and Eisenhower was president). This is one
of the more striking contrasts I've seen.

I thoroughly agree. IM has the greatest potential for producing aharmonic
(strike 1) spurious responses (strike 2) in the range of frequencies where
the ear is very sensitive (strike 3). Any nonlinearity that produces THD
will produce IM if you come up with the right test signal. Nonlinear
distortion tends to increase with frequency. It is possible for low pass
filtering to conceal the THD spuriae, while appropriate IM tests will bring
them to the forefront.

You can guess which EQ I used on the house speakers.

I like to see all artifacts at -100 or better, for any signal that is at
least 3 dB below clipping. I'll take -90 with a smile, and tolerate -80.

Thing is, this relevant and significant test applies to just one set of
products. Little can be inferred from them to the rest of the respective
product lines.

For example, I did similar tests on a small Behringer mixer a while back,
and found no such problems.

Besides, I'm not exactly a fan of graphic eqs. ;-)

[Scott Dorsey]

In article ,
George's Pro Sound Company wrote:

"Paul Stamler" wrote in message
news
Hi folks:

I had the computer rigged for testing stuff, so I figured I'd run some
comparisons between a Yamaha GQ1031BII graphic equalizer and a Behringer
Ultragraph Pro (their analog EQ). (I'd borrowed both for a live-sound gig
the next day.)

Paul, one should not need those test to know the behringer analouge eq sound
like ass
your ears should have been more than adaquate

No, but you need those tests to know WHY they are that way.

I'd like to see a comparison test on that awful Alesis equalizer too.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Mark]

On Oct 20, 9:16*am, (Scott Dorsey) wrote:
In article ,
George's Pro Sound Company wrote:



"Paul Stamler" wrote in message
news
Hi folks:

I had the computer rigged for testing stuff, so I figured I'd run some
comparisons between a Yamaha GQ1031BII graphic equalizer and a Behringer
Ultragraph Pro (their analog EQ). (I'd borrowed both for a live-sound gig
the next day.)

Paul, one should not need those test to know the behringer analouge eq sound
like ass
your ears should have been more than adaquate

No, but you need those tests to know WHY they are that way.

I'd like to see a comparison test on that awful Alesis equalizer too.
--scott

--
"C'est un Nagra. *C'est suisse, et tres, tres precis."

Paul,

Did the results change much at 6dB down at +4dBu instead of +10 dBu?


Mark

[Arny Krueger]

"Chel van Gennip" wrote in message

Mark schreef:
On Oct 20, 9:16 am, (Scott Dorsey)
wrote:
In article
,
George's Pro Sound Company wrote:
"Paul Stamler" wrote in
message
news Hi folks:
I had the computer rigged for testing stuff, so I
figured I'd run some comparisons between a Yamaha
GQ1031BII graphic equalizer and a Behringer
Ultragraph Pro (their analog EQ). (I'd borrowed both
for a live-sound gig the next day.)
Paul, one should not need those test to know the
behringer analouge eq sound like ass
your ears should have been more than adaquate
No, but you need those tests to know WHY they are that
way.

I'd like to see a comparison test on that awful Alesis
equalizer too. --scott

Did the results change much at 6dB down at +4dBu instead
of +10 dBu?


And does your CardDeluxe have balanced inputs that can
handle +10db

It might, just. As a rule audio interfaces have some headroom, from 6 to 10
dB or more above +4 dbu.

However, their performance in the last dB before clipping should not be
depended on.

I'm presuming that the usual trick of front-ending the audio interface with
an attenuator was used.

Some AES standards for digital audio testing can be read as recommending
that the tests be run 1 to 3 dB below FS. I presume that applies to not only
testing the interfaces themselves, but also to using them as test equipment.

I always loop my test equipment and collect test equipment residuals under
conditions that duplicate the test itself, before I start characterizing the
UUT.

[Arny Krueger]

"Scott Dorsey" wrote in message

In article

I'd like to see a comparison test on that awful Alesis
equalizer too.

http://www.pcavtech.com/sig_proc/MEQ230/index.htm

I don't see where the operating level for the tests is recorded on the web
page. My recollection is +4 into a 10K load, all knobs set to detent flat.
But, it might have been much higher, maybe +10 or more. It had to be pretty
high to get the noise floor so far down.

[Eeyore]

Arny Krueger wrote:

Besides, I'm not exactly a fan of graphic eqs. ;-)

Me neither.

Graham

[Eeyore]

Scott Dorsey wrote:

George's Pro Sound Company wrote:
"Paul Stamler" wrote in message

I had the computer rigged for testing stuff, so I figured I'd run some
comparisons between a Yamaha GQ1031BII graphic equalizer and a Behringer
Ultragraph Pro (their analog EQ). (I'd borrowed both for a live-sound gig
the next day.)

Paul, one should not need those test to know the behringer analouge eq sound
like ass your ears should have been more than adaquate

No, but you need those tests to know WHY they are that way.

I'd like to see a comparison test on that awful Alesis equalizer too.

The DSP one ?

Graham

[Scott Dorsey]

Eeyore wrote:
Scott Dorsey wrote:

George's Pro Sound Company wrote:
"Paul Stamler" wrote in message

I had the computer rigged for testing stuff, so I figured I'd run some
comparisons between a Yamaha GQ1031BII graphic equalizer and a Behringer
Ultragraph Pro (their analog EQ). (I'd borrowed both for a live-sound gig
the next day.)

Paul, one should not need those test to know the behringer analouge eq sound
like ass your ears should have been more than adaquate

No, but you need those tests to know WHY they are that way.

I'd like to see a comparison test on that awful Alesis equalizer too.

The DSP one ?

No, the 230 that Arny actually provided some measurements on. The DSP one
is probably a big step up. Hell, that Behringer is probably a big step up.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Paul Stamler]

"Mark" wrote in message
...

Did the results change much at 6dB down at +4dBu instead of +10 dBu?

I didn't look. Just made the one test.

Peace,
Paul

[Paul Stamler]

"Chel van Gennip" wrote in message
...

Did the results change much at 6dB down at +4dBu instead of +10 dBu?


Mark


And does your CardDeluxe have balanced inputs that can handle +10db

Yes. It's set for input and output sensitivity of -10dBV (because I use it a
lot for remastering 78s from a consumer-level preamp); maximum in/out is
+12dBu.

Peace,
Paul

[George's Pro Sound Company]

"Scott Dorsey" wrote in message
...
In article ,
George's Pro Sound Company wrote:

"Paul Stamler" wrote in message
news
Hi folks:

I had the computer rigged for testing stuff, so I figured I'd run some
comparisons between a Yamaha GQ1031BII graphic equalizer and a Behringer
Ultragraph Pro (their analog EQ). (I'd borrowed both for a live-sound
gig
the next day.)

Paul, one should not need those test to know the behringer analouge eq
sound
like ass
your ears should have been more than adaquate

No, but you need those tests to know WHY they are that way.

what I was amazed at was when w hooked the K-T up
it was just as awful, not in distortion but in not being at all accurate on
the iso centers
and the amount it affected adjecent bands
that being said, one can get a good sound from one, though I am not sure why
looking at its tests
George

[Paul Stamler]

"Arny Krueger" wrote in message
...

I'm presuming that the usual trick of front-ending the audio interface
with
an attenuator was used.

Yes.

I always loop my test equipment and collect test equipment residuals under
conditions that duplicate the test itself, before I start characterizing
the
UUT.

The loopback is he

http://www.twin-x.com/groupdiy/displ...0135&pos=-2682

Peace,
Paul

[Paul Stamler]

"Arny Krueger" wrote in message
...

I always loop my test equipment and collect test equipment residuals under
conditions that duplicate the test itself, before I start characterizing
the
UUT.

Easier URL for the loopback test, done through the attenuator:

http://twin-x.com/groupdiy/displayimage.php?pos=-2682

Peace,
Paul

[Paul Stamler]

"Chel van Gennip" wrote in message
...

Here tests are beyond specifications of both the tested equipment and
the test equipment. The things showing in the results might be a result
of one system amplifying 1 dB without any distortion, with the sliders
at unity gain.

Begging your pardon? These tests were run at 6dB over nominal level of
+4dBu, which is 9dB below the clipping level of both pieces of equipment.
That's a minimal amount of headroom over nominal, and there's still 9dB to
go before the DUT's overload point. Any competent pro gear should handle
that with ease -- as the Yamaha does.

As for the test setup, I refer you to the loopback test; it was putting
out -2dBFS, receiving -22dBFS. Again, that should be well within the
capacity of professional equipment, and it is.

Peace,
Paul

[Paul Stamler]

"Arny Krueger" wrote in message
. ..

Thing is, this relevant and significant test applies to just one set of
products. Little can be inferred from them to the rest of the respective
product lines.

Of course.

For example, I did similar tests on a small Behringer mixer a while back,
and found no such problems.

Besides, I'm not exactly a fan of graphic eqs. ;-)

Nor am I, but in the hall I was working on Friday, you really need an EQ,
and there wasn't a parametric available.

Peace,
Paul

[Paul Stamler]

"Chel van Gennip" wrote in message
...
Arny Krueger schreef:
"Paul Stamler" wrote in message
news
(I'd borrowed both for a live-sound gig the
next day.) One of the tests I did was 19kHz + 19.5kHz,
mixed at equal levels and fed into the unit under test at
+10dBu equivalent level (same P-P level as a +10dBu sine
wave), or 6dB above nominal +4dBu, not an outrageous
level.

...
I like to see all artifacts at -100 or better, for any signal that is at
least 3 dB below clipping. I'll take -90 with a smile, and tolerate -80.


This test was NOT done 3 dB below clipping!

No, it was done 9dB below clipping.

Peace,
Paul

[Scott Dorsey]

George's Pro Sound Company wrote:
it was just as awful, not in distortion but in not being at all accurate on
the iso centers
and the amount it affected adjecent bands
that being said, one can get a good sound from one, though I am not sure why
looking at its tests

What amazed me about the Alesis is that when I'd move one slider, I'd get
response changes several bands away... and sometimes unpredictable ones.
I'd start cutting at one frequency and another unrelated frequency would
start to feed back. It was completely uncontrollable.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Arny Krueger]

"Paul Stamler" wrote in message

"Arny Krueger" wrote in message
...

I'm presuming that the usual trick of front-ending the
audio interface with an attenuator was used.

Yes.

I always loop my test equipment and collect test
equipment residuals under conditions that duplicate the
test itself, before I start characterizing the UUT.

The loopback is he

http://www.twin-x.com/groupdiy/displ...0135&pos=-2682


Other than some minor differences in FFT size and scaling, that looks a lot
like:

http://www.pcavtech.com/soundcards/C..._2496-a_FS.gif


The rest of my collection of CardDeluxe test results is at

http://www.pcavtech.com/soundcards/C...luxe/index.htm

[Arny Krueger]

"Glenn Carlson" wrote in message

On 20 Oct 2008 14:30:22 -0400, Scott Dorsey wrote:

George's Pro Sound Company wrote:
it was just as awful, not in distortion but in not
being at all accurate on the iso centers
and the amount it affected adjecent bands
that being said, one can get a good sound from one,
though I am not sure why looking at its tests

What amazed me about the Alesis is that when I'd move
one slider, I'd get response changes several bands
away... and sometimes unpredictable ones. I'd start
cutting at one frequency and another unrelated frequency
would start to feed back. It was completely
uncontrollable. --scott

A friend of mine returned 3 of them to the store because
of that behavior! What a lump that unit is.

That's a sign of a common form of breakage.

[Paul Stamler]

"Arny Krueger" wrote in message
. ..
The loopback is he


http://www.twin-x.com/groupdiy/displ...0135&pos=-2682


Other than some minor differences in FFT size and scaling, that looks a
lot
like:

http://www.pcavtech.com/soundcards/C..._2496-a_FS.gif

As it ought to. Incidentally, I looked hard at Arny's collection of tests
before settling on the CardDeluxe. If I coulda afforded a Lynx I woulda, but
this gave me the best bang for the buck. The one thing I wish is that the
CardDeluxe let me switch back and forth between nominal -10dBV and +4dBu in
software, rather than having to open up the box and play with hardware
switches.

Peace,
Paul

The rest of my collection of CardDeluxe test results is at

http://www.pcavtech.com/soundcards/C...luxe/index.htm

[Paul Stamler]

"Chel van Gennip" wrote in message
...
Paul Stamler schreef:
"Chel van Gennip" wrote in message
...

Here tests are beyond specifications of both the tested equipment and
the test equipment. The things showing in the results might be a result
of one system amplifying 1 dB without any distortion, with the sliders
at unity gain.

Begging your pardon? These tests were run at 6dB over nominal level of
+4dBu, which is 9dB below the clipping level of both pieces of
equipment.
That's a minimal amount of headroom over nominal, and there's still 9dB
to
go before the DUT's overload point. Any competent pro gear should handle
that with ease -- as the Yamaha does.

As for the test setup, I refer you to the loopback test; it was putting
out -2dBFS, receiving -22dBFS. Again, that should be well within the
capacity of professional equipment, and it is.


I see the loopback test uses an attenuator, and gives an input value of
-22dB FS, and the other tests have an input level that shows as -2 dB FS
at the center frequency and a lot of energy at other frequencies,
indicating a total input level of about 0 dB FS. No way to tell this way
if the actual input level was eg. +2db FS. Normally an ADC does not
handle levels 6dB above FS very well.

The loopback was done in exactly the same way as the Yamaha and Behringer
tests, with the exception that the output cable bearing the test signal was
connected directly to the attenuator rather than first passing through a DUT
(in this case an EQ). The signals were normalized after recording
to -0.2dBFS.

By the way, here's another way of looking at the results:

Loopback: Highest distortion product(s): -114dB @ 18.5kHz, 20kHz
Yamaha: Highest distortion product: -97dB @ 500Hz
Behringer: Highest distortion product(s): -59dB @ 18.5kHz, 20kHz

Incidentally, the Yamaha's highest harmonic distortion product at 440Hz was
2nd harmonic at -92dB. Pretty darn clean.

Like I said, this is a useful test.

Peace,
Paul

[Arny Krueger]

"Paul Stamler" wrote in message

"Arny Krueger" wrote in message
. ..
The loopback is he


http://www.twin-x.com/groupdiy/displ...0135&pos=-2682

Other than some minor differences in FFT size and
scaling, that looks a lot like:

http://www.pcavtech.com/soundcards/C..._2496-a_FS.gif

As it ought to.

Exactly.

Incidentally, I looked hard at Arny's
collection of tests before settling on the CardDeluxe. If
I coulda afforded a Lynx I woulda, but this gave me the
best bang for the buck.

One can get fairly close to the LynxTwo level of performance with some of
the recent eMu cards, if published tests are reliable. They are only a tiny
fraction of the price of the LynxTwo. I currently use a M-Audio AP 24192
for most of my measurements.

The one thing I wish is that the
CardDeluxe let me switch back and forth between nominal
-10dBV and +4dBu in software, rather than having to open
up the box and play with hardware switches.

At one point I hooked an external dual pot up to the gain-setting pins on my
Card Deluxe in such a way that I had continuously-variable input
sensitivity.

[Mark]

On Oct 21, 8:15*am, "Arny Krueger" wrote:
"Paul Stamler" wrote in message



"Arny Krueger" wrote in message
...
The loopback is he

http://www.twin-x.com/groupdiy/displ...andom&cat=1013...
Other than some minor differences in FFT size and
scaling, that looks a lot like:
http://www.pcavtech.com/soundcards/C..._2496-a_FS.gif
As it ought to.

Exactly.

Incidentally, I looked hard at Arny's
collection of tests before settling on the CardDeluxe. If
I coulda afforded a Lynx I woulda, but this gave me the
best bang for the buck.

One can get fairly close to the LynxTwo level of performance with some of
the recent eMu cards, if published tests are reliable. They are only a tiny
fraction of the price of the LynxTwo. *I currently use a M-Audio AP 24192
for most of my measurements.

*The one thing I wish is that the
CardDeluxe let me switch back and forth between nominal
-10dBV and +4dBu in software, rather than having to open
up the box and play with hardware switches.

At one point I hooked an external dual pot up to the gain-setting pins on my
Card Deluxe in such a way that I had continuously-variable input
sensitivity.

Paul,

You seem to be catching a lot of grief about these tests. I want to
point out that doing actual instrumented measurement tests like this
is COMMENDABLE compared to subjective listening tests. So even
though you are catching grief, it is still a good thing. And you have
good answers for most of the questions.

Anytime someone "hears" something bad, the very next step should be a
quantitative measurement to verify and determine what and why it
sounds bad.

thanks

Mark

[Paul Stamler]

"Mark" wrote in message
...

You seem to be catching a lot of grief about these tests. I want to
point out that doing actual instrumented measurement tests like this
is COMMENDABLE compared to subjective listening tests. So even
though you are catching grief, it is still a good thing. And you have
good answers for most of the questions.

Anytime someone "hears" something bad, the very next step should be a
quantitative measurement to verify and determine what and why it
sounds bad.

Well, I agree. I hear stuff, especially from cheap gear, and one of my chief
interests is to understand *why* I'm hearing it. And if the quantitative
tests I perform don't find anything, I look for new tests that might.

The flip side, of course, is that if you can find a way to test for
something scruddy, you can use that test to design it out of the next piece
of gear.

Peace,
Paul

[Arny Krueger]

"Glenn Carlson" wrote in message

On Mon, 20 Oct 2008 15:51:42 -0400, Arny Krueger wrote:

"Glenn Carlson" wrote in message

On 20 Oct 2008 14:30:22 -0400, Scott Dorsey wrote:

George's Pro Sound Company wrote:
it was just as awful, not in distortion but in not
being at all accurate on the iso centers
and the amount it affected adjecent bands
that being said, one can get a good sound from one,
though I am not sure why looking at its tests

What amazed me about the Alesis is that when I'd move
one slider, I'd get response changes several bands
away... and sometimes unpredictable ones. I'd start
cutting at one frequency and another unrelated
frequency would start to feed back. It was completely
uncontrollable. --scott

A friend of mine returned 3 of them to the store because
of that behavior! What a lump that unit is.

That's a sign of a common form of breakage.

Could be!
But 3 in a row?

Common mishandling in packing and shipping, for example.

Or, a the store is simply recycling returns.GC is well known to me for doing
that.

I believe that the failure is induced by dropping the eq onto a flat
surface, face down.

Manufacturing defect maybe?

Could also be.

[Scott Dorsey]

Arny Krueger wrote:

A friend of mine returned 3 of them to the store because
of that behavior! What a lump that unit is.

That's a sign of a common form of breakage.

Could be!
But 3 in a row?

Common mishandling in packing and shipping, for example.

Or, a the store is simply recycling returns.GC is well known to me for doing
that.

I believe that the failure is induced by dropping the eq onto a flat
surface, face down.

Manufacturing defect maybe?

Could also be.

I have used two of the things and they were both the same way. If this
is a defect caused by mishandling, it has happened to an awful high
proportion of the units discussed.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."