I made a test CD using computer software (CEP) to generate sine waveforms to
test a home audio system... (the equipment you guys record for! ) BUT it sounded
very harsh... so I changed the levels down to -3db, and then to -6db, and now it
sounds better...

Isn't that weird?

What levels should I expect HI-FI gear to work at? It seems a waste to
not use the full spectrum, but I do see music CDs that seem to go up to
clipping and still sound OK... normally this system sounds very good with CDs.

Anyway - is there some trick to making a test CD in regards to levels?

Does CEP create bad waves?

thanks

Bob wrote:

I made a test CD using computer software (CEP) to generate sine waveforms to
test a home audio system... (the equipment you guys record for! ) BUT it sounded
very harsh... so I changed the levels down to -3db, and then to -6db, and now it
sounds better...

Isn't that weird?

I bet a nickel that if you look at the output with a scope, you will find
that your CD player has a linearity problem slightly below 0dBFS. This is
sadly very common with consumer players.

What levels should I expect HI-FI gear to work at? It seems a waste to
not use the full spectrum, but I do see music CDs that seem to go up to
clipping and still sound OK... normally this system sounds very good with CDs.

I bet it would sound better with a good DAC... even a used DAC In the Box
maybe.

Anyway - is there some trick to making a test CD in regards to levels?

No, I think you have a player problem. At least you aren't anywhere near
the first one.
--scott

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

"Scott Dorsey"
Bob

I made a test CD using computer software (CEP) to generate sine waveforms
to
test a home audio system... (the equipment you guys record for! ) BUT it
sounded
very harsh... so I changed the levels down to -3db, and then to -6db, and
now it
sounds better...

Isn't that weird?

I bet a nickel that if you look at the output with a scope, you will find
that your CD player has a linearity problem slightly below 0dBFS. This is
sadly very common with consumer players.


** More utter bull**** from Dorsey to whom all facts are optional.


What levels should I expect HI-FI gear to work at? It seems a waste to
not use the full spectrum, but I do see music CDs that seem to go up to
clipping and still sound OK... normally this system sounds very good with
CDs.

I bet it would sound better with a good DAC... even a used DAC In the Box
maybe.


** Pig ignorant audiophool garbage.


Anyway - is there some trick to making a test CD in regards to levels?

No, I think you have a player problem.


** Practically zero chance that is the case.



.......... Phil

"Scott Dorsey" wrote in message
...
Bob wrote:

I made a test CD using computer software (CEP) to generate sine
waveforms to
test a home audio system... (the equipment you guys record for! ) BUT it
sounded
very harsh... so I changed the levels down to -3db, and then to -6db, and
now it
sounds better...

Isn't that weird?

I bet a nickel that if you look at the output with a scope, you will find
that your CD player has a linearity problem slightly below 0dBFS. This is
sadly very common with consumer players.

You want to really find nonlinearity problems in DACs...generate three tones
in CEP, at 9.6kHz, 19kHz and 19.5kHz, each at (say) -1dBFS, then mix them
down at equal levels so the composite wave totals -0.1dBFS. Play that back
through your DAC, attenuate a few dB, and record the attenuated result into
a pro-quality sound card. You'll be surprised how much gunk a lot of DACs
put out.

For DACs, this seems to be a real torture test. Oh, it's a torture test for
tweeters too, so *don't* play it back through your speakers!

Peace,
Paul

On Tue, 11 Jan 2005 06:01:14 GMT, "Paul Stamler"
wrote:

You want to really find nonlinearity problems in DACs...generate three tones
in CEP, at 9.6kHz, 19kHz and 19.5kHz, each at (say) -1dBFS, then mix them
down at equal levels so the composite wave totals -0.1dBFS. Play that back
through your DAC, attenuate a few dB, and record the attenuated result into
a pro-quality sound card. You'll be surprised how much gunk a lot of DACs
put out.

But everybody knows that digital is (by definition) perfect. It's
been proven repeatedly and endlessly in ABCDE testing. Because
"Truth is stranger than fact."

Chris Hornbeck
"Conscious that we must have sprung from somewhere, we are
lured to the riddle of our origins." -Malcolm W. Browne

digital can never be perfect, long live the acetate ripping needle!!!!!

"siguy" wrote in message
oups.com...
digital can never be perfect, long live the acetate ripping needle!!!!!

Exactly. If you're not going to get perfect, you may as well get as far away
from it as you can eh? :-)

MrT.

"Chris Hornbeck" wrote in message
...

But everybody knows that digital is (by definition) perfect.

Exactly "by definition" only.

In the real world we even come a hell of a lot closer than anyone would have
expected 30 years ago!

MrT.

"Paul Stamler" wrote in message


You want to really find nonlinearity problems in DACs...generate
three tones in CEP, at 9.6kHz, 19kHz and 19.5kHz, each at (say)
-1dBFS, then mix them down at equal levels so the composite wave
totals -0.1dBFS. Play that back through your DAC, attenuate a few dB,
and record the attenuated result into a pro-quality sound card.
You'll be surprised how much gunk a lot of DACs put out.

Agreed.

Been there, done that.

That's one reason why I've been telling people around here to set peak
levels at -1 dB for years.

It's also the reason why I've been telling people to do their Rightmark
tests with 18 & 20 KHz or 19 & 20 KHz CCIR-like IM test tones, not the
SMPTE-like easy stuff that it uses by default.

It can find a lot of converters with their hands in the cookie jar, too.

"siguy" wrote in message
oups.com

digital can never be perfect, long live the acetate ripping
needle!!!!!

Yes, I regret to admit that I just spent a lot of my winter vacation camping
in the Anza-Barrago (California) and Baja Mexico (Erindera) deserts and
failed to bring back any cactus needles.

But everybody knows that digital is (by definition) perfect.

Huh ???

May I please see that definition....

No, I think you have a player problem.


** Practically zero chance that is the case.


So what is the problem?

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

Paul Stamler wrote:
Scott Dorsey writes:

I bet a nickel that if you look at the output with a scope, you will find
that your CD player has a linearity problem slightly below 0dBFS. This is
sadly very common with consumer players.

You want to really find nonlinearity problems in DACs...generate three tones
in CEP, at 9.6kHz, 19kHz and 19.5kHz, each at (say) -1dBFS, then mix them
down at equal levels so the composite wave totals -0.1dBFS. Play that back
through your DAC, attenuate a few dB, and record the attenuated result into
a pro-quality sound card. You'll be surprised how much gunk a lot of DACs
put out.

I can believe it. I'm not eve going that far, though, I am just talking
about analogue electronics after the converters that go into clipping before
the converter does.
--scott


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

"Laurence Payne" wrote in
message
No, I think you have a player problem.


** Practically zero chance that is the case.


So what is the problem?

Probably a player problem.

I say this because its hard to generate a clipped waveform in Audition/CE
without going out of your way.

But, the initial defaults are to make a wave that peaks at 0 dB.

On Tue, 11 Jan 2005 10:42:15 -0500, "Arny Krueger"
wrote:

No, I think you have a player problem.


** Practically zero chance that is the case.


So what is the problem?

Probably a player problem.

So it's what Scott said? :-)

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

"Laurence Payne" wrote in
message
On Tue, 11 Jan 2005 10:42:15 -0500, "Arny Krueger"
wrote:

No, I think you have a player problem.


** Practically zero chance that is the case.


So what is the problem?

Probably a player problem.

So it's what Scott said? :-)

(Putting fireproof suit on)

Yup!

;-)

In article writes:

I made a test CD using computer software (CEP) to generate sine waveforms to
test a home audio system... (the equipment you guys record for! ) BUT it
sounded
very harsh... so I changed the levels down to -3db, and then to -6db, and now
it
sounds better...

If you made the disk at full scale, that may have indeed put too hot a
signal into your system from the CD player. In fact, a lot of CD
players themselves can't even handle a full scale signal without some
distortion.

Anyway - is there some trick to making a test CD in regards to levels?

Make it at -10 dBFS or lower, maybe -20.


--
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 he double-m-eleven-double-zero at yahoo

Anyway - is there some trick to making a test CD in regards to levels?

Make it at -10 dBFS or lower, maybe -20.

Make SOME tracks at that level, others at full-scale.
Looks like your playback setup fails the 0dB test. Isn't that what
tests are for?

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

Just burnt a CD with 1k tone at 100%, then a little under. Computer
plays it OK. Sony CD player (into the same playback system) can't
cope with the 100% tone. Sounds like vinyl mistracking.

Took it next door to the crappy set-top DVD player. TV cabinet
shakes a bit with the 1k tone, but it's clean.

I guess the culprit has to be the CD player.

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

Chris Hornbeck wrote:
On Tue, 11 Jan 2005 06:01:14 GMT, "Paul Stamler"
wrote:

You want to really find nonlinearity problems in DACs...generate
three tones
in CEP, at 9.6kHz, 19kHz and 19.5kHz, each at (say) -1dBFS, then mix
them
down at equal levels so the composite wave totals -0.1dBFS. Play
that back
through your DAC, attenuate a few dB, and record the attenuated
result into
a pro-quality sound card. You'll be surprised how much gunk a lot of
DACs
put out.

But everybody knows that digital is (by definition) perfect. It's
been proven repeatedly and endlessly in ABCDE testing. Because
"Truth is stranger than fact."

Yes, that's a good test, but before you go blame digital for everything
bad,
try the same test with analog tape or vinal. Ha.

Mark

....which is surprising, given how hot current popular music cd's are
mastered and replicated. One would think the better consumer players
manufactured now would handle the max levels possible.

Laurence Payne wrote:

Just burnt a CD with 1k tone at 100%, then a little under. Computer
plays it OK. Sony CD player (into the same playback system) can't
cope with the 100% tone. Sounds like vinyl mistracking.

Took it next door to the crappy set-top DVD player. TV cabinet
shakes a bit with the 1k tone, but it's clean...

In article ,
Laurence Payne wrote:
Just burnt a CD with 1k tone at 100%, then a little under. Computer
plays it OK. Sony CD player (into the same playback system) can't
cope with the 100% tone. Sounds like vinyl mistracking.

Took it next door to the crappy set-top DVD player. TV cabinet
shakes a bit with the 1k tone, but it's clean.

I guess the culprit has to be the CD player.

Yes. As I said, this is sadly very common. The analogue section clips
or it least gets very ugly at a lower level than the converter itself.

Personally I believe there is no excuse for that sort of thing.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On 11 Jan 2005 12:41:08 -0500, (Scott Dorsey) wrote:

As I said, this is sadly very common. The analogue section clips
or it least gets very ugly at a lower level than the converter itself.

Personally I believe there is no excuse for that sort of thing.

Maybe they should be using LM301's in the I/V convertor. I
hear they're just about perfect.

Chris Hornbeck
"Conscious that we must have sprung from somewhere, we are
lured to the riddle of our origins." -Malcolm W. Browne

"Scott Dorsey" wrote in message
...
In article ,
Laurence Payne wrote:
Just burnt a CD with 1k tone at 100%, then a little under. Computer
plays it OK. Sony CD player (into the same playback system) can't
cope with the 100% tone. Sounds like vinyl mistracking.

Took it next door to the crappy set-top DVD player. TV cabinet
shakes a bit with the 1k tone, but it's clean.

I guess the culprit has to be the CD player.

Yes. As I said, this is sadly very common. The analogue section clips
or it least gets very ugly at a lower level than the converter itself.

Personally I believe there is no excuse for that sort of thing.

I'm not entirely sure the problem's in the analog sections. Standard output
from a DAC at 0dBS is 2Vrms, which is about 5.64Vpp. Most of the analog
stages run from +/- 15V rails, and such a stage should have no problem with
5.64Vpp. I would guess, without going inside a player with a scope, that
what's going on is digital clipping someplace, perhaps in the filtering
stage.

Peace,
Paul

"siguy" wrote in message
oups.com

digital can never be perfect, long live the acetate ripping
needle!!!!!

Actually, the fault being described is probably on the analog side of the
converters.

"Chris Hornbeck" wrote in message

On 11 Jan 2005 12:41:08 -0500, (Scott Dorsey) wrote:

As I said, this is sadly very common. The analogue section clips
or it least gets very ugly at a lower level than the converter
itself.

Or, there is clipping in the digital filter becuase its possible for signal
values to go above FS among intermediate values in a digital filter.

Personally I believe there is no excuse for that sort of thing.

Agreed, and it seems to be a lot less common. One thing that seems to have
happened is that a lot of very cheap suppliers have been blown off the
market by good suppliers like AKM and Crystal.

Maybe they should be using LM301's in the I/V convertor. I
hear they're just about perfect.

In fact they are usually using 5558s, which are something like a
dual-section, fixed-compensated LM301.

"Arny Krueger" wrote in message

Agreed.

Been there, done that.

That's one reason why I've been telling people around here to set peak
levels at -1 dB for years.

It's also the reason why I've been telling people to do their Rightmark
tests with 18 & 20 KHz or 19 & 20 KHz CCIR-like IM test tones, not the
SMPTE-like easy stuff that it uses by default.

It can find a lot of converters with their hands in the cookie jar, too.


LPs can sound a bit crappy with FS 19 + 19.x signals 'encoded' too ! Just
as well nothing remotely like that happens in real life - sort of a 'natural
de-emphasis' factor helps there.

geoff

On Tue, 11 Jan 2005 12:37:10 -0500, Greg Taylor
wrote:

...which is surprising, given how hot current popular music cd's are
mastered and replicated. One would think the better consumer players
manufactured now would handle the max levels possible.

Today's commercial recordings aren't QUITE sustained sine tones at
100% level :-)

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

On 11 Jan 2005 12:41:08 -0500, (Scott Dorsey) wrote:

Yes. As I said, this is sadly very common. The analogue section clips
or it least gets very ugly at a lower level than the converter itself.

Personally I believe there is no excuse for that sort of thing.

I wonder if there's any adjustment possible. This particular DC
player has always sounded "louder" than the original material.
Almost as if Sony thought it would sound more impressive if the signal
ran into a bit of limiting.

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

On Wed, 12 Jan 2005 08:15:12 +1300, "Geoff Wood"
wrote:

LPs can sound a bit crappy with FS 19 + 19.x signals 'encoded' too ! Just
as well nothing remotely like that happens in real life - sort of a 'natural
de-emphasis' factor helps there.

Yeah. It's of little practical significance how any player reproduces
the sort of test tone we're discussing.

CubaseFAQ www.laurencepayne.co.uk/CubaseFAQ.htm
"Possibly the world's least impressive web site": George Perfect

"Geoff Wood" wrote in message

"Arny Krueger" wrote in message

Agreed.

Been there, done that.

That's one reason why I've been telling people around here to set
peak levels at -1 dB for years.

It's also the reason why I've been telling people to do their
Rightmark tests with 18 & 20 KHz or 19 & 20 KHz CCIR-like IM test
tones, not the SMPTE-like easy stuff that it uses by default.

It can find a lot of converters with their hands in the cookie jar,
too.

LPs can sound a bit crappy with FS 19 + 19.x signals 'encoded' too !

I suppose. LP playback is usually far from linear in the top octave.

Just as well nothing remotely like that happens in real life - sort
of a 'natural de-emphasis' factor helps there.

As bad as real world signals get in the top octave can be found in my
www.pcabx.com keys jangling test clip:

http://64.41.69.21/technical/sample_...s-spectral.gif

http://64.41.69.21/technical/sample_...s-spectrum.gif

"Greg Taylor" wrote in message


...which is surprising, given how hot current popular music cd's are
mastered and replicated. One would think the better consumer players
manufactured now would handle the max levels possible.

They do. Even some of the better $40 DVD players do FS sine waves at all
frequencies well. This problem was more common about 5 years ago.

Laurence Payne wrote:

Just burnt a CD with 1k tone at 100%, then a little under. Computer
plays it OK. Sony CD player (into the same playback system) can't
cope with the 100% tone.

In article writes:

...which is surprising, given how hot current popular music cd's are
mastered and replicated. One would think the better consumer players
manufactured now would handle the max levels possible.

I guess they figure that if it's already distorted, a little more
distortion is probably better than none.

--
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 he double-m-eleven-double-zero at yahoo

Paul, I'm not sure of anything either (not having the player or the
waveform here in front of me) but I certainly have had similar
experiences. There used to be a highly-regarded brand of outboard DAC
called "Audio Alchemy" that was effusively recommended by Stereophile
and all the other magazines; after all, every audiophile knows that
separates always sound better than combined components, right?

Well, a few years ago I took one of those DACs and fed it a simple
midrange sine wave at about 1 dB below full scale (via its S-P/DIF
input), and the clipping was easily audible. If I recall correctly,
the audible distortion on sine waves began 1 - 2 dB lower than that.

You're certainly right that there can be clipping (sometimes quite
severe) in digital filters--but with such a simple waveform, I'd
consider the problem somewhat more likely to be on the analog side.
Anyway, most of this "free diagnosis over the Web" stuff is worth about
what you pay for it ...

--best regards

David Satz wrote:
Paul, I'm not sure of anything either (not having the player or the
waveform here in front of me) but I certainly have had similar
experiences. There used to be a highly-regarded brand of outboard DAC
called "Audio Alchemy" that was effusively recommended by Stereophile
and all the other magazines; after all, every audiophile knows that
separates always sound better than combined components, right?

Well, a few years ago I took one of those DACs and fed it a simple
midrange sine wave at about 1 dB below full scale (via its S-P/DIF
input), and the clipping was easily audible. If I recall correctly,
the audible distortion on sine waves began 1 - 2 dB lower than that.

Was this the DAC-In-The-Box or one of the more expensive models?
As I recall, those had a generic Crystal sigma-delta chip inside.
I thought they sounded decent, though.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Paul Stamler wrote:
"Scott Dorsey" wrote in message
...
In article ,
Laurence Payne wrote:
Just burnt a CD with 1k tone at 100%, then a little under.
Computer
plays it OK. Sony CD player (into the same playback system) can't
cope with the 100% tone. Sounds like vinyl mistracking.

Took it next door to the crappy set-top DVD player. TV cabinet
shakes a bit with the 1k tone, but it's clean.

I guess the culprit has to be the CD player.

Yes. As I said, this is sadly very common. The analogue section
clips
or it least gets very ugly at a lower level than the converter
itself.

Personally I believe there is no excuse for that sort of thing.

I'm not entirely sure the problem's in the analog sections. Standard
output
from a DAC at 0dBS is 2Vrms, which is about 5.64Vpp. Most of the
analog
stages run from +/- 15V rails, and such a stage should have no
problem with
5.64Vpp. I would guess, without going inside a player with a scope,
that
what's going on is digital clipping someplace, perhaps in the
filtering
stage.

Peace,
Paul


Actually that misunderstanding may be part of the problem.

If you start with a (say) 5vp-p square wave and then filter it to
REMOVE the harmonics, the resulting sine wave will be GREATER than 5
Vp-p.

This is true for analog as well as digital filters that someone
mentioed previously.

If you don't belive this, look up the first Fourier coefficent for a
square wave or try a test on the bench. The harmonics actually LOWER
the peak value of the square wave. Without the harmonics the sine wave
exceeds the original value of the square wave.

In a CD player, if the DAC can output 5vp-p, the reconstruction filter
removes harmoincs and the resulting sine wave applied ot the anlog
section can exceed 5vp-p. Some designers may not be aware of this and
design the analog section for 5vp-p.

The same problem occurs in radio modulation limiters. If you set a
hard limit to prevent overmodulation, and then filter to remove the
resulting harmonics (which you must do) the filtered signal can exceed
the original limited output and you can overmodulate.

Mark

Mark wrote:

If you start with a (say) 5vp-p square wave and then filter it to
REMOVE the harmonics, the resulting sine wave will be GREATER than 5
Vp-p.

This is true for analog as well as digital filters that someone
mentioed previously.

This is correct, but I'm not sure it's relevant. Yes, overshoot will
result in values that exceed the nominal 0 dBFS value. But that's just
something the converter designer should design around.

In a CD player, if the DAC can output 5vp-p, the reconstruction filter
removes harmoincs and the resulting sine wave applied ot the anlog
section can exceed 5vp-p. Some designers may not be aware of this and
design the analog section for 5vp-p.

It might also happen in a sigma-delta system if the digital filter
used in the oversampling process overflows.

Either way the result is the same, and the result is unacceptable.
I don't think it matters WHY this happens so much as THAT it happens
and that we, as users, shouldn't put up with it.

The same problem occurs in radio modulation limiters. If you set a
hard limit to prevent overmodulation, and then filter to remove the
resulting harmonics (which you must do) the filtered signal can exceed
the original limited output and you can overmodulate.

Yup. Or even limiting before the NAB emphasis filter for FM can cause this.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

On Tue, 11 Jan 2005 18:20:39 GMT, "Paul Stamler"
wrote:

I'm not entirely sure the problem's in the analog sections. Standard output
from a DAC at 0dBS is 2Vrms, which is about 5.64Vpp. Most of the analog
stages run from +/- 15V rails, and such a stage should have no problem with
5.64Vpp. I would guess, without going inside a player with a scope, that
what's going on is digital clipping someplace, perhaps in the filtering
stage.

Lots of interesting responses, too. Not yet mentioned are the big
out-of-band switching transients, and maybe overshoots, that the
summing junction I/V stage has to live with. Comparable in order
of nastiness magnitude with mic preamps, perhaps.

It's a mean world out there, little electrons. Be safe.

Chris Hornbeck
"Conscious that we must have sprung from somewhere, we are
lured to the riddle of our origins." -Malcolm W. Browne

"Laurence Payne"

No, I think you have a player problem.


** Practically zero chance that is the case.


So what is the problem?


** Bob's most likely problem is that his test disk has serious distortion.

The other possibility is that his pre-amp following the CD player cannot
accept 2 volt rms without distortion - many older domestic amps were only
designed for 200mV - 500 mV signals on the aux inputs.




.............. Phil

"Arny Krueger"


digital can never be perfect, long live the acetate ripping
needle!!!!!

Actually, the fault being described is probably on the analog side of the
converters.


** Only if you include the amplifier following the CD player.




.............. Phil