I am wanting to obtain the best possible quality when burning audio CD's.

I have been advised to burn at the slowest speed possible but under XP, I
cannot get Roxio or Nero to go lower than 16x on either of my burners.

Is this the software? if so is there an ideal burning software suite?

Many thanks


AlunP

"Alun P" alun.priddle@NOSPAMblueyonderDOTcoDOTuk wrote in
. uk:

I have been advised to burn at the slowest speed possible but under
XP, I cannot get Roxio or Nero to go lower than 16x on either of my
burners.

New burners can burn at full speed with no problem.

--
Lucas Tam )
Please delete "REMOVE" from the e-mail address when replying.
http://members.ebay.com/aboutme/coolspot18/

Alun P alun.priddle@nospamblueyonderdotcodotuk wrote:
I am wanting to obtain the best possible quality when burning audio CD's.

I have been advised to burn at the slowest speed possible but under XP, I
cannot get Roxio or Nero to go lower than 16x on either of my burners.

That advice is esssentially superstition, unless you have objective
evidence to show that the slowest burning speed on your setup yields
better performance than other speeds.

consider this:

http://www.cdrfaq.org/faq03.html#S3-31


--
-S
Your a boring little troll. How does it feel? Go blow your bad breath elsewhere.

Steven Sullivan wrote in
:

I am wanting to obtain the best possible quality when burning audio
CD's.

I have been advised to burn at the slowest speed possible but under
XP, I cannot get Roxio or Nero to go lower than 16x on either of my
burners.

That advice is esssentially superstition, unless you have objective
evidence to show that the slowest burning speed on your setup yields
better performance than other speeds.

I believe this "supertition" stemmed from the days when highspeed media was
not available. When 4x drives came out on the market, the vast majority of
CDRs were only 2X. Whenever you tried to burn a 2X rated CDR at 4X speeds,
it would yield unreliable discs.

Anyhow, it got stuck in people's mind that you need to burn at a lower
speed for reliable burns... but it's not true anymore. As long as you don't
burn faster than the media's rated speed (the speed of the media is listed
on the disc or Nero can detect the maximum speed for you), you'll be fine.

--
Lucas Tam )
Please delete "REMOVE" from the e-mail address when replying.
http://members.ebay.com/aboutme/coolspot18/

Bits are bits. If your burner reliably burns 1s and 0s at it's highest
speed (and it damn well should!) then burn away.

I defy _anyone_ who thinks they can hear the difference between a disc
burned at 1x and 50x to prove in a double-blind ABX text. It can't be
done. This is the beauty of digital.

On 2004-10-28 01:36:20 +1000, "Alun P"
alun.priddle@NOSPAMblueyonderDOTcoDOTuk said:

I am wanting to obtain the best possible quality when burning audio CD's.

I have been advised to burn at the slowest speed possible but under XP,
I cannot get Roxio or Nero to go lower than 16x on either of my burners.

Is this the software? if so is there an ideal burning software suite?

Many thanks


AlunP

Bits are bits. If your burner reliably burns 1s and 0s at it's highest
speed (and it damn well should!) then burn away.

I defy _anyone_ who thinks they can hear the difference between a disc
burned at 1x and 50x to prove in a double-blind ABX text. It can't be
done. This is the beauty of digital.

Largely true, but there are some interesting boundary situations:

- If a disc burned at 50x has a higher low-level-media bit-error rate
than one burned at a lower speed, then it may be more prone to
exhibit the symptoms of error-correction-algorithm failure (brief
muting, pops and ticks) during playback. This might be a
significant problem in some cases, especially as discs intended for
high-speed burning seem to tend to have thinner, somewhat-less-
reflective dye layers, and it may be harder for some older CD
players to get a strong-enough RF signal off of the disc to decode
properly.

- Similarly, a disc burned with a low level of contrast between pits
and lands may be harder for the CD player's focusing and tracking
servos to read accurately, leading to mistracking, a reluctance to
play, increased vulnerability to shock and vibration during
playback, and so forth.

- Burning at 50x results in a *fast* rotation of the disc... I've
heard burners which sounded a lot like small vacuum cleaners at
these speeds. If the disc's physical balance is off, or if the
spindle hole isn't properly centered or has some plastic residue on
its edge, the disc can wobble while being burned, and the writer
may not be able to track the pregroove accurately enough to give a
good burn.

- There have been some reports (largely anecdodal) which suggest that
there are some (poorly-designed) CD players which can allow digital
circuitry noise, and motor (tracking-servo) noise to feed back
through the power supply rails into the analog electronics. Discs
which are harder to track may result in different amounts of this
digital-and-motor noise ending up contaminating the analog audio
signal.

Digital's a neat thing, and the use of sophisticated error-correction
codes on CD give it a lot of resistance to reasonable numbers of bit
errors, RF signal dropouts, etc. due to burning problems. However,
one cannot make a silk purse out of a sow's ear... if the physical
quality of the burn is bad enough, either the tracking system or the
Reed-Solomon error correction coding system will eventually be
overwhelmed, and the CD will tend to mistrack or will suffer from
audible noise.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

"Dave Platt" wrote in message
...
Bits are bits. If your burner reliably burns 1s and 0s at it's highest
speed (and it damn well should!) then burn away.

I defy _anyone_ who thinks they can hear the difference between a disc
burned at 1x and 50x to prove in a double-blind ABX text. It can't be
done. This is the beauty of digital.

Largely true, but there are some interesting boundary situations:

- If a disc burned at 50x has a higher low-level-media bit-error rate
than one burned at a lower speed, then it may be more prone to
exhibit the symptoms of error-correction-algorithm failure (brief
muting, pops and ticks) during playback. This might be a
significant problem in some cases, especially as discs intended for
high-speed burning seem to tend to have thinner, somewhat-less-
reflective dye layers, and it may be harder for some older CD
players to get a strong-enough RF signal off of the disc to decode
properly.

- Similarly, a disc burned with a low level of contrast between pits
and lands may be harder for the CD player's focusing and tracking
servos to read accurately, leading to mistracking, a reluctance to
play, increased vulnerability to shock and vibration during
playback, and so forth.

- Burning at 50x results in a *fast* rotation of the disc... I've
heard burners which sounded a lot like small vacuum cleaners at
these speeds. If the disc's physical balance is off, or if the
spindle hole isn't properly centered or has some plastic residue on
its edge, the disc can wobble while being burned, and the writer
may not be able to track the pregroove accurately enough to give a
good burn.

- There have been some reports (largely anecdodal) which suggest that
there are some (poorly-designed) CD players which can allow digital
circuitry noise, and motor (tracking-servo) noise to feed back
through the power supply rails into the analog electronics. Discs
which are harder to track may result in different amounts of this
digital-and-motor noise ending up contaminating the analog audio
signal.

Digital's a neat thing, and the use of sophisticated error-correction
codes on CD give it a lot of resistance to reasonable numbers of bit
errors, RF signal dropouts, etc. due to burning problems. However,
one cannot make a silk purse out of a sow's ear... if the physical
quality of the burn is bad enough, either the tracking system or the
Reed-Solomon error correction coding system will eventually be
overwhelmed, and the CD will tend to mistrack or will suffer from
audible noise.

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

There are borderline issues which can cause varying results.
I use a Yamaha 16X burner. It produces oustanding discs - I have examined
the HF patterns produced on a variety of consumer decks using an
oscilloscope. The quality of the HF pattern is excellent for a burned disc.
Nevertheless, I have seen the occasional player (usually Philips based)
which has problems playing a 16X burned disc, but re-burning another at 12X
fixed the problem. Acts like a grating issue.

YMMV.

Mark Z.

Lucas Tam wrote in message ...
Steven Sullivan wrote in
:

I am wanting to obtain the best possible quality when burning audio
CD's.

I have been advised to burn at the slowest speed possible but under
XP, I cannot get Roxio or Nero to go lower than 16x on either of my
burners.

That advice is esssentially superstition, unless you have objective
evidence to show that the slowest burning speed on your setup yields
better performance than other speeds.

I believe this "supertition" stemmed from the days when highspeed media was
not available. When 4x drives came out on the market, the vast majority of
CDRs were only 2X. Whenever you tried to burn a 2X rated CDR at 4X speeds,
it would yield unreliable discs.

Anyhow, it got stuck in people's mind that you need to burn at a lower
speed for reliable burns... but it's not true anymore. As long as you don't
burn faster than the media's rated speed (the speed of the media is listed
on the disc or Nero can detect the maximum speed for you), you'll be fine.

For that matter, there was a time when a computer couldn't handle the
data fast enough, and so for the faster drive speeds the data wasn't
getting to it fast enough to keep the buffer full. You needed to
keep the record speed down in order to burn the CD properly, otherwise
it would be unreadable.

I just got my first CDRW drive back in December, in this 1GHz hand me
down. The books I had around were old enough where this was an issue,
and it made it sound like there could be serious problems if one
wasn't careful. But none of it was relevant once computers got
to a certain point.

Michael

(Michael Black) wrote in news:6447bcd3.0410271915.30d91cc8
@posting.google.com:

I just got my first CDRW drive back in December, in this 1GHz hand me
down. The books I had around were old enough where this was an issue,
and it made it sound like there could be serious problems if one
wasn't careful. But none of it was relevant once computers got
to a certain point.

These days almost all CDRW and DVDRW drives incorporate "burn proof"
technology. It basically stops the burning processes if the buffer drops
below an acceptable level.

When burn-proof came out... I thought it was a wonderfu invention. It sure
saved me a whole bunch of bad discs!

--
Lucas Tam )
Please delete "REMOVE" from the e-mail address when replying.
http://members.ebay.com/aboutme/coolspot18/

Many thanks for the overwhelming information, The reason for the post,
initially, was that my car cd player, an Alpine, does not like a lot of CD-R
discs (yes it is compatible!!) however, branded cd-r audio media burned at
slow speeds play perfectly.

The people at Alpine suggested birning at 1x or 2 x to overcome the problem,
as I said, neither of the two burning suites I have will go that low, maybe
its windows XP, I dont know?

AlunP

"Lucas Tam" wrote in message
Anyhow, it got stuck in people's mind that you need to burn at a lower
speed for reliable burns... but it's not true anymore. As long as you
don't
burn faster than the media's rated speed (the speed of the media is listed
on the disc or Nero can detect the maximum speed for you), you'll be fine.


I do CD duplication for a business , and I find the highest speeds may WRITE
OK, but are less likely to READ OK on a range of players. Yes, using the
appropriate high speed media.

I find 8x best on my bank of 14 12x Plexwriters, and 16x on Plextor
Premium (which can successfully write up to 52x) , over a variety of media
brands.

With the Plextor Premium PlexTools software you can do BLER tests on the
resultant discs, but you can't do a BLER test that necessarily reflects
*exactly* what will happen with that media in an audio CD player.

geoff

"Lucas Tam" wrote in message
...
(Michael Black) wrote in
news:6447bcd3.0410271915.30d91cc8
@posting.google.com:

I just got my first CDRW drive back in December, in this 1GHz hand me
down. The books I had around were old enough where this was an issue,
and it made it sound like there could be serious problems if one
wasn't careful. But none of it was relevant once computers got
to a certain point.

These days almost all CDRW and DVDRW drives incorporate "burn proof"
technology. It basically stops the burning processes if the buffer drops
below an acceptable level.

..... and does not restart 'perfectly', but well enough.

When burn-proof came out... I thought it was a wonderfu invention. It sure
saved me a whole bunch of bad discs!

If you were/are having constant buffer under-runs, then you have a problem
that should or could be addressed specifically. Treat the cause and not the
symptom...

geoff

"Kai Howells" wrote in message
u...
Bits are bits. If your burner reliably burns 1s and 0s at it's highest
speed (and it damn well should!) then burn away.

I defy _anyone_ who thinks they can hear the difference between a disc
burned at 1x and 50x to prove in a double-blind ABX text. It can't be
done. This is the beauty of digital.

Try taking a digital photo while a-creeping , and b- sprinting. See which
comes out sharper.

geoff

"Alun P" alun.priddle@NOSPAMblueyonderDOTcoDOTuk wrote in message
o.uk...
Many thanks for the overwhelming information, The reason for the post,
initially, was that my car cd player, an Alpine, does not like a lot of
CD-R discs (yes it is compatible!!) however, branded cd-r audio media
burned at slow speeds play perfectly.


many car players (and ghetto-blasters, and carousels) do not like CD-Rs.

The people at Alpine suggested birning at 1x or 2 x to overcome the
problem,

Their best possible advice in the circumstances, other than getting a
new-generation player than doesn't have problems with CD-Rs.

geoff

"Geoff Wood" -nospam wrote in
:

These days almost all CDRW and DVDRW drives incorporate "burn proof"
technology. It basically stops the burning processes if the buffer
drops below an acceptable level.

.... and does not restart 'perfectly', but well enough.

My drives always restarted perfectly.

When burn-proof came out... I thought it was a wonderfu invention. It
sure saved me a whole bunch of bad discs!

If you were/are having constant buffer under-runs, then you have a
problem that should or could be addressed specifically. Treat the
cause and not the symptom...

This was back when I had a Pentium III 450 - I was running other
applications while burning, so the buffer would drop due to disk activity.
Not much I could have done but maybe purchase a new hard drive.

--
Lucas Tam )
Please delete "REMOVE" from the e-mail address when replying.
http://members.ebay.com/aboutme/coolspot18/

Alun P alun.priddle@nospamblueyonderdotcodotuk wrote:
Many thanks for the overwhelming information, The reason for the post,
initially, was that my car cd player, an Alpine, does not like a lot of CD-R
discs (yes it is compatible!!) however, branded cd-r audio media burned at
slow speeds play perfectly.


Well, what you wrote was that you wanted the 'best possible quality'
CDRs.

I'd say your problem has to do with the quality of the *player*, not
the CDRs.



--
-S
Your a boring little troll. How does it feel? Go blow your bad breath elsewhere.

"Lucas Tam" wrote in message
.. .
"Geoff Wood" -nospam wrote in
:

These days almost all CDRW and DVDRW drives incorporate "burn proof"
technology. It basically stops the burning processes if the buffer
drops below an acceptable level.

.... and does not restart 'perfectly', but well enough.

My drives always restarted perfectly.

So you have an electron microscope. OK.

When burn-proof came out... I thought it was a wonderfu invention. It
sure saved me a whole bunch of bad discs!

If you were/are having constant buffer under-runs, then you have a
problem that should or could be addressed specifically. Treat the
cause and not the symptom...

This was back when I had a Pentium III 450 - I was running other
applications while burning, so the buffer would drop due to disk activity.
Not much I could have done but maybe purchase a new hard drive.

.... or not write while running other applications like I'm sure 90% of us
don't.


geoff.

"Kai Howells" wrote in message
u...
Bits are bits. If your burner reliably burns 1s and 0s at it's highest
speed (and it damn well should!) then burn away.

I defy _anyone_ who thinks they can hear the difference between a disc
burned at 1x and 50x to prove in a double-blind ABX text. It can't be
done. This is the beauty of digital.

What a load of crap. You don't need double blind tests with bits, just use
the appropriate measurement tools and see how many C1 errors there are.
You will find lots! Now of course you will say, who cares, the error
correction fixes that. But errors usually increase rather than decrease over
time, so having lower errors to start with is always a good thing IMO. And
caters better for marginal players.

Those of us who have actually looked at these things have found a quality
difference at varying speeds even with the same disk and drive.
Of course it is not necessary for you to care, some people do though

TonyP.

What I should have clarified was that as long as the bit-level error
rate is below the threshold that the Reed-Solomon codes can deal with,
then you will not hear any difference... Errors may increase over time,
although newer equipment may read it better, so errors may decrease,
but that is beside the point...

As long as the errors are corrected by the ECC coding, then you will
not hear anything different. There are those that will claim they can
hear the difference between different media, some brands are better
than others because they sound clearer or warmer - digital is just not
like that, either it is OK or it's not, a single-bit error in the audio
bitstream in the least-significant bit will probably not be audible,
but it's just as likely to be there as it is in the MSB, and you sure
as hell will hear that - it's not subtle =)

As I said, I defy _anyone_ who thinks they can hear the difference
between a disc burned at 1x and 50x to prove in a double-blind ABX
text. It can't be done. (provided that the disc is actually playable)

=)

Cheers,
Kai

On 2004-10-29 19:51:54 +1000, "TonyP" said:


"Kai Howells" wrote in message
u...
Bits are bits. If your burner reliably burns 1s and 0s at it's highest
speed (and it damn well should!) then burn away.

I defy _anyone_ who thinks they can hear the difference between a disc
burned at 1x and 50x to prove in a double-blind ABX text. It can't be
done. This is the beauty of digital.

What a load of crap. You don't need double blind tests with bits, just use
the appropriate measurement tools and see how many C1 errors there are.
You will find lots! Now of course you will say, who cares, the error
correction fixes that. But errors usually increase rather than decrease over
time, so having lower errors to start with is always a good thing IMO. And
caters better for marginal players.

Those of us who have actually looked at these things have found a quality
difference at varying speeds even with the same disk and drive.
Of course it is not necessary for you to care, some people do though

TonyP.

On Thu, 28 Oct 2004 06:02:56 GMT, "Alun P"
alun.priddle@NOSPAMblueyonderDOTcoDOTuk wrote:

Many thanks for the overwhelming information, The reason for the post,
initially, was that my car cd player, an Alpine, does not like a lot of CD-R
discs (yes it is compatible!!) however, branded cd-r audio media burned at
slow speeds play perfectly.

The people at Alpine suggested birning at 1x or 2 x to overcome the problem,
as I said, neither of the two burning suites I have will go that low, maybe
its windows XP, I dont know?

The old advice (backed up with experimental data) was that 2X was the
optimal burn speed. (Better than 1X, strangely).

You can't blame Windows XP. Today's media is optimised for faster
speeds. A good burner/software will test a disk when inserted and
only offer appropriate burn speeds. Too SLOW a speed can be bad.
Somehow, you're never told the very highest speeds are risky - I guess
the Marketing department would object :-)

Some musicians jealously hoard remaining stocks of "slow" media. I
find speeds between 10X and 20X reliable. Some players are fussy.

On Thu, 28 Oct 2004 19:55:25 +1300, "Geoff Wood"
-nospam wrote:

These days almost all CDRW and DVDRW drives incorporate "burn proof"
technology. It basically stops the burning processes if the buffer drops
below an acceptable level.

.... and does not restart 'perfectly', but well enough.

I've turned burnproof off for audio disks. I suspect that, on the
rare occasions that it's required, the burn continues but you get an
audio glitch. I'd rather the burn failed. I haven't got time to
listen right through checking every audio CD I send out :-)

On Thu, 28 Oct 2004 19:59:02 +1300, "Geoff Wood"
-nospam wrote:

The people at Alpine suggested birning at 1x or 2 x to overcome the
problem,

Their best possible advice in the circumstances, other than getting a
new-generation player than doesn't have problems with CD-Rs.

But, with modern media, advice impossible to follow, even if it would
help. Which it wouldn't, WITH MODERN MEDIA.

On Thu, 28 Oct 2004 16:15:44 GMT, Lucas Tam
wrote:

If you were/are having constant buffer under-runs, then you have a
problem that should or could be addressed specifically. Treat the
cause and not the symptom...

This was back when I had a Pentium III 450 - I was running other
applications while burning, so the buffer would drop due to disk activity.
Not much I could have done but maybe purchase a new hard drive.

You could have not run other applications while burning. Still good
advice, unless you're quite confident your computer is well
over-powered.

On Fri, 29 Oct 2004 23:43:08 +1000, Kai Howells
wrote:

As I said, I defy _anyone_ who thinks they can hear the difference
between a disc burned at 1x and 50x to prove in a double-blind ABX
text. It can't be done. (provided that the disc is actually playable)

Jitter?

Jitter *should* not be a problem in any well-designed digital circuit.
Most DACs (SACD aside, as that's a raw, 1-bit bitstream) process data
in words - which is a group of (usually) 16 bits.

The DAC will need to have a buffer on it, to hold the incoming bits
until it's been handed 16 of them, and then it should look at the value
it holds and output the appropriate analogue voltage. The DAC should
run off it's own clock, or have it's clock synchronised with the rest
of the system, so it should be reading entire words on a regular basis.
Now, all these assumptions I've been making may not be valid with SACD,
as that's a whole different kettle of fish, but then you can't (yet)
burn SACD discs so the point it moot =)

For this same reason, I can't see how different optical cables can
sound different - there are many who will swear black & blue that co-ax
is the supreme digital interconnect, whereas glass optical cables sound
too bright and plastic sounds too dull.
Effects, such as jitter, and single-bit errors *CAN'T* have this kind
of effect on the sound - it is digital, you need all kinds of DSP
equipment to change the tone of a digital signal, and this equipment
has to inherently understand the meaning of the bits that it's
processing. a "dumb" wire just can't do this.

If anyone can explain how different digital cables can make different
sounds, then I'd love to hear it =)

Cheers,
Kai

On 2004-10-30 06:53:42 +1000, Laurence Payne
said:

On Fri, 29 Oct 2004 23:43:08 +1000, Kai Howells
wrote:

As I said, I defy _anyone_ who thinks they can hear the difference
between a disc burned at 1x and 50x to prove in a double-blind ABX
text. It can't be done. (provided that the disc is actually playable)

Jitter?

In article ,
Kai Howells wrote:

Jitter *should* not be a problem in any well-designed digital circuit.

Agreed.

Most DACs (SACD aside, as that's a raw, 1-bit bitstream) process data
in words - which is a group of (usually) 16 bits.

The DAC will need to have a buffer on it, to hold the incoming bits
until it's been handed 16 of them, and then it should look at the value
it holds and output the appropriate analogue voltage. The DAC should
run off it's own clock, or have it's clock synchronised with the rest
of the system, so it should be reading entire words on a regular basis.

Well, it's the clock synchronization which can be the tricky part, and
from what I've heard there are a fair number of DACs which get it wrong.

It's usually not a problem for single-box CD players, since in these
players the DAC-chip is being clocked by the player's rather stable,
low-jitter quartz crystal oscillator, and the timing of the
conversions from analog to digital don't have significant timing jitter.

The task is harder for two-box transport/DAC systems. The data read
by the transport is being clocked out of the transport box (onto the
optical or coax cable) with stable, quartz-driving timing in most
cases. However, the DAC-box doesn't have direct access to this clock
signal. The DAC-box must "recover" (reconstruct) the clock signal
from the edge transitions embedded in the S/PDIF or other
digital-audio data stream. This is usually done via a voltage-
controlled-oscillator / phase-locked-loop arrangement.

Unfortunately, the VCO/PLL arrangement is vulnerable to some amount of
wandering... the clock signal it outputs contains significantly more
timing jitter than the original quartz-oscillator clock in the DAC.
This jitter can be made worse if the digital-audio signal is
"contaminated" by the effects of signal reflections (due e.g. to
improper termination of the 75-ohm line), or if there's electrical
noise inside the DAC-box due to poor power supply bypassing or
PC-board layout, or etc.

A poor DAC-box (and there have apparently been quite a few) may take
the reconstructed signal right out of the S/PDIF receiver chip's
VCO/PLL, and feed it right into the DAC chip. This can result in
significant jitter in the timing conversions. Jitter in time has
results similar to error in amplitude, and it's been shown that it
_can_ be audible if the jitter is severe enough (audibility seems to
depend both on jitter amplitude, and on the frequency spectrum of the
jitter itself).

A really good DAC-box will run the reconstructed clock signal through
several levels of cleanup and stabilization logic (filtering), and
will often buffer multiple audio samples through a RAM FIFO of some
sort.

Whether the jitter in consumer-grade DAC-boxes (or A/V receivers) is
severe enough to be audible under typical listening conditions is
another matter. I've heard plenty of anecdotal reports that certain
DAC-boxes are widely believed to "reveal" differences between
different CD transports, and they're often lauded for this. In my
opinion, DAC-boxes should be *criticized* for making transport
differences audible, for just the reasons you allude to... a properly
designed DAC-box should simply not *care* about small timing
variations in the incoming signal, as these differences are irrelevant
to proper conversion of the digital signal to analog.

For this same reason, I can't see how different optical cables can
sound different - there are many who will swear black & blue that co-ax
is the supreme digital interconnect, whereas glass optical cables sound
too bright and plastic sounds too dull.

Yeah, I agree, that's excessive. In most cases I view this as being
due to human suggestibility.

Effects, such as jitter, and single-bit errors *CAN'T* have this kind
of effect on the sound - it is digital, you need all kinds of DSP
equipment to change the tone of a digital signal, and this equipment
has to inherently understand the meaning of the bits that it's
processing. a "dumb" wire just can't do this.

Not strictly true, I think. Any problems in the conversion equipment
which alter the _timing_ of the sample conversions can have an effect
on the reconstructed waveform which is essentially equivalent to an
error in the _amplitude_ of the sample conversions. It's distortion,
and (if severe enough) it can be audible.

If anyone can explain how different digital cables can make different
sounds, then I'd love to hear it =)

Use a DAC-box which has poor clock-reconstruction circuitry. Compare
a high-quality 75-ohm digital-audio interconnect, with another cable
whose impedance is poorly controlled and is significantly different
than 75 ohms. You may find that you can hear a difference (even if
the DAC-box is not reporting gross errors with the bitstream), and if
you put an RF spectrum analyzer or jitter-analyzer on the clock signal
being fed to the DAC-chip you may be able to measure the effects of
jitter caused by signal reflections on the digital interconnect cable.

Remember, only the data's digital. The signal flowing over the S/PDIF
cable isn't digital - it's bandlimited RF, which just plays digital on
television (i.e. its edge transitions are far from instantaneous, and
it may have detectable signal reflections due to impedance variations
and mistermination).

--
Dave Platt AE6EO
Hosting the Jade Warrior home page: http://www.radagast.org/jade-warrior
I do _not_ wish to receive unsolicited commercial email, and I will
boycott any company which has the gall to send me such ads!

On Sat, 30 Oct 2004 08:30:34 +1000, Kai Howells
wrote:

Jitter *should* not be a problem in any well-designed digital circuit.
Most DACs (SACD aside, as that's a raw, 1-bit bitstream) process data
in words - which is a group of (usually) 16 bits.

Unfortunately, some "audiophiles" think it's clever to separate CD
player and DAC. An integrated system, I believe, has fewer problems.

Thankyou David, that's a fantastic writeup.
I've saved it for re-reading - there's a lot of information there.

Cheers,
Kai

On 2004-10-30 09:22:20 +1000, (Dave Platt) said:

In article ,
Kai Howells wrote:

Jitter *should* not be a problem in any well-designed digital circuit.

Agreed.

Most DACs (SACD aside, as that's a raw, 1-bit bitstream) process data
in words - which is a group of (usually) 16 bits.

The DAC will need to have a buffer on it, to hold the incoming bits
until it's been handed 16 of them, and then it should look at the value
it holds and output the appropriate analogue voltage. The DAC should
run off it's own clock, or have it's clock synchronised with the rest
of the system, so it should be reading entire words on a regular basis.

Well, it's the clock synchronization which can be the tricky part, and
from what I've heard there are a fair number of DACs which get it wrong.

It's usually not a problem for single-box CD players, since in these
players the DAC-chip is being clocked by the player's rather stable,
low-jitter quartz crystal oscillator, and the timing of the
conversions from analog to digital don't have significant timing jitter.

The task is harder for two-box transport/DAC systems. The data read
by the transport is being clocked out of the transport box (onto the
optical or coax cable) with stable, quartz-driving timing in most
cases. However, the DAC-box doesn't have direct access to this clock
signal. The DAC-box must "recover" (reconstruct) the clock signal
from the edge transitions embedded in the S/PDIF or other
digital-audio data stream. This is usually done via a voltage-
controlled-oscillator / phase-locked-loop arrangement.

Unfortunately, the VCO/PLL arrangement is vulnerable to some amount of
wandering... the clock signal it outputs contains significantly more
timing jitter than the original quartz-oscillator clock in the DAC.
This jitter can be made worse if the digital-audio signal is
"contaminated" by the effects of signal reflections (due e.g. to
improper termination of the 75-ohm line), or if there's electrical
noise inside the DAC-box due to poor power supply bypassing or
PC-board layout, or etc.

A poor DAC-box (and there have apparently been quite a few) may take
the reconstructed signal right out of the S/PDIF receiver chip's
VCO/PLL, and feed it right into the DAC chip. This can result in
significant jitter in the timing conversions. Jitter in time has
results similar to error in amplitude, and it's been shown that it
_can_ be audible if the jitter is severe enough (audibility seems to
depend both on jitter amplitude, and on the frequency spectrum of the
jitter itself).

A really good DAC-box will run the reconstructed clock signal through
several levels of cleanup and stabilization logic (filtering), and
will often buffer multiple audio samples through a RAM FIFO of some
sort.

Whether the jitter in consumer-grade DAC-boxes (or A/V receivers) is
severe enough to be audible under typical listening conditions is
another matter. I've heard plenty of anecdotal reports that certain
DAC-boxes are widely believed to "reveal" differences between
different CD transports, and they're often lauded for this. In my
opinion, DAC-boxes should be *criticized* for making transport
differences audible, for just the reasons you allude to... a properly
designed DAC-box should simply not *care* about small timing
variations in the incoming signal, as these differences are irrelevant
to proper conversion of the digital signal to analog.

For this same reason, I can't see how different optical cables can
sound different - there are many who will swear black & blue that co-ax
is the supreme digital interconnect, whereas glass optical cables sound
too bright and plastic sounds too dull.

Yeah, I agree, that's excessive. In most cases I view this as being
due to human suggestibility.

Effects, such as jitter, and single-bit errors *CAN'T* have this kind
of effect on the sound - it is digital, you need all kinds of DSP
equipment to change the tone of a digital signal, and this equipment
has to inherently understand the meaning of the bits that it's
processing. a "dumb" wire just can't do this.

Not strictly true, I think. Any problems in the conversion equipment
which alter the _timing_ of the sample conversions can have an effect
on the reconstructed waveform which is essentially equivalent to an
error in the _amplitude_ of the sample conversions. It's distortion,
and (if severe enough) it can be audible.

If anyone can explain how different digital cables can make different
sounds, then I'd love to hear it =)

Use a DAC-box which has poor clock-reconstruction circuitry. Compare
a high-quality 75-ohm digital-audio interconnect, with another cable
whose impedance is poorly controlled and is significantly different
than 75 ohms. You may find that you can hear a difference (even if
the DAC-box is not reporting gross errors with the bitstream), and if
you put an RF spectrum analyzer or jitter-analyzer on the clock signal
being fed to the DAC-chip you may be able to measure the effects of
jitter caused by signal reflections on the digital interconnect cable.

Remember, only the data's digital. The signal flowing over the S/PDIF
cable isn't digital - it's bandlimited RF, which just plays digital on
television (i.e. its edge transitions are far from instantaneous, and
it may have detectable signal reflections due to impedance variations
and mistermination).

On Sat, 30 Oct 2004 18:23:55 +1000, Kai Howells
wrote:

Thankyou David, that's a fantastic writeup.
I've saved it for re-reading - there's a lot of information there.

Indeed there is - all of which can be negated by using an asynchronous
re-sampling DAC such as the 'gold standard' Benchmark DAC-1. This
device totally reclocks the incoming data, thereby *removing* (not
attenuating) any incoming datastream jitter.

Cheers,
Kai

On 2004-10-30 09:22:20 +1000, (Dave Platt) said:

In article ,
Kai Howells wrote:

Jitter *should* not be a problem in any well-designed digital circuit.

Agreed.

Most DACs (SACD aside, as that's a raw, 1-bit bitstream) process data
in words - which is a group of (usually) 16 bits.

The DAC will need to have a buffer on it, to hold the incoming bits
until it's been handed 16 of them, and then it should look at the value
it holds and output the appropriate analogue voltage. The DAC should
run off it's own clock, or have it's clock synchronised with the rest
of the system, so it should be reading entire words on a regular basis.

Well, it's the clock synchronization which can be the tricky part, and
from what I've heard there are a fair number of DACs which get it wrong.

It's usually not a problem for single-box CD players, since in these
players the DAC-chip is being clocked by the player's rather stable,
low-jitter quartz crystal oscillator, and the timing of the
conversions from analog to digital don't have significant timing jitter.

The task is harder for two-box transport/DAC systems. The data read
by the transport is being clocked out of the transport box (onto the
optical or coax cable) with stable, quartz-driving timing in most
cases. However, the DAC-box doesn't have direct access to this clock
signal. The DAC-box must "recover" (reconstruct) the clock signal
from the edge transitions embedded in the S/PDIF or other
digital-audio data stream. This is usually done via a voltage-
controlled-oscillator / phase-locked-loop arrangement.

Unfortunately, the VCO/PLL arrangement is vulnerable to some amount of
wandering... the clock signal it outputs contains significantly more
timing jitter than the original quartz-oscillator clock in the DAC.
This jitter can be made worse if the digital-audio signal is
"contaminated" by the effects of signal reflections (due e.g. to
improper termination of the 75-ohm line), or if there's electrical
noise inside the DAC-box due to poor power supply bypassing or
PC-board layout, or etc.

A poor DAC-box (and there have apparently been quite a few) may take
the reconstructed signal right out of the S/PDIF receiver chip's
VCO/PLL, and feed it right into the DAC chip. This can result in
significant jitter in the timing conversions. Jitter in time has
results similar to error in amplitude, and it's been shown that it
_can_ be audible if the jitter is severe enough (audibility seems to
depend both on jitter amplitude, and on the frequency spectrum of the
jitter itself).

A really good DAC-box will run the reconstructed clock signal through
several levels of cleanup and stabilization logic (filtering), and
will often buffer multiple audio samples through a RAM FIFO of some
sort.

Whether the jitter in consumer-grade DAC-boxes (or A/V receivers) is
severe enough to be audible under typical listening conditions is
another matter. I've heard plenty of anecdotal reports that certain
DAC-boxes are widely believed to "reveal" differences between
different CD transports, and they're often lauded for this. In my
opinion, DAC-boxes should be *criticized* for making transport
differences audible, for just the reasons you allude to... a properly
designed DAC-box should simply not *care* about small timing
variations in the incoming signal, as these differences are irrelevant
to proper conversion of the digital signal to analog.

For this same reason, I can't see how different optical cables can
sound different - there are many who will swear black & blue that co-ax
is the supreme digital interconnect, whereas glass optical cables sound
too bright and plastic sounds too dull.

Yeah, I agree, that's excessive. In most cases I view this as being
due to human suggestibility.

Effects, such as jitter, and single-bit errors *CAN'T* have this kind
of effect on the sound - it is digital, you need all kinds of DSP
equipment to change the tone of a digital signal, and this equipment
has to inherently understand the meaning of the bits that it's
processing. a "dumb" wire just can't do this.

Not strictly true, I think. Any problems in the conversion equipment
which alter the _timing_ of the sample conversions can have an effect
on the reconstructed waveform which is essentially equivalent to an
error in the _amplitude_ of the sample conversions. It's distortion,
and (if severe enough) it can be audible.

If anyone can explain how different digital cables can make different
sounds, then I'd love to hear it =)

Use a DAC-box which has poor clock-reconstruction circuitry. Compare
a high-quality 75-ohm digital-audio interconnect, with another cable
whose impedance is poorly controlled and is significantly different
than 75 ohms. You may find that you can hear a difference (even if
the DAC-box is not reporting gross errors with the bitstream), and if
you put an RF spectrum analyzer or jitter-analyzer on the clock signal
being fed to the DAC-chip you may be able to measure the effects of
jitter caused by signal reflections on the digital interconnect cable.

Remember, only the data's digital. The signal flowing over the S/PDIF
cable isn't digital - it's bandlimited RF, which just plays digital on
television (i.e. its edge transitions are far from instantaneous, and
it may have detectable signal reflections due to impedance variations
and mistermination).


--

Stewart Pinkerton | Music is Art - Audio is Engineering

On Sat, 30 Oct 2004 17:22:12 +0000 (UTC), Stewart Pinkerton
wrote:


Thankyou David, that's a fantastic writeup.
I've saved it for re-reading - there's a lot of information there.

Indeed there is - all of which can be negated by using an asynchronous
re-sampling DAC such as the 'gold standard' Benchmark DAC-1. This
device totally reclocks the incoming data, thereby *removing* (not
attenuating) any incoming datastream jitter.

So it creates a problem by not being fully integrated, then cleverly
solves it? :-)

"Laurence Payne" wrote in message
...
On Sat, 30 Oct 2004 17:22:12 +0000 (UTC), Stewart Pinkerton
wrote:


Thankyou David, that's a fantastic writeup.
I've saved it for re-reading - there's a lot of information there.

Indeed there is - all of which can be negated by using an asynchronous
re-sampling DAC such as the 'gold standard' Benchmark DAC-1. This
device totally reclocks the incoming data, thereby *removing* (not
attenuating) any incoming datastream jitter.

So it creates a problem by not being fully integrated, then cleverly
solves it? :-)

Nothing like "golden ears" audio types who don't understand the
fundamentals of how digital logic works (whether digital audio
or just plain binary gates, latches, etc.)

On Sat, 30 Oct 2004 19:21:11 +0100, Laurence Payne
wrote:

On Sat, 30 Oct 2004 17:22:12 +0000 (UTC), Stewart Pinkerton
wrote:


Thankyou David, that's a fantastic writeup.
I've saved it for re-reading - there's a lot of information there.

Indeed there is - all of which can be negated by using an asynchronous
re-sampling DAC such as the 'gold standard' Benchmark DAC-1. This
device totally reclocks the incoming data, thereby *removing* (not
attenuating) any incoming datastream jitter.

So it creates a problem by not being fully integrated, then cleverly
solves it? :-)

It creates no problem in and of itself, it simply offers a solution to
a problem created by others, viz the existence of multiple digital
signal sources, all of which may be fed through this DAC.
--

Stewart Pinkerton | Music is Art - Audio is Engineering

"Kai Howells" wrote in message
u...
As I said, I defy _anyone_ who thinks they can hear the difference
between a disc burned at 1x and 50x to prove in a double-blind ABX
text. It can't be done.
(provided that the disc is actually playable)

Which is kind of the point in keeping errors as low as possible. However I
do have disks that are "playable", and the errors are quite audible in many
CD players.

TonyP.