[Mark D. Zacharias]

Over the years I've recorded many albums to cd, using DcArt and CoolEdit
2000 to process them.

Lately with lots of HD space available I've taken to simply recording LP
wave files and processing only major pops, and normalizing the overall
level, storing the files for possible later processing.

I was considering the issue of subsonic filtering of these wave files and
the following 2 questions came to mind:


1. Does filtering in the digital domain introduce phase shift the same as
analog filtering? ( I would guess it does)

2. What would you guys recommend as a cut-off frequency and slope for LP
subsonic purposes? 20 Hz, 18 dB/octave, 15 hz at 6 dB/octave, etc?

I don't have any ready info on the tonearm resonce of my setup, so this
could only be general advice, I realize. Resonance does not seem to be a
particular issue, though, the arm seems quite stable in operation.


Thanks all,

Mark Z.

[[email protected]]

Mark D. Zacharias wrote:
1. Does filtering in the digital domain introduce phase shift the same as
analog filtering? ( I would guess it does)

If it has the same pole-zero response, yes it does. However,
unlike analog-based filters, one can build digital filters that
have a number of unique attributes. One example is a digital
filter can be built that are "non-causal," that is, in effect, that
the filter can act on something before it happens, i.e., it can
"look forward" in time. Now, there's no miracle of time travel
involved, simply the introduction of delays.

The advantage of a digital filter is that you CAN build a filter
that has exactly the same frequency and phase characteristics
of an equivalent analog filter. And an added advantage is that
you can build filters that don't, if needed.

2. What would you guys recommend as a cut-off frequency
and slope for LP subsonic purposes? 20 Hz, 18 dB/octave,
15 hz at 6 dB/octave, etc?

Don't necessarily limit yourself to typical analog properties of
filters. Consider, for example, a filter that is flat to within +-0.1
dB to 20 Hz and is down 100 dB at 18 Hz, for example.

I don't have any ready info on the tonearm resonce of my
setup, so this could only be general advice, I realize.
Resonance does not seem to be a particular issue,
though, the arm seems quite stable in operation.

Resonance IS a very important issue, even if the arm seems
stable. Resonance determines the ultimate low-frequency
cutoff of the playback system. It determines the low-end
frequency- and phase-response of the mechanical portion
of the playback system.

[Mark D. Zacharias]

wrote in message
ps.com...

Mark D. Zacharias wrote:
1. Does filtering in the digital domain introduce phase shift the same as
analog filtering? ( I would guess it does)

If it has the same pole-zero response, yes it does. However,
unlike analog-based filters, one can build digital filters that
have a number of unique attributes. One example is a digital
filter can be built that are "non-causal," that is, in effect, that
the filter can act on something before it happens, i.e., it can
"look forward" in time. Now, there's no miracle of time travel
involved, simply the introduction of delays.

The advantage of a digital filter is that you CAN build a filter
that has exactly the same frequency and phase characteristics
of an equivalent analog filter. And an added advantage is that
you can build filters that don't, if needed.

2. What would you guys recommend as a cut-off frequency
and slope for LP subsonic purposes? 20 Hz, 18 dB/octave,
15 hz at 6 dB/octave, etc?

Don't necessarily limit yourself to typical analog properties of
filters. Consider, for example, a filter that is flat to within +-0.1
dB to 20 Hz and is down 100 dB at 18 Hz, for example.

I don't have any ready info on the tonearm resonce of my
setup, so this could only be general advice, I realize.
Resonance does not seem to be a particular issue,
though, the arm seems quite stable in operation.

Resonance IS a very important issue, even if the arm seems
stable. Resonance determines the ultimate low-frequency
cutoff of the playback system. It determines the low-end
frequency- and phase-response of the mechanical portion
of the playback system.


Appreciate the quick reply.

How would one set about measuring tonearm resonance? (I'm suspecting one
could feed a signal to the tonearm / cartridge, and find the point of
maximum deflection in the under 20 hz region.)

Or is there a more traditional test record / chart recorder technique? (I
have neither at my disposal)


Mark Z.

[Arny Krueger]

"Mark D. Zacharias" wrote in
message
et
Over the years I've recorded many albums to cd, using
DcArt and CoolEdit 2000 to process them.

Lately with lots of HD space available I've taken to
simply recording LP wave files and processing only major
pops, and normalizing the overall level, storing the
files for possible later processing.
I was considering the issue of subsonic filtering of
these wave files and the following 2 questions came to
mind:

1. Does filtering in the digital domain introduce phase
shift the same as analog filtering? ( I would guess it
does)

Since you said "Cool Edit" you have a wide choice of filters that have
various phase shift characteristics. The CE FFT filter is relatively free of
phase shift, while the scientific filters have phase shift characteristics
that are similar to traditional analog filters.

Remember that as a rule, phase shift applied equally to both channels has
far less audible effect than the corresponding amplitude response changes.

2. What would you guys recommend as a cut-off frequency
and slope for LP subsonic purposes? 20 Hz, 18 dB/octave,
15 hz at 6 dB/octave, etc?

That is very much dependent on the application. As a rule, you want to set
this corner frequency as high as possible, and the slope as steep as
possible, without audibly impacting the music. If your recordings are
typical popular music, there are probably no musical notes with fundamentals
below either 32 or 42 Hz, depending on the type of electric bass that was
used.

The LPs themselves might have had their musical content agressively high
pass filtered during the production or cutting process. For example, Motown
high pass filtered *everything* they produced in Detroit at about 80 Hz with
something like an 18 dB/octave slope.

People like to talk about the phase shift that is inherent in traditional
analog filters with steep slopes, but FFT-based filters don't have the same
problems.

IME the most audible effects of filters are in the transition band, IOW the
band where the filter's gain is relatively high, and changing rapidly with
frequency. One benefit of steep filters is that they minimize the width of
the transition bands.

I don't have any ready info on the tonearm resonce of my
setup, so this could only be general advice, I realize.

The tone arm resonance is probably in the 5-25 Hz range, with light damping.
That means that the tone arm resonance is affecting (typically boosting and
muddying) frequencies up as high as 120 Hz.

Resonance does not seem to be a particular issue, though,
the arm seems quite stable in operation.

Since you said "Cool Edit", you should familiarize yourself with its
powerful built-in frequency analysis tool. Set the FFT size to max, and
compare what you're getting off the LPs to what you know about the low
frequency content of the instruments that were actually used.

[Arny Krueger]

"Mark D. Zacharias" wrote in
message

How would one set about measuring tonearm resonance? (I'm
suspecting one could feed a signal to the tonearm /
cartridge, and find the point of maximum deflection in
the under 20 hz region.)

Tone arm resonance often shows up, totally or in part, in Cool Edit's
frequency analysis of just about any old transcription unless it has been
high-pass filtered along the way.

Remember that its a rare piece of popular music that has any musical
fundamentals below 32 Hz. It was pretty common in the days of, to high-pass
LPs during production, to ensure trackability on crappy equipment.

By crappy equipment I mean a plastic or tin tone arm and platter, 2-pole
motor, cloth-covered cardboard chassis, and crystal cartridge. IOW half or
more of all LP playback equipment in consumer use in the days of vinyl.

I believe that Scott has recently said that the policy of rolling-off the
bass is being relaxed with some modern recordings, because very little
really crappy equipment is still in use.

The usual LP warps and non-flatness of the LP's substrate provides a source
of excitation for LF resonances. So, if there's a lot of stuff below 32 Hz,
and there is a discernable rise or peak, you are probably looking at
artifacts of the tone arm's resonance.

The *standard* way to find tone arm resonance was to have a phono preamp
that is flat below 50 Hz down to say 1 Hz, and use a test record with LF
test tones. Most RIAA preamps have some built-in high-pass filtering which
may mask this. The IEC variation to RIAA de-emphasis added a roll-off below
30 Hz or so, if memory serves.

The response peak around the tone arm's fundamental resonance is usually
non-trivial, to say the least. ;-)

[Serge Auckland]

Arny Krueger wrote:
"Mark D. Zacharias" wrote in
message

How would one set about measuring tonearm resonance? (I'm
suspecting one could feed a signal to the tonearm /
cartridge, and find the point of maximum deflection in
the under 20 hz region.)

Tone arm resonance often shows up, totally or in part, in Cool Edit's
frequency analysis of just about any old transcription unless it has been
high-pass filtered along the way.

Remember that its a rare piece of popular music that has any musical
fundamentals below 32 Hz. It was pretty common in the days of, to high-pass
LPs during production, to ensure trackability on crappy equipment.

By crappy equipment I mean a plastic or tin tone arm and platter, 2-pole
motor, cloth-covered cardboard chassis, and crystal cartridge. IOW half or
more of all LP playback equipment in consumer use in the days of vinyl.

I believe that Scott has recently said that the policy of rolling-off the
bass is being relaxed with some modern recordings, because very little
really crappy equipment is still in use.

The usual LP warps and non-flatness of the LP's substrate provides a source
of excitation for LF resonances. So, if there's a lot of stuff below 32 Hz,
and there is a discernable rise or peak, you are probably looking at
artifacts of the tone arm's resonance.

The *standard* way to find tone arm resonance was to have a phono preamp
that is flat below 50 Hz down to say 1 Hz, and use a test record with LF
test tones. Most RIAA preamps have some built-in high-pass filtering which
may mask this. The IEC variation to RIAA de-emphasis added a roll-off below
30 Hz or so, if memory serves.

The response peak around the tone arm's fundamental resonance is usually
non-trivial, to say the least. ;-)


The way I have always tested for tonearm/cartridge resonance is to use a
test record which has a mid-frequency tone modulated with a very low
frequency tone, starting at around 5 Hz and going up to about 15 Hz. If
you listen to the tone and view the cartridge, you will see the
frequency at which the arm/cartridge visibly resonates. The reproduction
of the pure tone also audibly warbles at the LF resonant frequency.

The HFN test record has such bands, as do several others.

As the test is essentially visual (although the audible warble also
helps) it is independent of the RIAA preamp.

S.

[[email protected]]

Phase shifts in analog and most digital filters arise because of the
caused signal delays vary with frequency and the lengths of these
delays change with the amount of boost or cut. In the digital domain
you can have Symmetric FIR filters, which are phase linear, so the
delay caused by processing is constant across the whole spectrum. Both
have their drawbacks: Whenever you equalize you alter the signal in
time and freq domain. In the analog style eq these delays are different
at different frequencies. This changes the "sound" more in stead of the
timing and also has an effect on depth and imaging. The linear phase
type will equally precede and follow the signal so that it can, when
used with high q's, actually cause audible pre-echo effect on
transients. Linear phase sounds more sweet but is not suitable for
every kind of music. Anyway,.. there is a use for both,.. I'll guess
you have to listen. And then there are a lot of VERY BAD filter
plugins, esp HPfilters, watch out! Listen, change, listen, change,
listen, ...break..., listen, change, ...break..., listen, change,
listen, back, listen again, and decide, then, ...break... confirm your
decision and do it!

Henk

[Mark D. Zacharias]

Serge Auckland wrote:
Arny Krueger wrote:
"Mark D. Zacharias" wrote in
message

How would one set about measuring tonearm resonance? (I'm
suspecting one could feed a signal to the tonearm /
cartridge, and find the point of maximum deflection in
the under 20 hz region.)

Tone arm resonance often shows up, totally or in part, in Cool Edit's
frequency analysis of just about any old transcription unless it has
been high-pass filtered along the way.

Remember that its a rare piece of popular music that has any musical
fundamentals below 32 Hz. It was pretty common in the days of, to
high-pass LPs during production, to ensure trackability on crappy
equipment. By crappy equipment I mean a plastic or tin tone arm and
platter,
2-pole motor, cloth-covered cardboard chassis, and crystal
cartridge. IOW half or more of all LP playback equipment in consumer
use in the days of vinyl. I believe that Scott has recently said that the
policy of
rolling-off the bass is being relaxed with some modern recordings,
because very little really crappy equipment is still in use.

The usual LP warps and non-flatness of the LP's substrate provides a
source of excitation for LF resonances. So, if there's a lot of
stuff below 32 Hz, and there is a discernable rise or peak, you are
probably looking at artifacts of the tone arm's resonance.

The *standard* way to find tone arm resonance was to have a phono
preamp that is flat below 50 Hz down to say 1 Hz, and use a test
record with LF test tones. Most RIAA preamps have some built-in
high-pass filtering which may mask this. The IEC variation to RIAA
de-emphasis added a roll-off below 30 Hz or so, if memory serves.

The response peak around the tone arm's fundamental resonance is
usually non-trivial, to say the least. ;-)


The way I have always tested for tonearm/cartridge resonance is to
use a test record which has a mid-frequency tone modulated with a
very low frequency tone, starting at around 5 Hz and going up to
about 15 Hz. If you listen to the tone and view the cartridge, you
will see the frequency at which the arm/cartridge visibly resonates.
The reproduction of the pure tone also audibly warbles at the LF
resonant frequency.
The HFN test record has such bands, as do several others.

As the test is essentially visual (although the audible warble also
helps) it is independent of the RIAA preamp.

S.

This sounds VERY cool. Any idea where one might pick up a copy of this test
record these days?

I have a few old test records around by Shure, JVC and Ortofon, but the docs
are lacking as I recall. The Ortofon LP label is in German, and IIRC the JVC
test record label is in Japanese.

Maybe I can find such a test on one of them.

Thanks to all for the replies.


Mark Z.

[Serge Auckland]

Mark D. Zacharias wrote:
Serge Auckland wrote:
Arny Krueger wrote:
"Mark D. Zacharias" wrote in
message

How would one set about measuring tonearm resonance? (I'm
suspecting one could feed a signal to the tonearm /
cartridge, and find the point of maximum deflection in
the under 20 hz region.)
Tone arm resonance often shows up, totally or in part, in Cool Edit's
frequency analysis of just about any old transcription unless it has
been high-pass filtered along the way.

Remember that its a rare piece of popular music that has any musical
fundamentals below 32 Hz. It was pretty common in the days of, to
high-pass LPs during production, to ensure trackability on crappy
equipment. By crappy equipment I mean a plastic or tin tone arm and
platter,
2-pole motor, cloth-covered cardboard chassis, and crystal
cartridge. IOW half or more of all LP playback equipment in consumer
use in the days of vinyl. I believe that Scott has recently said that the
policy of
rolling-off the bass is being relaxed with some modern recordings,
because very little really crappy equipment is still in use.

The usual LP warps and non-flatness of the LP's substrate provides a
source of excitation for LF resonances. So, if there's a lot of
stuff below 32 Hz, and there is a discernable rise or peak, you are
probably looking at artifacts of the tone arm's resonance.

The *standard* way to find tone arm resonance was to have a phono
preamp that is flat below 50 Hz down to say 1 Hz, and use a test
record with LF test tones. Most RIAA preamps have some built-in
high-pass filtering which may mask this. The IEC variation to RIAA
de-emphasis added a roll-off below 30 Hz or so, if memory serves.

The response peak around the tone arm's fundamental resonance is
usually non-trivial, to say the least. ;-)


The way I have always tested for tonearm/cartridge resonance is to
use a test record which has a mid-frequency tone modulated with a
very low frequency tone, starting at around 5 Hz and going up to
about 15 Hz. If you listen to the tone and view the cartridge, you
will see the frequency at which the arm/cartridge visibly resonates.
The reproduction of the pure tone also audibly warbles at the LF
resonant frequency.
The HFN test record has such bands, as do several others.

As the test is essentially visual (although the audible warble also
helps) it is independent of the RIAA preamp.

S.

This sounds VERY cool. Any idea where one might pick up a copy of this test
record these days?

The test record is available from Hi-Fi News Accessories club. It can be
ordered on-line at www.hifiaccessoriesclub.com

Cost is £ sterling 25, or about $45, plus postage etc.

S.

[Mark D. Zacharias]

Serge Auckland wrote:
Mark D. Zacharias wrote:
Serge Auckland wrote:
Arny Krueger wrote:
"Mark D. Zacharias" wrote in
message

How would one set about measuring tonearm resonance? (I'm
suspecting one could feed a signal to the tonearm /
cartridge, and find the point of maximum deflection in
the under 20 hz region.)
Tone arm resonance often shows up, totally or in part, in Cool
Edit's frequency analysis of just about any old transcription
unless it has been high-pass filtered along the way.

Remember that its a rare piece of popular music that has any
musical fundamentals below 32 Hz. It was pretty common in the days
of, to high-pass LPs during production, to ensure trackability on
crappy equipment. By crappy equipment I mean a plastic or tin tone
arm and platter,
2-pole motor, cloth-covered cardboard chassis, and crystal
cartridge. IOW half or more of all LP playback equipment in
consumer use in the days of vinyl. I believe that Scott has recently
said
that the policy of
rolling-off the bass is being relaxed with some modern recordings,
because very little really crappy equipment is still in use.

The usual LP warps and non-flatness of the LP's substrate provides
a source of excitation for LF resonances. So, if there's a lot of
stuff below 32 Hz, and there is a discernable rise or peak, you are
probably looking at artifacts of the tone arm's resonance.

The *standard* way to find tone arm resonance was to have a phono
preamp that is flat below 50 Hz down to say 1 Hz, and use a test
record with LF test tones. Most RIAA preamps have some built-in
high-pass filtering which may mask this. The IEC variation to RIAA
de-emphasis added a roll-off below 30 Hz or so, if memory serves.

The response peak around the tone arm's fundamental resonance is
usually non-trivial, to say the least. ;-)


The way I have always tested for tonearm/cartridge resonance is to
use a test record which has a mid-frequency tone modulated with a
very low frequency tone, starting at around 5 Hz and going up to
about 15 Hz. If you listen to the tone and view the cartridge, you
will see the frequency at which the arm/cartridge visibly resonates.
The reproduction of the pure tone also audibly warbles at the LF
resonant frequency.
The HFN test record has such bands, as do several others.

As the test is essentially visual (although the audible warble also
helps) it is independent of the RIAA preamp.

S.

This sounds VERY cool. Any idea where one might pick up a copy of
this test record these days?

The test record is available from Hi-Fi News Accessories club. It can
be ordered on-line at www.hifiaccessoriesclub.com

Cost is £ sterling 25, or about $45, plus postage etc.

S.

Thanks!

mz

[Robert Orban]

In article om,
says...



The advantage of a digital filter is that you CAN build a filter
that has exactly the same frequency and phase characteristics
of an equivalent analog filter. And an added advantage is that
you can build filters that don't, if needed.

I would modify this statement as follows:

The advantage of a digital filter is that you CAN build a filter that
approximates the frequency and phase characteristics of an equivalent analog
filter to whatever accuracy you desire, provided that you allow addition of
an arbitrary pure delay to the filter. Strictly speaking, the added delay
causes the phase response of the digital filter to be very different from
the phase shift of the analog filter, although this is an academic argument
for most purposes(however not for live sound, sound for picture, and a
significant number of other real-world applications).

If you demand that your digital filter be minimum phase, then the digital
filter can never have exactly the same magnitude and phase response as the
prototype analog filter you are trying to duplicate. The differences will
often be quite large, particulaly as you approach Nyquist. The common
transforms (bilinear, matched-z, impulse-invariant, and step-invariant) all
have errors of some sort.

Regarding the main subject of this thread, IIR subsonic digital filters are
a challenge to implement correctly because they have high noise gain.
Highest performance requires choosing a low-noise structure and also making
sure that the finite word length arithmetic that implements the filters has
a sufficient number of bits. There is are a lot of research papers on this
subject.