August Helmbright wrote:
I have a new direct drive turntable with built-in strobe to correct
the speed. Unfortunately, the adjustements with the slider bar are
rather coarse, and I can't get the speed to be perfect - at best I get
it ever so slightly slow or fast. In the locked center position on the
slider bar, the table is slightly slow.

Since I only use the turntable for input into a computer and process
the sound files with DC-V filtering software for making CD-Rs (I never
actually "listen" to records), I wonder if someone can confirm my
thinking on how to determine how much the adjust the speed with my
sound filtering software, as follows. The example below assumes a
correction for a turntable that is too slow. It also deals only with
33 1/3. If my thinking is right, the same process could be adapted for
45 RPM, or modified to correct a speed that is too fast.

a. There are 130 dots around the circumference of my turntable for
33 1/3 RPM on the 60Hz band.

b. While a record is playing, to make sure stylus / record friction
is taken into account, I use a stopwatch to time the amount of time it
takes for 130 dots to move past the strobe counter-clockwise, to judge
how much time it takes for the turntable to lose one full turn
relative to correct speed. While it is possible I could be off by a
fraction of a second due to measurement errors(holding my head still,
etc.), I suspect this tiny difference would be inaudible.

c. I convert the time it takes to gain or lose a turn to minutes and
decimals.

d. I calculate the number of revolutions that should have occured at
correct speed by multiplying the result in (c) by 33 1/3.

e. To the result in (d), I subract 1 to arrive at the number of
revolutions at the turntable's actual speed.

f. I calculate the percentage slow by dividing the difference
between (d) and (e) by (e). [Not by (d), because the "percent change"
to apply to the sound file with filtering software is based on
correcting the actual file, which is at (e) speed.]

g. I run the speed filter in DC-V and reduce speed by the percent
(in my practice it's about a quarter of a percent) obtained in (f).

A quarter of a percent slow doesn't sound like much, but it does have
the effect of turning A=440 into A=438.9, and I'd rather know that
when I make CD-Rs, they're correctly pitched relative to the source
(assuming the studio mastering and LP pressing was right).

I don't know if anyone has suggested this, but why don't you disable the
slider bar and replace it with a multi-turn potentiometer?

It is almost certain that the slider bar is just a resistor. Replace it
with something that is similar in total resistance to the best setting
of the slider but with better resolution. If necessary, you can use a
narrow range high turn pot in series with a fix resistor.

If the regulator on the turntable is of high quality, you should be able
to match the speed such that A(440) is indistinguishable from a tuning fork.

August Helmbright wrote:
I have a new direct drive turntable with built-in strobe to correct
the speed. Unfortunately, the adjustements with the slider bar are
rather coarse, and I can't get the speed to be perfect - at best I get
it ever so slightly slow or fast. In the locked center position on the
slider bar, the table is slightly slow.

Since I only use the turntable for input into a computer and process
the sound files with DC-V filtering software for making CD-Rs (I never
actually "listen" to records), I wonder if someone can confirm my
thinking on how to determine how much the adjust the speed with my
sound filtering software, as follows. The example below assumes a
correction for a turntable that is too slow. It also deals only with
33 1/3. If my thinking is right, the same process could be adapted for
45 RPM, or modified to correct a speed that is too fast.

a. There are 130 dots around the circumference of my turntable for
33 1/3 RPM on the 60Hz band.

b. While a record is playing, to make sure stylus / record friction
is taken into account, I use a stopwatch to time the amount of time it
takes for 130 dots to move past the strobe counter-clockwise, to judge
how much time it takes for the turntable to lose one full turn
relative to correct speed. While it is possible I could be off by a
fraction of a second due to measurement errors(holding my head still,
etc.), I suspect this tiny difference would be inaudible.

c. I convert the time it takes to gain or lose a turn to minutes and
decimals.

d. I calculate the number of revolutions that should have occured at
correct speed by multiplying the result in (c) by 33 1/3.

e. To the result in (d), I subract 1 to arrive at the number of
revolutions at the turntable's actual speed.

f. I calculate the percentage slow by dividing the difference
between (d) and (e) by (e). [Not by (d), because the "percent change"
to apply to the sound file with filtering software is based on
correcting the actual file, which is at (e) speed.]

g. I run the speed filter in DC-V and reduce speed by the percent
(in my practice it's about a quarter of a percent) obtained in (f).

A quarter of a percent slow doesn't sound like much, but it does have
the effect of turning A=440 into A=438.9, and I'd rather know that
when I make CD-Rs, they're correctly pitched relative to the source
(assuming the studio mastering and LP pressing was right).

I don't know if anyone has suggested this, but why don't you disable the
slider bar and replace it with a multi-turn potentiometer?

It is almost certain that the slider bar is just a resistor. Replace it
with something that is similar in total resistance to the best setting
of the slider but with better resolution. If necessary, you can use a
narrow range high turn pot in series with a fix resistor.

If the regulator on the turntable is of high quality, you should be able
to match the speed such that A(440) is indistinguishable from a tuning fork.

August Helmbright wrote:
I have a new direct drive turntable with built-in strobe to correct
the speed. Unfortunately, the adjustements with the slider bar are
rather coarse, and I can't get the speed to be perfect - at best I get
it ever so slightly slow or fast. In the locked center position on the
slider bar, the table is slightly slow.

Since I only use the turntable for input into a computer and process
the sound files with DC-V filtering software for making CD-Rs (I never
actually "listen" to records), I wonder if someone can confirm my
thinking on how to determine how much the adjust the speed with my
sound filtering software, as follows. The example below assumes a
correction for a turntable that is too slow. It also deals only with
33 1/3. If my thinking is right, the same process could be adapted for
45 RPM, or modified to correct a speed that is too fast.

a. There are 130 dots around the circumference of my turntable for
33 1/3 RPM on the 60Hz band.

b. While a record is playing, to make sure stylus / record friction
is taken into account, I use a stopwatch to time the amount of time it
takes for 130 dots to move past the strobe counter-clockwise, to judge
how much time it takes for the turntable to lose one full turn
relative to correct speed. While it is possible I could be off by a
fraction of a second due to measurement errors(holding my head still,
etc.), I suspect this tiny difference would be inaudible.

c. I convert the time it takes to gain or lose a turn to minutes and
decimals.

d. I calculate the number of revolutions that should have occured at
correct speed by multiplying the result in (c) by 33 1/3.

e. To the result in (d), I subract 1 to arrive at the number of
revolutions at the turntable's actual speed.

f. I calculate the percentage slow by dividing the difference
between (d) and (e) by (e). [Not by (d), because the "percent change"
to apply to the sound file with filtering software is based on
correcting the actual file, which is at (e) speed.]

g. I run the speed filter in DC-V and reduce speed by the percent
(in my practice it's about a quarter of a percent) obtained in (f).

A quarter of a percent slow doesn't sound like much, but it does have
the effect of turning A=440 into A=438.9, and I'd rather know that
when I make CD-Rs, they're correctly pitched relative to the source
(assuming the studio mastering and LP pressing was right).

I don't know if anyone has suggested this, but why don't you disable the
slider bar and replace it with a multi-turn potentiometer?

It is almost certain that the slider bar is just a resistor. Replace it
with something that is similar in total resistance to the best setting
of the slider but with better resolution. If necessary, you can use a
narrow range high turn pot in series with a fix resistor.

If the regulator on the turntable is of high quality, you should be able
to match the speed such that A(440) is indistinguishable from a tuning fork.

On Fri, 12 Dec 2003 00:56:12 GMT, henryf
wrote:

Here's a government document that describes what's been
happening with frequency control in the North American power
system:
http://www.ornl.gov/sci/btc/apps/Res...NLTM200341.pdf

I found the conclusions on page 17 interesting:

"The preliminary investigation found that frequency is not
being controlled as tightly today as it was years ago.
Average frequency error doubled (from 0.002 to 0.004 Hz
standard deviation) ..."


The significance of this depends on how long the *Average * is
compared to how long August Helmbright spends counting dots. .

On Fri, 12 Dec 2003 00:56:12 GMT, henryf
wrote:

Here's a government document that describes what's been
happening with frequency control in the North American power
system:
http://www.ornl.gov/sci/btc/apps/Res...NLTM200341.pdf

I found the conclusions on page 17 interesting:

"The preliminary investigation found that frequency is not
being controlled as tightly today as it was years ago.
Average frequency error doubled (from 0.002 to 0.004 Hz
standard deviation) ..."


The significance of this depends on how long the *Average * is
compared to how long August Helmbright spends counting dots. .

On Fri, 12 Dec 2003 00:56:12 GMT, henryf
wrote:

Here's a government document that describes what's been
happening with frequency control in the North American power
system:
http://www.ornl.gov/sci/btc/apps/Res...NLTM200341.pdf

I found the conclusions on page 17 interesting:

"The preliminary investigation found that frequency is not
being controlled as tightly today as it was years ago.
Average frequency error doubled (from 0.002 to 0.004 Hz
standard deviation) ..."


The significance of this depends on how long the *Average * is
compared to how long August Helmbright spends counting dots. .

Robert Gault wrote:

I don't know if anyone has suggested this, but why don't you disable the
slider bar and replace it with a multi-turn potentiometer?

It is almost certain that the slider bar is just a resistor. Replace it
with something that is similar in total resistance to the best setting
of the slider but with better resolution. If necessary, you can use a
narrow range high turn pot in series with a fix resistor.

If the regulator on the turntable is of high quality, you should be able
to match the speed such that A(440) is indistinguishable from a tuning
fork.


Part of the problem is that the Stanton website doesn't have
schematics for their turntables. I saw a photo on the web
that showed the slider of a non-Stanton turntable plugging
into a control board that looked like it had a trim pot on it.
Who knows, maybe Stanton turntables already have an internal
trim pot that can be used to fine tune the speed? It would
make sense to me, but the Stanton web site is mum.

Robert Gault wrote:

I don't know if anyone has suggested this, but why don't you disable the
slider bar and replace it with a multi-turn potentiometer?

It is almost certain that the slider bar is just a resistor. Replace it
with something that is similar in total resistance to the best setting
of the slider but with better resolution. If necessary, you can use a
narrow range high turn pot in series with a fix resistor.

If the regulator on the turntable is of high quality, you should be able
to match the speed such that A(440) is indistinguishable from a tuning
fork.


Part of the problem is that the Stanton website doesn't have
schematics for their turntables. I saw a photo on the web
that showed the slider of a non-Stanton turntable plugging
into a control board that looked like it had a trim pot on it.
Who knows, maybe Stanton turntables already have an internal
trim pot that can be used to fine tune the speed? It would
make sense to me, but the Stanton web site is mum.

Robert Gault wrote:

I don't know if anyone has suggested this, but why don't you disable the
slider bar and replace it with a multi-turn potentiometer?

It is almost certain that the slider bar is just a resistor. Replace it
with something that is similar in total resistance to the best setting
of the slider but with better resolution. If necessary, you can use a
narrow range high turn pot in series with a fix resistor.

If the regulator on the turntable is of high quality, you should be able
to match the speed such that A(440) is indistinguishable from a tuning
fork.


Part of the problem is that the Stanton website doesn't have
schematics for their turntables. I saw a photo on the web
that showed the slider of a non-Stanton turntable plugging
into a control board that looked like it had a trim pot on it.
Who knows, maybe Stanton turntables already have an internal
trim pot that can be used to fine tune the speed? It would
make sense to me, but the Stanton web site is mum.

henryf wrote in message .net...
Robert Gault wrote:

I don't know if anyone has suggested this, but why don't you disable the
slider bar and replace it with a multi-turn potentiometer?

It is almost certain that the slider bar is just a resistor. Replace it
with something that is similar in total resistance to the best setting
of the slider but with better resolution. If necessary, you can use a
narrow range high turn pot in series with a fix resistor.

If the regulator on the turntable is of high quality, you should be able
to match the speed such that A(440) is indistinguishable from a tuning
fork.


Part of the problem is that the Stanton website doesn't have
schematics for their turntables. I saw a photo on the web
that showed the slider of a non-Stanton turntable plugging
into a control board that looked like it had a trim pot on it.
Who knows, maybe Stanton turntables already have an internal
trim pot that can be used to fine tune the speed? It would
make sense to me, but the Stanton web site is mum.

The person I talked to by phone at Stanton said they have no fine speed adjustment.

henryf wrote in message .net...
Robert Gault wrote:

I don't know if anyone has suggested this, but why don't you disable the
slider bar and replace it with a multi-turn potentiometer?

It is almost certain that the slider bar is just a resistor. Replace it
with something that is similar in total resistance to the best setting
of the slider but with better resolution. If necessary, you can use a
narrow range high turn pot in series with a fix resistor.

If the regulator on the turntable is of high quality, you should be able
to match the speed such that A(440) is indistinguishable from a tuning
fork.


Part of the problem is that the Stanton website doesn't have
schematics for their turntables. I saw a photo on the web
that showed the slider of a non-Stanton turntable plugging
into a control board that looked like it had a trim pot on it.
Who knows, maybe Stanton turntables already have an internal
trim pot that can be used to fine tune the speed? It would
make sense to me, but the Stanton web site is mum.

The person I talked to by phone at Stanton said they have no fine speed adjustment.

henryf wrote in message .net...
Robert Gault wrote:

I don't know if anyone has suggested this, but why don't you disable the
slider bar and replace it with a multi-turn potentiometer?

It is almost certain that the slider bar is just a resistor. Replace it
with something that is similar in total resistance to the best setting
of the slider but with better resolution. If necessary, you can use a
narrow range high turn pot in series with a fix resistor.

If the regulator on the turntable is of high quality, you should be able
to match the speed such that A(440) is indistinguishable from a tuning
fork.


Part of the problem is that the Stanton website doesn't have
schematics for their turntables. I saw a photo on the web
that showed the slider of a non-Stanton turntable plugging
into a control board that looked like it had a trim pot on it.
Who knows, maybe Stanton turntables already have an internal
trim pot that can be used to fine tune the speed? It would
make sense to me, but the Stanton web site is mum.

The person I talked to by phone at Stanton said they have no fine speed adjustment.

Many thanks for the "scratch test" method, which had frankly never
occured to me, simple as it seems once you think about it.

Upon further extremely close study of the waveforms resulting from the
radial scratches, I was able to position the curser such that it
seemed to my eye (at very fine resolution on the screen) to be in the
same position for each of 5 successive scratches, resulting in four
temporal measurements. In fact the largest differnce in the 4
measurements, showing that my eye for waveforms isn't perfect, was
0.0002 seconds. I averaged the 4 measurements and came up with 1.8081
instead of the correct 1.8 seconds, confirming what the strobe dots
show from a qualitative standpoint - that the turntable is slightly
too slow.

The amount by which I'm speeding up my wav files is 0.45% (0.0045),
which is the best the program allows in terms of decimal precision.
Assuming the 1.8081 is an absolutely correct measurement, the
difference between 0.45% and what is would be calculated based on the
scratch test results in a difference of approximately 0.0089 Hz
relative to 440 Hz. I'm satisfied I've gotten this thing much closer
to being right than will ever make a difference to my (or anyone
else's) ears. I'm also satisfied that the scratch test - particularly
based on averaging several extremely close measurements to minimize
the effects of small eyeballing errors with the waveforms - is much
more accurate that the stopwatch & strobe dot approach I had first
tried.

Many thanks for the "scratch test" method, which had frankly never
occured to me, simple as it seems once you think about it.

Upon further extremely close study of the waveforms resulting from the
radial scratches, I was able to position the curser such that it
seemed to my eye (at very fine resolution on the screen) to be in the
same position for each of 5 successive scratches, resulting in four
temporal measurements. In fact the largest differnce in the 4
measurements, showing that my eye for waveforms isn't perfect, was
0.0002 seconds. I averaged the 4 measurements and came up with 1.8081
instead of the correct 1.8 seconds, confirming what the strobe dots
show from a qualitative standpoint - that the turntable is slightly
too slow.

The amount by which I'm speeding up my wav files is 0.45% (0.0045),
which is the best the program allows in terms of decimal precision.
Assuming the 1.8081 is an absolutely correct measurement, the
difference between 0.45% and what is would be calculated based on the
scratch test results in a difference of approximately 0.0089 Hz
relative to 440 Hz. I'm satisfied I've gotten this thing much closer
to being right than will ever make a difference to my (or anyone
else's) ears. I'm also satisfied that the scratch test - particularly
based on averaging several extremely close measurements to minimize
the effects of small eyeballing errors with the waveforms - is much
more accurate that the stopwatch & strobe dot approach I had first
tried.

Many thanks for the "scratch test" method, which had frankly never
occured to me, simple as it seems once you think about it.

Upon further extremely close study of the waveforms resulting from the
radial scratches, I was able to position the curser such that it
seemed to my eye (at very fine resolution on the screen) to be in the
same position for each of 5 successive scratches, resulting in four
temporal measurements. In fact the largest differnce in the 4
measurements, showing that my eye for waveforms isn't perfect, was
0.0002 seconds. I averaged the 4 measurements and came up with 1.8081
instead of the correct 1.8 seconds, confirming what the strobe dots
show from a qualitative standpoint - that the turntable is slightly
too slow.

The amount by which I'm speeding up my wav files is 0.45% (0.0045),
which is the best the program allows in terms of decimal precision.
Assuming the 1.8081 is an absolutely correct measurement, the
difference between 0.45% and what is would be calculated based on the
scratch test results in a difference of approximately 0.0089 Hz
relative to 440 Hz. I'm satisfied I've gotten this thing much closer
to being right than will ever make a difference to my (or anyone
else's) ears. I'm also satisfied that the scratch test - particularly
based on averaging several extremely close measurements to minimize
the effects of small eyeballing errors with the waveforms - is much
more accurate that the stopwatch & strobe dot approach I had first
tried.

On 12 Dec 2003 11:40:23 -0800, (August
Helmbright) wrote:

Many thanks for the "scratch test" method, which had frankly never
occured to me, simple as it seems once you think about it.

Upon further extremely close study of the waveforms resulting from the
radial scratches, I was able to position the curser such that it
seemed to my eye (at very fine resolution on the screen) to be in the
same position for each of 5 successive scratches, resulting in four
temporal measurements. In fact the largest differnce in the 4
measurements, showing that my eye for waveforms isn't perfect, was
0.0002 seconds. I averaged the 4 measurements and came up with 1.8081
instead of the correct 1.8 seconds, confirming what the strobe dots
show from a qualitative standpoint - that the turntable is slightly
too slow.

The amount by which I'm speeding up my wav files is 0.45% (0.0045),
which is the best the program allows in terms of decimal precision.
Assuming the 1.8081 is an absolutely correct measurement, the
difference between 0.45% and what is would be calculated based on the
scratch test results in a difference of approximately 0.0089 Hz
relative to 440 Hz. I'm satisfied I've gotten this thing much closer
to being right than will ever make a difference to my (or anyone
else's) ears. I'm also satisfied that the scratch test - particularly
based on averaging several extremely close measurements to minimize
the effects of small eyeballing errors with the waveforms - is much
more accurate that the stopwatch & strobe dot approach I had first
tried.

Thanks for your kind words. You would not believe how difficult it is
to get vinyl lovers to attack records with a Stanley knife.

On 12 Dec 2003 11:40:23 -0800, (August
Helmbright) wrote:

Many thanks for the "scratch test" method, which had frankly never
occured to me, simple as it seems once you think about it.

Upon further extremely close study of the waveforms resulting from the
radial scratches, I was able to position the curser such that it
seemed to my eye (at very fine resolution on the screen) to be in the
same position for each of 5 successive scratches, resulting in four
temporal measurements. In fact the largest differnce in the 4
measurements, showing that my eye for waveforms isn't perfect, was
0.0002 seconds. I averaged the 4 measurements and came up with 1.8081
instead of the correct 1.8 seconds, confirming what the strobe dots
show from a qualitative standpoint - that the turntable is slightly
too slow.

The amount by which I'm speeding up my wav files is 0.45% (0.0045),
which is the best the program allows in terms of decimal precision.
Assuming the 1.8081 is an absolutely correct measurement, the
difference between 0.45% and what is would be calculated based on the
scratch test results in a difference of approximately 0.0089 Hz
relative to 440 Hz. I'm satisfied I've gotten this thing much closer
to being right than will ever make a difference to my (or anyone
else's) ears. I'm also satisfied that the scratch test - particularly
based on averaging several extremely close measurements to minimize
the effects of small eyeballing errors with the waveforms - is much
more accurate that the stopwatch & strobe dot approach I had first
tried.

Thanks for your kind words. You would not believe how difficult it is
to get vinyl lovers to attack records with a Stanley knife.

On 12 Dec 2003 11:40:23 -0800, (August
Helmbright) wrote:

Many thanks for the "scratch test" method, which had frankly never
occured to me, simple as it seems once you think about it.

Upon further extremely close study of the waveforms resulting from the
radial scratches, I was able to position the curser such that it
seemed to my eye (at very fine resolution on the screen) to be in the
same position for each of 5 successive scratches, resulting in four
temporal measurements. In fact the largest differnce in the 4
measurements, showing that my eye for waveforms isn't perfect, was
0.0002 seconds. I averaged the 4 measurements and came up with 1.8081
instead of the correct 1.8 seconds, confirming what the strobe dots
show from a qualitative standpoint - that the turntable is slightly
too slow.

The amount by which I'm speeding up my wav files is 0.45% (0.0045),
which is the best the program allows in terms of decimal precision.
Assuming the 1.8081 is an absolutely correct measurement, the
difference between 0.45% and what is would be calculated based on the
scratch test results in a difference of approximately 0.0089 Hz
relative to 440 Hz. I'm satisfied I've gotten this thing much closer
to being right than will ever make a difference to my (or anyone
else's) ears. I'm also satisfied that the scratch test - particularly
based on averaging several extremely close measurements to minimize
the effects of small eyeballing errors with the waveforms - is much
more accurate that the stopwatch & strobe dot approach I had first
tried.

Thanks for your kind words. You would not believe how difficult it is
to get vinyl lovers to attack records with a Stanley knife.

On 11 Dec 2003 14:19:33 -0800, (August Helmbright)
wrote:

Here's my report on using the 2 methods -- the one I suggested (stop
watch counting strobe dots movement past the light) and the other with
a radially scratched record.

With the stop watch method, I ended up with needing to speed up the
wav file by 0.56% (.0056). My program (DC-V) won't allow a speed
correction more precise than four decimal points.

With the radial scratch, there was some difficulty in counting the
exact right point to mark for each scratch, since the wave forms
looked slightly different for each time around. Therefore, I marked
several in a row and measured the time (down to thousanths of a
second) between each one and averaged them. The result of this
analysis was that I needed to speed up the wav file by 0.44% (.0044).

I suppose I shouldn't mention this, but it is extremely unlikely that your table
ran at a constant speed over the 15-20 minutes that it takes to play a record.
In fact it's been shown that needle drag, which is proportional to groove
modulation affects the speed, and the modulation of course varies as you play
the record. In fact the drag is LEAST when playing the empty grooves, so when
you timed it using those grooves, chance are the record was playing a bit faster
than it was during music grooves.

If you want to see this effect, you can watch the strobe marks as you set the
needle down on the record.

The other major factor is the temperature change of the motor and circuitry
which is why the table usually take 3-10 minutes of operation before it becomes
relatively stable in speed. Plus any voltage variation in main voltage may
affect the speed, depending on the circuitry in the drive system.

And of course, as I mentioned, it's not likely the original recording table and
tape recorders were accurate to the degree you are attempting to be. But I
applaud you for your effort to make a silk purse out of a sow's ear!

---------------------------------------------------------------
Alan Peterman al at scn.rain.com Tigard, OR
As I grow older, the days seem longer and the years seem shorter.

On 11 Dec 2003 14:19:33 -0800, (August Helmbright)
wrote:

Here's my report on using the 2 methods -- the one I suggested (stop
watch counting strobe dots movement past the light) and the other with
a radially scratched record.

With the stop watch method, I ended up with needing to speed up the
wav file by 0.56% (.0056). My program (DC-V) won't allow a speed
correction more precise than four decimal points.

With the radial scratch, there was some difficulty in counting the
exact right point to mark for each scratch, since the wave forms
looked slightly different for each time around. Therefore, I marked
several in a row and measured the time (down to thousanths of a
second) between each one and averaged them. The result of this
analysis was that I needed to speed up the wav file by 0.44% (.0044).

I suppose I shouldn't mention this, but it is extremely unlikely that your table
ran at a constant speed over the 15-20 minutes that it takes to play a record.
In fact it's been shown that needle drag, which is proportional to groove
modulation affects the speed, and the modulation of course varies as you play
the record. In fact the drag is LEAST when playing the empty grooves, so when
you timed it using those grooves, chance are the record was playing a bit faster
than it was during music grooves.

If you want to see this effect, you can watch the strobe marks as you set the
needle down on the record.

The other major factor is the temperature change of the motor and circuitry
which is why the table usually take 3-10 minutes of operation before it becomes
relatively stable in speed. Plus any voltage variation in main voltage may
affect the speed, depending on the circuitry in the drive system.

And of course, as I mentioned, it's not likely the original recording table and
tape recorders were accurate to the degree you are attempting to be. But I
applaud you for your effort to make a silk purse out of a sow's ear!

---------------------------------------------------------------
Alan Peterman al at scn.rain.com Tigard, OR
As I grow older, the days seem longer and the years seem shorter.

On 11 Dec 2003 14:19:33 -0800, (August Helmbright)
wrote:

Here's my report on using the 2 methods -- the one I suggested (stop
watch counting strobe dots movement past the light) and the other with
a radially scratched record.

With the stop watch method, I ended up with needing to speed up the
wav file by 0.56% (.0056). My program (DC-V) won't allow a speed
correction more precise than four decimal points.

With the radial scratch, there was some difficulty in counting the
exact right point to mark for each scratch, since the wave forms
looked slightly different for each time around. Therefore, I marked
several in a row and measured the time (down to thousanths of a
second) between each one and averaged them. The result of this
analysis was that I needed to speed up the wav file by 0.44% (.0044).

I suppose I shouldn't mention this, but it is extremely unlikely that your table
ran at a constant speed over the 15-20 minutes that it takes to play a record.
In fact it's been shown that needle drag, which is proportional to groove
modulation affects the speed, and the modulation of course varies as you play
the record. In fact the drag is LEAST when playing the empty grooves, so when
you timed it using those grooves, chance are the record was playing a bit faster
than it was during music grooves.

If you want to see this effect, you can watch the strobe marks as you set the
needle down on the record.

The other major factor is the temperature change of the motor and circuitry
which is why the table usually take 3-10 minutes of operation before it becomes
relatively stable in speed. Plus any voltage variation in main voltage may
affect the speed, depending on the circuitry in the drive system.

And of course, as I mentioned, it's not likely the original recording table and
tape recorders were accurate to the degree you are attempting to be. But I
applaud you for your effort to make a silk purse out of a sow's ear!

---------------------------------------------------------------
Alan Peterman al at scn.rain.com Tigard, OR
As I grow older, the days seem longer and the years seem shorter.

Alan Peterman wrote in message . ..
I suppose I shouldn't mention this, but it is extremely unlikely that your table
ran at a constant speed over the 15-20 minutes that it takes to play a record.
In fact it's been shown that needle drag, which is proportional to groove
modulation affects the speed, and the modulation of course varies as you play
the record. In fact the drag is LEAST when playing the empty grooves, so when
you timed it using those grooves, chance are the record was playing a bit faster
than it was during music grooves.

If you want to see this effect, you can watch the strobe marks as you set the
needle down on the record.

The other major factor is the temperature change of the motor and circuitry
which is why the table usually take 3-10 minutes of operation before it becomes
relatively stable in speed. Plus any voltage variation in main voltage may
affect the speed, depending on the circuitry in the drive system.

And of course, as I mentioned, it's not likely the original recording table and
tape recorders were accurate to the degree you are attempting to be. But I
applaud you for your effort to make a silk purse out of a sow's ear!

---------------------------------------------------------------
Alan Peterman al at scn.rain.com Tigard, OR
As I grow older, the days seem longer and the years seem shorter.

While you are correct from a technical standpoint, it is also true
that voices and musical instruments in live performance do not
maintain absolute accuracy in pitch either. Pitches change slightly as
instruments warm up - particularly noticeable if you're measuring
frequencies scientifically in woodwind instruments, where the bores
contract as the wood expands due to heat. That being said, the human
ear is very forgiving of small deviations, which is why the equal
tuning system used in pianos and other keyboard instruments (every
note is just a bit out of tune compared to the overtone series)
doesn't offend greatly.

Since I'm working for musical rather than scientific precision,
anything I can do to get *average* turntable speed as close as
possible to true 33 1/3 RPM (either mechanically or, as I'm doing,
using software to change the speed of a wav file) gets me well within
human precision, even for someone with a very good musical ear. If
it's scientific precision we need, sown to the smallest fraction of a
Hertz (thank God it's not - we'd all go nuts), then there's probably:
(1) no musical performance live that would fill the bill; and (2) no
method of reproducing a performance in recorded form that is exactly
in sync at all times with the original.

Alan Peterman wrote in message . ..
I suppose I shouldn't mention this, but it is extremely unlikely that your table
ran at a constant speed over the 15-20 minutes that it takes to play a record.
In fact it's been shown that needle drag, which is proportional to groove
modulation affects the speed, and the modulation of course varies as you play
the record. In fact the drag is LEAST when playing the empty grooves, so when
you timed it using those grooves, chance are the record was playing a bit faster
than it was during music grooves.

If you want to see this effect, you can watch the strobe marks as you set the
needle down on the record.

The other major factor is the temperature change of the motor and circuitry
which is why the table usually take 3-10 minutes of operation before it becomes
relatively stable in speed. Plus any voltage variation in main voltage may
affect the speed, depending on the circuitry in the drive system.

And of course, as I mentioned, it's not likely the original recording table and
tape recorders were accurate to the degree you are attempting to be. But I
applaud you for your effort to make a silk purse out of a sow's ear!

---------------------------------------------------------------
Alan Peterman al at scn.rain.com Tigard, OR
As I grow older, the days seem longer and the years seem shorter.

While you are correct from a technical standpoint, it is also true
that voices and musical instruments in live performance do not
maintain absolute accuracy in pitch either. Pitches change slightly as
instruments warm up - particularly noticeable if you're measuring
frequencies scientifically in woodwind instruments, where the bores
contract as the wood expands due to heat. That being said, the human
ear is very forgiving of small deviations, which is why the equal
tuning system used in pianos and other keyboard instruments (every
note is just a bit out of tune compared to the overtone series)
doesn't offend greatly.

Since I'm working for musical rather than scientific precision,
anything I can do to get *average* turntable speed as close as
possible to true 33 1/3 RPM (either mechanically or, as I'm doing,
using software to change the speed of a wav file) gets me well within
human precision, even for someone with a very good musical ear. If
it's scientific precision we need, sown to the smallest fraction of a
Hertz (thank God it's not - we'd all go nuts), then there's probably:
(1) no musical performance live that would fill the bill; and (2) no
method of reproducing a performance in recorded form that is exactly
in sync at all times with the original.

Alan Peterman wrote in message . ..
I suppose I shouldn't mention this, but it is extremely unlikely that your table
ran at a constant speed over the 15-20 minutes that it takes to play a record.
In fact it's been shown that needle drag, which is proportional to groove
modulation affects the speed, and the modulation of course varies as you play
the record. In fact the drag is LEAST when playing the empty grooves, so when
you timed it using those grooves, chance are the record was playing a bit faster
than it was during music grooves.

If you want to see this effect, you can watch the strobe marks as you set the
needle down on the record.

The other major factor is the temperature change of the motor and circuitry
which is why the table usually take 3-10 minutes of operation before it becomes
relatively stable in speed. Plus any voltage variation in main voltage may
affect the speed, depending on the circuitry in the drive system.

And of course, as I mentioned, it's not likely the original recording table and
tape recorders were accurate to the degree you are attempting to be. But I
applaud you for your effort to make a silk purse out of a sow's ear!

---------------------------------------------------------------
Alan Peterman al at scn.rain.com Tigard, OR
As I grow older, the days seem longer and the years seem shorter.

While you are correct from a technical standpoint, it is also true
that voices and musical instruments in live performance do not
maintain absolute accuracy in pitch either. Pitches change slightly as
instruments warm up - particularly noticeable if you're measuring
frequencies scientifically in woodwind instruments, where the bores
contract as the wood expands due to heat. That being said, the human
ear is very forgiving of small deviations, which is why the equal
tuning system used in pianos and other keyboard instruments (every
note is just a bit out of tune compared to the overtone series)
doesn't offend greatly.

Since I'm working for musical rather than scientific precision,
anything I can do to get *average* turntable speed as close as
possible to true 33 1/3 RPM (either mechanically or, as I'm doing,
using software to change the speed of a wav file) gets me well within
human precision, even for someone with a very good musical ear. If
it's scientific precision we need, sown to the smallest fraction of a
Hertz (thank God it's not - we'd all go nuts), then there's probably:
(1) no musical performance live that would fill the bill; and (2) no
method of reproducing a performance in recorded form that is exactly
in sync at all times with the original.

(August Helmbright) wrote in message . com...

...., it is also true
that voices and musical instruments in live performance do not
maintain absolute accuracy in pitch either. Pitches change slightly as
instruments warm up - particularly noticeable if you're measuring
frequencies scientifically in woodwind instruments, where the bores
contract as the wood expands due to heat.

I can't help myself here, I have to add that there also is an effect
of the CO2 in the exhaled air. The CO2 content goes up towards the end
of a breath, which reduces the speed of sound within the instrument,
and so the pitch drops.

http://www.speech.kth.se/qpsr/tmh/19...30-037-043.pdf

....but, as stated in this article, it is likely that good musicians
compensate for these effects such that the effects are kept about as
small as what is audible.

So, it all comes down to what is audible. I seem to recall that a
frequency modulation of 0.2% was audible in the days of wow and
flutter measurements (correct me if I am wrong), that should give a
hint regarding the precision needed.

(August Helmbright) wrote in message . com...

...., it is also true
that voices and musical instruments in live performance do not
maintain absolute accuracy in pitch either. Pitches change slightly as
instruments warm up - particularly noticeable if you're measuring
frequencies scientifically in woodwind instruments, where the bores
contract as the wood expands due to heat.

I can't help myself here, I have to add that there also is an effect
of the CO2 in the exhaled air. The CO2 content goes up towards the end
of a breath, which reduces the speed of sound within the instrument,
and so the pitch drops.

http://www.speech.kth.se/qpsr/tmh/19...30-037-043.pdf

....but, as stated in this article, it is likely that good musicians
compensate for these effects such that the effects are kept about as
small as what is audible.

So, it all comes down to what is audible. I seem to recall that a
frequency modulation of 0.2% was audible in the days of wow and
flutter measurements (correct me if I am wrong), that should give a
hint regarding the precision needed.

(August Helmbright) wrote in message . com...

...., it is also true
that voices and musical instruments in live performance do not
maintain absolute accuracy in pitch either. Pitches change slightly as
instruments warm up - particularly noticeable if you're measuring
frequencies scientifically in woodwind instruments, where the bores
contract as the wood expands due to heat.

I can't help myself here, I have to add that there also is an effect
of the CO2 in the exhaled air. The CO2 content goes up towards the end
of a breath, which reduces the speed of sound within the instrument,
and so the pitch drops.

http://www.speech.kth.se/qpsr/tmh/19...30-037-043.pdf

....but, as stated in this article, it is likely that good musicians
compensate for these effects such that the effects are kept about as
small as what is audible.

So, it all comes down to what is audible. I seem to recall that a
frequency modulation of 0.2% was audible in the days of wow and
flutter measurements (correct me if I am wrong), that should give a
hint regarding the precision needed.