"Bob Cain" wrote

In double blind testing done recently no one can detect a difference
between 48 kHz and any rate above that and only a very few can tell the
difference between 44.1 and 48 kHz.


Can you provide links to the above tests?

'Fraid not. The results were recently reported in one of the pro audio
newsgroups or mailing lists I subscribe to. It may have been done by Arny
Kruger on rec.pro.audio but I am not certain.

Arny, If you are out there, I'd really like to see those results.

Bob, you didn't answer my question if whether you think the reason some
people can tell the difference between 44.1 and 48 is because it puts the
filters up at a higher frequency out of the audible range. Even though you
can be annoying to debate with, I do value your opinion.

If the dither is done right, the noise cannot be distinguished from the
addition of random white noise. If you hear an artifact it's because the
dithering was not done right and with modern software that just won't
happen.

You won't hear artifacts at most *normal* listening volumes and most
*normal* program materials. IMO 16 bit is barely adequate. I'm in good
company with that belief as Sony seems to agree with me. If not, why would
they bother with 20 bit HiMD or 20 bit "Super Bit Mapping" on their studio
DAT machines?

Yes, whatever analog tape does to sound it seems to be euphonic to most
people. I've read that is not true of newer, less aclimatized ears but
I've no link to back that up.

Have you given time to any of the DSP plugins which attempt to model the
tape process on digital recordings? I haven't but then I came to audio
via DSP so I have little frame of reference with tape recording.

I haven't listened to them but I've read about them. From what I've read,
boosting low midrange around 400 Hz is a big part of it, just where I said
the old analog engineers used to like to cut here. Hmmm.... I don't really
care enough about it to bother any serious testing. Like I keep saying
although I'm taking a hard line that 44/16 is not quite adequate, I don't
really give a crap. My ATRAC recordings with a $90 mic transferred
digitally to CD sound just fine to me. My Sony 20 bit SBM DAT sounds
better, but I'd rather fit a MD in my shirt pocket than haul the 25 pound
DAT machine around in it's road case. I'd probably prefer a 100 lb. Otari
mastering machine to the DAT, but who cares?

I do wish CD's were 48/20 instead of 44/16.

BTW, I've always wondered why 44.1. Why not 44 or 45? Do you know?

I merely supplied an explanation and
context for why the original post claimed "digital sucks".

Yeah, me too. It sucks, I suppose, if one is really attached to the tape
artifacts.

I have no idea how old John "Daffy Duck" is, but I sounds like he's never
heard an LP, so tape artifacts can't be the reason in his case!

Julian

"Julian Adamaitis" wrote in message

"Bob Cain" wrote

In double blind testing done recently no one can detect a
difference between 48 kHz and any rate above that and only a very
few can tell the difference between 44.1 and 48 kHz.


Can you provide links to the above tests?

'Fraid not. The results were recently reported in one of the pro
audio newsgroups or mailing lists I subscribe to. It may have been
done by Arny Kruger on rec.pro.audio but I am not certain.

Arny, If you are out there, I'd really like to see those results.

Please see

http://www.pcabx.com/technical/low_pass/index.htm

http://www.pcabx.com/technical/sample_rates/index.htm

Bob, you didn't answer my question if whether you think the reason
some people can tell the difference between 44.1 and 48 is because
it
puts the filters up at a higher frequency out of the audible range.

It's something about how the human ear is made.


You won't hear artifacts at most *normal* listening volumes and most
*normal* program materials. IMO 16 bit is barely adequate. I'm in
good company with that belief as Sony seems to agree with me. If
not, why would they bother with 20 bit HiMD or 20 bit "Super Bit
Mapping" on their studio DAT machines?

It sells hardware with numbers, as opposed to audible performance
improvements.

"Arny Krueger" wrote
Please see

http://www.pcabx.com/technical/low_pass/index.htm

http://www.pcabx.com/technical/sample_rates/index.htm

Thanks. Lot's of graphs and sound files. What do you conclude from them?

Bob, you didn't answer my question if whether you think the reason
some people can tell the difference between 44.1 and 48 is because
it
puts the filters up at a higher frequency out of the audible range.

It's something about how the human ear is made.

What'd ya mean?

Julian

"Julian Adamaitis" wrote in message

"Arny Krueger" wrote
Please see

http://www.pcabx.com/technical/low_pass/index.htm

http://www.pcabx.com/technical/sample_rates/index.htm

Thanks. Lot's of graphs and sound files. What do you conclude from
them?

The idea is to listen to the sound files for yourself using the free,
provided DBT comparison software, and reach your own conclusion.

Bob, you didn't answer my question if whether you think the reason
some people can tell the difference between 44.1 and 48 is because
it
puts the filters up at a higher frequency out of the audible
range.

It's something about how the human ear is made.

What'd ya mean?

The human ear is pretty much incapable of hearing the difference
between a well-done 44.1 KHz sample rate file and one made at any
higher sample rate. The ability to hear differences in frequency
response goes south around 15-16 KHz.

"Arny Krueger" wrote

The idea is to listen to the sound files for yourself using the free,
provided DBT comparison software, and reach your own conclusion.

I know that's the idea. I just wonder what conclusions you have reached.

Julian

Julian Adamaitis wrote:
"Arny Krueger" wrote

The idea is to listen to the sound files for yourself using the
free,
provided DBT comparison software, and reach your own conclusion.

I know that's the idea. I just wonder what conclusions you have
reached.

If you code music with 14 bits and a sample rate of 32 KHz, audible
artifacts due to the sampling might just intrude.

Arny Krueger wrote:
The human ear is pretty much incapable of hearing the difference
between a well-done 44.1 KHz sample rate file and one made at any
higher sample rate. The ability to hear differences in frequency
response goes south around 15-16 KHz.


Many folks are tested as to hearing range, and you might be surprised at
some of the results. As for myself, Many Moons Ago (I do not think I
could do this today) I was working as a security guard in a store.
Well, one night at closing they forgot to turn on the ultrasonic motion
detector alarm system... I informed the manager it had not been
activated. He ask how I knew. I told him I could not hear it.

He thought I was kidding... Went back and ... Turned it on (I heard it
clearly after it was turned on)

Not everyone has the same high frequency cut off
--
John F Davis, in Delightful Detroit. WA8YXM(at)arrl(dot)net
"Nothing adds excitement like something that is none of your business"
Diabetic? http://community.compuserve.com/diabetes

John in Detroit wrote:
Arny Krueger wrote:
The human ear is pretty much incapable of hearing the difference
between a well-done 44.1 KHz sample rate file and one made at any
higher sample rate. The ability to hear differences in frequency
response goes south around 15-16 KHz.


Many folks are tested as to hearing range, and you might be
surprised
at some of the results.

No, I'm not.

Hearing a pure tone at XX KHz is not the same as hearing the removal
of all information above XX KHz.

As for myself, Many Moons Ago (I do not
think I could do this today) I was working as a security guard in a
store. Well, one night at closing they forgot to turn on the
ultrasonic motion detector alarm system... I informed the manager it
had not been activated. He ask how I knew. I told him I could not
hear it.

See above. I heard ultrasonic alarms, as well. Make it intense enough
and maybe somehow you'll notice it. But is it really the same as
hearing intelligble sounds?

He thought I was kidding... Went back and ... Turned it on (I heard
it clearly after it was turned on)

Not everyone has the same high frequency cut off

Agreed. My 16 KKz number is optimistic for older folks. 12 KHz would
be a better number for some of them.

Arny Krueger wrote:

Please see

http://www.pcabx.com/technical/low_pass/index.htm

http://www.pcabx.com/technical/sample_rates/index.htm

Sorry for the misattribution, Arny. Someone recently posted
(somewhere) that they had done DBT testing with a number of
subjects and that the results showed some discrimination
between 44.1 and 48 but none above that. Anybody remember
who that was and where it was reported?

You won't hear artifacts at most *normal* listening volumes and most
*normal* program materials. IMO 16 bit is barely adequate. I'm in
good company with that belief as Sony seems to agree with me. If
not, why would they bother with 20 bit HiMD or 20 bit "Super Bit
Mapping" on their studio DAT machines?


It sells hardware with numbers, as opposed to audible performance
improvements.

Doesn't SBM shape the low order noise so as to give a higher
perceptual dynamic range than 16 bits can with random
dither? IIRC, they claimed it to be around 18 bits.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein

"Bob Cain" wrote

Doesn't SBM shape the low order noise so as to give a higher perceptual
dynamic range than 16 bits can with random dither? IIRC, they claimed it
to be around 18 bits.

sounds about right. My only point is that Sony thought there was something
needed to be done, also I've heard talk HiMD is capable of 20 bit, but all I
know is talk.

Julian

Bob Cain wrote:
Arny Krueger wrote:

Julian Adamaitis wrote:

You won't hear artifacts at most *normal* listening volumes and
most
*normal* program materials. IMO 16 bit is barely adequate. I'm
in
good company with that belief as Sony seems to agree with me. If
not, why would they bother with 20 bit HiMD or 20 bit "Super Bit
Mapping" on their studio DAT machines?

It sells hardware with numbers, as opposed to audible performance
improvements.

Doesn't SBM shape the low order noise so as to give a higher
perceptual dynamic range than 16 bits can with random
dither? IIRC, they claimed it to be around 18 bits.

Yes, as does any other well-engineered generic approach that shapes
the noise floor in according to the sensitivity of the human ear.
Trouble is, even in an *unshaped* 16 bit system, the noise floor due
to quantization is well below the noise floor of real-world musical
sources.

Some years ago I challenged Glen Zelniker developer of another
proprietary noise shaping scheme to demonstrate the effectiveness of
his system with DBTs, based on real-world musical program material.
Didn't happen, did it?

Arny Krueger wrote:

Some years ago I challenged Glen Zelniker developer of another
proprietary noise shaping scheme to demonstrate the effectiveness of
his system with DBTs, based on real-world musical program material.
Didn't happen, did it?

A few years ago, we did a set of single blind tests on noise shaping systems,
and what was interesting is that people perceived changes in tonality of
the program material as a result of the different noise shaping algorithms.
As for me, I liked the straight Gaussian best.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

Arny Krueger wrote:

Trouble is, even in an *unshaped* 16 bit system, the noise floor due
to quantization is well below the noise floor of real-world musical
sources.

Granted in full. However, this is in regard to using SBM
for recording where the extra 12 dB of potential headroom
can help a lot when, as is often the case, you just don't
know before hand where the performance will peak.


Bob
--

"Things should be described as simply as possible, but no
simpler."

A. Einstein

Bob Cain wrote:
Arny Krueger wrote:

Trouble is, even in an *unshaped* 16 bit system, the noise floor
due
to quantization is well below the noise floor of real-world musical
sources.

Granted in full. However, this is in regard to using SBM
for recording where the extra 12 dB of potential headroom
can help a lot when, as is often the case, you just don't
know before hand where the performance will peak.

I guess this was more important when the DAT format (16 bits) was as
far as we could practically go.

Today just about every audio interface worth its salt does 24 data
bits, and has analog performance that is 6-12 dB better than the best
that 16 bits can do.

In article writes:

Doesn't SBM shape the low order noise so as to give a higher
perceptual dynamic range than 16 bits can with random
dither? IIRC, they claimed it to be around 18 bits.

It's been a while but as I recall, Sony's SBM (Super Bit Mapping) was
a 20-bit A/D converter which, in that era was a pretty good 16-bit
converter with a 20-bit outout, with their own flavor of noise shaped
dithering. When truncated by a 16-bit recorder, it had pretty close to
16-bit resolution, pretty close to accurate.


--
I'm really Mike Rivers )
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me he double-m-eleven-double-zero at yahoo

In article writes:

I do wish CD's were 48/20 instead of 44/16.

With CD players selling for under $20, and CDs with less than 10 dB of
dynamic range what difference do you think it would really make? You
think this would prompt manufacturers to make better CD players, or
producer to make CDs with greater dynamic range?

BTW, I've always wondered why 44.1. Why not 44 or 45? Do you know?

There are various verisons of this piece of history. The most
interesting one was that Sony looked at recorded classical works
(seems like I recall that Mrs. Stokowski was involved) and decided
that in order to fit what they wanted on to a single disk, they needed
the playing time afforded by 44.1 kHz, so that's what they used.
Another explanation was that the digital hardware used for recording,
editing, and mastering were based on videotape, and 44.1 kHz sample
rate fit nicely in with the video scan rate.

The original Soundstream digital recorders ran at 50 kHz sample rate,
which they decided was the optimum.


--
I'm really Mike Rivers )
However, until the spam goes away or Hell freezes over,
lots of IP addresses are blocked from this system. If
you e-mail me and it bounces, use your secret decoder ring
and reach me he double-m-eleven-double-zero at yahoo

Mike Rivers wrote:
In article
writes:

I do wish CD's were 48/20 instead of 44/16.

With CD players selling for under $20, and CDs with less than 10 dB
of
dynamic range what difference do you think it would really make? You
think this would prompt manufacturers to make better CD players, or
producer to make CDs with greater dynamic range?

BTW, I've always wondered why 44.1. Why not 44 or 45? Do you
know?


From:
http://www.cdrfaq.org/faq02.html#S2-35

According to John Watkinson's _The Art of Digital Audio_, 2nd edition,
page 104, the choice of frequency is an artifact of the equipment used
during early digital audio research. Storing digital audio on a hard
drive was impractical, because the capacity needed for significant
amounts of 1 Mbps audio was expensive. Instead, they used video
recorders, storing samples as black and white levels. If you take the
number of 16-bit stereo samples you can get on a line, and multiply it
by the number of recorded lines in a field and the number of fields per
second, you get the sampling rate. It turned out that both NTSC and PAL
formats (the video standards used in US/Japan and Europe, respectively)
could handle a rate of 44100 samples per second. This rate was carried
over into the definition of the compact disc.

Mark wrote:

From:
http://www.cdrfaq.org/faq02.html#S2-35

According to John Watkinson's _The Art of Digital Audio_, 2nd
edition,
page 104, the choice of frequency is an artifact of the equipment
used
during early digital audio research. Storing digital audio on a hard
drive was impractical, because the capacity needed for significant
amounts of 1 Mbps audio was expensive. Instead, they used video
recorders, storing samples as black and white levels. If you take
the
number of 16-bit stereo samples you can get on a line, and multiply
it
by the number of recorded lines in a field and the number of fields
per second, you get the sampling rate. It turned out that both NTSC
and PAL formats (the video standards used in US/Japan and Europe,
respectively) could handle a rate of 44100 samples per second. This
rate was carried over into the definition of the compact disc.

Since 44.1 is mild overkill, looks like they did a good job of picking
a common sample rate.

On 4 Apr 2005 10:25:07 -0400, (Mike Rivers) wrote:

BTW, I've always wondered why 44.1. Why not 44 or 45? Do you know?

There are various verisons of this piece of history. The most
interesting one was that Sony looked at recorded classical works
(seems like I recall that Mrs. Stokowski was involved) and decided
that in order to fit what they wanted on to a single disk, they needed
the playing time afforded by 44.1 kHz, so that's what they used.
Another explanation was that the digital hardware used for recording,
editing, and mastering were based on videotape, and 44.1 kHz sample
rate fit nicely in with the video scan rate.

A related story, told by Ken Pohlmann, is that Herbert von Karajan
was asked how long a playing time was needed. His Beethoven Ninth
would fit in 74 minutes, and the rest is history. The disk diameter
was limited by the opening in auto dashes, pit size and spiral
pitch by contemporary lasers.

Chris Hornbeck
6x9=42