Carey Carlan wrote:
Yes, 14 bits are plenty for good dynamics.

Notice how grainy the flute sounds, though? It's not horrible, but it's
not completely natural. On a modern 16-bit converter it would be much
more smooth.

But considering the technology of the day when it was recorded...
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

James Boyk wrote in
:

Carey Carlan wrote: Yes, 14 bits are plenty for good dynamics.

The BBC published a Research Dept. paper long ago with the conclusion
that linear PCM required 22 bits to capture the dynamics of an
"unconditioned" signal (uncompressed, etc.) from sources of widest
dynamic range (e.g., big band); 20 bits for the actual dynamic range,
and 2 more to allow for possible mis-setting of the window.

We're talking specific examples here. Take your ears out of the book and
listen. Given the room acoustics, the amount of background noise, the
power of the given instruments, can your ears truly say that, in THIS
recording, the dynamic range was compromised? I'm not asking if the
recording is perfect. I'm just asking if the dynamic range of the
recording was adequate to record the performance you hear?

Somewhere, sometime a 22 bit recording is necessary. This wasn't it.

James Boyk wrote in
:

The BBC published a Research Dept. paper long ago with the conclusion
that linear PCM required 22 bits to capture the dynamics of an
"unconditioned" signal (uncompressed, etc.) from sources of widest
dynamic range (e.g., big band); 20 bits for the actual dynamic range,
and 2 more to allow for possible mis-setting of the window.

LOL. No ****. I mean, do you really need BBC to tell you 20-22 bits
are needed to cover some 120-130dB from the threshold of hearing to
the threshold of pain? "Published a paper"???. Man, I'm still
laughing. How much more pompous can you get?
Vladan
www.geocities.com/vla_dan_l
www.mp3.com/lesly , www.mp3.com/shook , www.mp3.com/lesly2
www.kunsttick.com/artists/vuskovic/indexdat.htm

In article writes:

Notice how grainy the flute sounds, though? It's not horrible, but it's
not completely natural.

It's a funky wooden flute. I think it sounds quite realistic, but then
I wasn't there.



--
I'm really Mike Rivers - )

In article Vladan
writes:
James Boyk wrote in
:

The BBC published a Research Dept. paper long ago with the conclusion
that linear PCM required 22 bits to capture the dynamics of an
"unconditioned" signal (uncompressed, etc.) from sources of widest
dynamic range (e.g., big band); 20 bits for the actual dynamic range,
and 2 more to allow for possible mis-setting of the window.

LOL. No ****. I mean, do you really need BBC to tell you 20-22 bits
are needed to cover some 120-130dB from the threshold of hearing to
the threshold of pain? "Published a paper"???. Man, I'm still
laughing. How much more pompous can you get?
Vladan

The BBC knows something about his topic. You're free to disparage their work
when you have presented data to refute theirs.

-Jay
--
x------- Jay Kadis ------- x---- Jay's Attic Studio ----x
x Lecturer, Audio Engineer x Dexter Records x
x CCRMA, Stanford University x http://www.offbeats.com/ x
x-------- http://ccrma-www.stanford.edu/~jay/ ----------x

On Mon, 11 Aug 2003 21:37:11 -0700, James Boyk
wrote:

And so pompous you think that a statistical matter has an absolute answer.

????????? Sorry, I'm too stupid fior your wisdom.

Vladan
www.geocities.com/vla_dan_l
www.mp3.com/lesly , www.mp3.com/shook , www.mp3.com/lesly2
www.kunsttick.com/artists/vuskovic/indexdat.htm

On Tue, 12 Aug 2003 14:35:49 +0000 (UTC), (Jay
Kadis) wrote:

The BBC knows something about his topic. You're free to disparage their work
when you have presented data to refute theirs.

I don't disagree with BBC. 20-22db is exactly how much bits is needed
to cover whole dynamic range of effective hearing, from silence to
pain. But when someone as pompous as Boyk comes in with "BBC published
a paper" on that, that's just plain ridiculous and funny. It may be
good enough to impress his students: ... wow this "professeure" of
ours, he reads AES papers, wow, one day we may end being so educate,
like he is. I can't help but LOL.

Vladan
www.geocities.com/vla_dan_l
www.mp3.com/lesly , www.mp3.com/shook , www.mp3.com/lesly2
www.kunsttick.com/artists/vuskovic/indexdat.htm

In article "Arny Krueger"
writes:
"James Boyk" wrote in message

Carey Carlan wrote: Yes, 14 bits are plenty for good dynamics.

The BBC published a Research Dept. paper long ago with the conclusion
that linear PCM required 22 bits to capture the dynamics of an
"unconditioned" signal (uncompressed, etc.) from sources of widest
dynamic range (e.g., big band); 20 bits for the actual dynamic range,
and 2 more to allow for possible mis-setting of the window.

There's some totally unreasonable assumptions here - like that the big band
would be playing in a NC10 room, and that the musicians themselves
contribute only music to the sound level in the room.


They're interested in the limiting cases, not the average.

I'm surprised that measurements on the deck of an fully-operational aircraft
carrier deck were left out (and not compared to some empty chamber in a dry
cave 1,000 feet underground)!

Back in the real world, if you ever find music that peaks say 80 dB over
room tone:

(a) be sure you are using ear protection
(b) make a note of it, because it's an outrageously exceptional condition


Our studio noise floor is about 21 dBA. The output of a Musicman 2x12 guitar
amp is often 110 dB. That's not unusual at all in popular music. (Well, maybe
the 21 dB noise floor is a bit atypical...)

I agree on the hearing protection.

-Jay
--
x------- Jay Kadis ------- x---- Jay's Attic Studio ----x
x Lecturer, Audio Engineer x Dexter Records x
x CCRMA, Stanford University x http://www.offbeats.com/ x
x-------- http://ccrma-www.stanford.edu/~jay/ ----------x

I think you guys are inventing an argument where there doesn't need to
be one. Surely the BBC's research told them they didn't need MORE than
22 bits to capture the dynamics of an "unconditioned" signal. They
obviously know that not all signals "require" that much resolution.

ulysses


In article , Carey
Carlan wrote:

James Boyk wrote in
:

Carey Carlan wrote: Yes, 14 bits are plenty for good dynamics.

The BBC published a Research Dept. paper long ago with the conclusion
that linear PCM required 22 bits to capture the dynamics of an
"unconditioned" signal (uncompressed, etc.) from sources of widest
dynamic range (e.g., big band); 20 bits for the actual dynamic range,
and 2 more to allow for possible mis-setting of the window.

We're talking specific examples here. Take your ears out of the book and
listen. Given the room acoustics, the amount of background noise, the
power of the given instruments, can your ears truly say that, in THIS
recording, the dynamic range was compromised? I'm not asking if the
recording is perfect. I'm just asking if the dynamic range of the
recording was adequate to record the performance you hear?

Somewhere, sometime a 22 bit recording is necessary. This wasn't it.

"Jay Kadis" wrote in message

In article "Arny Krueger"
writes:
"James Boyk" wrote in message

Carey Carlan wrote: Yes, 14 bits are plenty for good dynamics.

The BBC published a Research Dept. paper long ago with the
conclusion that linear PCM required 22 bits to capture the dynamics
of an "unconditioned" signal (uncompressed, etc.) from sources of
widest dynamic range (e.g., big band); 20 bits for the actual
dynamic range, and 2 more to allow for possible mis-setting of the
window.

There's some totally unreasonable assumptions here - like that the
big band would be playing in a NC10 room, and that the musicians
themselves contribute only music to the sound level in the room.

They're interested in the limiting cases, not the average.

I have nothing against scientific investigation, but I favor a complete
analysis of the noise floor when people start making claims about what kind
of dynamic range it takes to record and reproduce music.

My method for analyzing that is to look at raw tracks from a recording
session. Obviously recording starts before musicians start playing. The
sound level at that point sets the background noise level. I don't adjust
recording levels while musicians are playing. So, the peak level on the
track, compared to background noise level at the beginning or end of the
recording set the recording's actual dynamic range. On the best day of my
life I see something like 70 dB, and that's with close micing of singers and
acoustical instruments and direct connection to digital electronic
instruments.

A secondary experiment is to bring the noise floor in by stages. I've made
recordings where I turn the mic preamps off, start recording and then turn
on the mic preamps, but the room is still totally empty except me. Then I
record room tone with just me in the room. Then I record room tone during
rehearsal, but when nobody's doing nothing. Then I record room tone with the
powered instruments turned on. etc.

I'm surprised that measurements on the deck of an fully-operational
aircraft carrier deck were left out (and not compared to some empty
chamber in a dry cave 1,000 feet underground)!

Back in the real world, if you ever find music that peaks say 80 dB
over room tone:

(a) be sure you are using ear protection

(b) make a note of it, because it's an outrageously exceptional
condition

Our studio noise floor is about 21 dBA.

And that's with musicians in place, ready to play? Of course not! Musicians
raise the noise floor even higher when they start to play (exclusive of
course, of the sounds made by their musical instruments).

The output of a Musician 2x12 guitar amp is often 110 dB.

The fallacy here is that the guitar amp itself significantly raises the
noise floor in the room when it's turned on, set to go, but the guitarist
isn't playing.

That's not unusual at all in
popular music. (Well, maybe the 21 dB noise floor is a bit
atypical...)

A room's 21 dB noise floor is history, as soon as there are people in the
room. They breath and move around. At 21 dB SPL, just about everything is
pretty noisy.

I agree on the hearing protection.

"LeBaron & Alrich" wrote in message

Jay Kadis wrote:

Our studio noise floor is about 21 dBA. The output of a Musicman
2x12 guitar amp is often 110 dB. That's not unusual at all in
popular music.

But it's a great reason not to be in that room when that amp is
cranking.

Or use ear protection.

However, I've never seen an amp set up to play at 110 dB that was exactly
quiet before the guitarist started playing.

In article "Arny Krueger"
writes:
"LeBaron & Alrich" wrote in message

Jay Kadis wrote:

Our studio noise floor is about 21 dBA. The output of a Musicman
2x12 guitar amp is often 110 dB. That's not unusual at all in
popular music.

But it's a great reason not to be in that room when that amp is
cranking.

Or use ear protection.

However, I've never seen an amp set up to play at 110 dB that was exactly
quiet before the guitarist started playing.


Yeah, but noise gates can alter that. While it's true that guitar amps are
basically compressors, they are not necessarily always in the mix even if
they're on.

-Jay
--
x------- Jay Kadis ------- x---- Jay's Attic Studio ----x
x Lecturer, Audio Engineer x Dexter Records x
x CCRMA, Stanford University x http://www.offbeats.com/ x
x-------- http://ccrma-www.stanford.edu/~jay/ ----------x

"Jay Kadis" wrote in message
...
In article writes:
Jay Kadis wrote:

Our studio noise floor is about 21 dBA. The output of a Musicman 2x12
guitar
amp is often 110 dB. That's not unusual at all in popular music.

But it's a great reason not to be in that room when that amp is
cranking.

--
ha

One of many, actually.

I've been using a nice little Blues Jr. myself. Reasonable levels with
good
tone.

-Jay

--
x------- Jay Kadis ------- x---- Jay's Attic Studio ----x
x Lecturer, Audio Engineer x Dexter Records x
x CCRMA, Stanford University x http://www.offbeats.com/ x
x-------- http://ccrma-www.stanford.edu/~jay/ ----------x

I see that you're at Stanford, are you familiar with Curt Emery in El
Cerrito? He makes a little 8 watt tube head that's great for recording. Very
simple, Class A. Better living through dynamics. I just did a bunch of stuff
with it and was quite pleased.

Darryl

(LeBaron & Alrich) wrote in message ...
Jay Kadis wrote:

Our studio noise floor is about 21 dBA. The output of a Musicman 2x12 guitar
amp is often 110 dB. That's not unusual at all in popular music.

But it's a great reason not to be in that room when that amp is
cranking.

What, due to the presence of popular music?

The BBC knows something about his topic. You're free to disparage their
work
when you have presented data to refute theirs.

I don't disagree with BBC. 20-22db is exactly how much bits is needed
to cover whole dynamic range of effective hearing, from silence to
pain. But when someone as pompous as Boyk comes in with "BBC published
a paper" on that, that's just plain ridiculous and funny. It may be
good enough to impress his students: ... wow this "professeure" of
ours, he reads AES papers, wow, one day we may end being so educate,
like he is. I can't help but LOL.

Not taking any sides here, but is it possible that this BBC paper that was
published on & anon ago was published BEFORE there were 20-bit (or
22-bit/24-bit,etc) convertors? If so, then 14 bits may have been the best
reference they could have come up with at the time.

Just wondering.

NeilH

James Boyk wrote:


Any experimental conclusions in this field will be statistical; for instance, the BBC work concluded that with *source of widest dynamic range and with the mike signal from that source *unconditioned, *22 bits was necessary to conceal *"granularity distortion" for *99% of listeners. Each starred element is relevant and important; for instance, 1% of listeners _could_ still hear the granularity. Those who think that something like this--something statistical and empirical--can be replaced by "a priori" thinking are missing the boat.


What is "granularity distortion"? Is that the same as
quantization noise or does it refer to some other
phenomenon?


Bob
--

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

A. Einstein

On 13 Aug 2003 00:04:44 GMT, (NeilH011) wrote:

Not taking any sides here, but is it possible that this BBC paper that was
published on & anon ago was published BEFORE there were 20-bit (or
22-bit/24-bit,etc) convertors? If so, then 14 bits may have been the best
reference they could have come up with at the time.

Just wondering.

Quite possible, afterall, all knoweledge has to be writen somewhere
for referwnce, but nowdays it's really unnescessarry, if not
unapropriate to call on such an paper. I'm sure the man was posing.
Vladan
www.geocities.com/vla_dan_l
www.mp3.com/lesly , www.mp3.com/shook , www.mp3.com/lesly2
www.kunsttick.com/artists/vuskovic/indexdat.htm

NeilH011 wrote:
...is it possible that this BBC paper ...was published BEFORE there were 20-bit (or
22-bit/24-bit,etc) convertors? If so, then 14 bits may have been the best
reference they could have come up with at the time.

Yes, the paper dates from '77 or so if I remember correctly. But for purposes of the experiment, the researcher was cleverly able to simulate having more bits than he really had. Of course it would be best to do it with true 22-bit or better yet 24-bit converters; unfortunately, there aren't any true 24-bit converters even today, so far as I know.

James Boyk

Buster Mudd wrote:

(LeBaron & Alrich) wrote...
Jay Kadis wrote:

Our studio noise floor is about 21 dBA. The output of a Musicman 2x12
guitar amp is often 110 dB. That's not unusual at all in popular
music.

But it's a great reason not to be in that room when that amp is
cranking.

What, due to the presence of popular music?

SPL does seem to be popular.

--
ha

James Boyk wrote:


Any experimental conclusions in this field will be statistical; for instance, the BBC work concluded that with *source of widest dynamic range and with the mike signal from that source *unconditioned, *22 bits was necessary to conceal *"granularity distortion" for *99% of listeners. Each starred element is relevant and important; for instance, 1% of listeners _could_ still hear the granularity. Those who think that something like this--something statistical and empirical--can be replaced by "a priori" thinking are missing the boat.

If you want to be statistical and empirical you have to define the
following:

1) source of widest dynamic range - how wide is it?
60dB? 130dB? 2000dB?

2) unconditioned - untreated, or any imaginable, or...?

3) conceal - make listeners unaware of? Make the effect inaudiable?
Under what conditions (unconditioned?). Is tested signal played at SPL
equal to source?

4) granularity - what is it?

5) 99% - how many in number?

Please check that "paper" and tell us.

Vladan
www.geocities.com/vla_dan_l
www.mp3.com/lesly , www.mp3.com/shook , www.mp3.com/lesly2
www.kunsttick.com/artists/vuskovic/indexdat.htm

Sorry folks, have lost the start of this thread, and cannot be regained ,
but Scott, you made me spill my cup of coffee !

geoff

Geoff Wood -nospam wrote:
Sorry folks, have lost the start of this thread, and cannot be regained ,
but Scott, you made me spill my cup of coffee !

You shoulda read the LINER NOTES first.
evil chuckle
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

"Scott Dorsey" wrote in message
...
Geoff Wood -nospam wrote:
Sorry folks, have lost the start of this thread, and cannot be regained ,
but Scott, you made me spill my cup of coffee !

You shoulda read the LINER NOTES first.
evil chuckle
--scott


I didn't get any liner notes. Should I sue somebody ?-)

geoff

"Geoff Wood" -nospam wrote:

"Scott Dorsey" wrote:
Geoff Wood -nospam wrote:
Sorry folks, have lost the start of this thread, and cannot be regained ,
but Scott, you made me spill my cup of coffee !

You shoulda read the LINER NOTES first.
evil chuckle
--scott

I didn't get any liner notes. Should I sue somebody ?-)

geoff

The liner notes are the equivilant of a small book; way too much to print and
include with the discs. So, at the very start, we decided we would keep the
liner notes on the Internet. They reside at: http://www.recaudiopro.net . I
believe the notes are also available as a large text file from one of the
members.

Harvey Gerst
Indian Trail Recording Studio
http://www.ITRstudio.com/

"Harvey Gerst" wrote in message I didn't get any
liner notes. Should I sue somebody ?-)

geoff

The liner notes are the equivilant of a small book; way too much to print
and
include with the discs. So, at the very start, we decided we would keep
the
liner notes on the Internet. They reside at: http://www.recaudiopro.net .
I
believe the notes are also available as a large text file from one of the
members.


There *was* a smilie there Harv !

Without printing the 'book' one needs to be diligent to keep up with the
notes, or one's drycleaning bill can get a bit much .....


geoff

On Tue, 26 Aug 2003 07:45:08 +1200, "Geoff Wood"
-nospam wrote:

Without printing the 'book' one needs to be diligent to keep up with the
notes, or one's drycleaning bill can get a bit much .....

Knight: "That rabbit's dynamite."

King: "Go and change your armor."


Chris Hornbeck