[Adrian Tuddenham[_2_]]

The Victor recordings of Jelly-Roll-Morton made in the ballroom of the
Webster Hotel, Chicago, in September and December of 1926 have
incredibly good balance between the instruments.

In particular, the trombone is heard as an accompanying instrument; yet
a trombone can easily be 10dB louder than the rest of the band put
together. The unmuted trumpet and the clarinet sound equally loud.

With a single mic recording (which this almost certainly was), the
balance is normally obtained by altering the distance of the various
instruments from the mic. In the case of the trumpet versus the
clarinet, this seems to be the case (and the trumpet plays side-on to
the mic when unmuted) but the trombone does not sound as far distant in
the acoustics of the room as its level would suggest.

The mic was probably the nominally omnidirectional Wente/Thuras design
which went with the Western Electric recording kit, so there was no dead
side to it.

How have they managed to balance the trombone and still keep it sounding
as if it is in the right perspective?



--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Scott Dorsey]

Adrian Tuddenham wrote:
The Victor recordings of Jelly-Roll-Morton made in the ballroom of the
Webster Hotel, Chicago, in September and December of 1926 have
incredibly good balance between the instruments.

In particular, the trombone is heard as an accompanying instrument; yet
a trombone can easily be 10dB louder than the rest of the band put
together. The unmuted trumpet and the clarinet sound equally loud.

What you need to know about acoustic recordings:

1. Everybody was grouped around one big horn. Some instruments were
put far away from the horn, some were stuck right up inside it.

2. A lot of instruments were modified in order to be louder. For
example, it was common to use fiddles with horns attached to the
F-holes in order to amplify them somewhat. Woodwinds would often have
larger bells, sometimes with an angle to point them directly out.
Sometimes woodwind parts would actually be played on other instruments
(clarinets replaced by soprano sax, etc) in order to bring the levels up.

3. Everything was massively slew-limited by the inability of the diaphragm
in the horn to keep up. This brings down the level of peaky instruments
like trumpets a lot.

With a single mic recording (which this almost certainly was), the
balance is normally obtained by altering the distance of the various
instruments from the mic. In the case of the trumpet versus the
clarinet, this seems to be the case (and the trumpet plays side-on to
the mic when unmuted) but the trombone does not sound as far distant in
the acoustics of the room as its level would suggest.

I am assuming this was an acoustic recording, and in the acoustic recording
world everything centered around getting everything as loud as possible.
Since the total volume in the studio reflects the actual needle excursion,
if there was a decision to be made about balances it would be made by making
the soft instrument louder whenever possible, regardless of tone.

The mic was probably the nominally omnidirectional Wente/Thuras design
which went with the Western Electric recording kit, so there was no dead
side to it.

How have they managed to balance the trombone and still keep it sounding
as if it is in the right perspective?

By 1926, it could well have been an electric recording made with an omni
microphone. Even if this was the case, the acoustic recording techniques
were pretty entrenched and I would not be surprised if they still applied.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Adrian Tuddenham[_2_]]

Scott Dorsey wrote:

Adrian Tuddenham wrote:
The Victor recordings of Jelly-Roll-Morton made in the ballroom of the
Webster Hotel, Chicago, in September and December of 1926 have
incredibly good balance between the instruments.

In particular, the trombone is heard as an accompanying instrument; yet
a trombone can easily be 10dB louder than the rest of the band put
together. The unmuted trumpet and the clarinet sound equally loud.

What you need to know about acoustic recordings:
[..]

This one wasn't acoustic by a long way.

By 1926, it could well have been an electric recording made with an omni
microphone.

It was definitely electric and almost certainly made with the
Wente/Thuras 'omni' mic copled to the WE moving-iron (rubber line
damped) lateral cutterhead.

Even if this was the case, the acoustic recording techniques
were pretty entrenched and I would not be surprised if they still applied.

Not by that stage. I can only think of one electrically-recorded 78
with an acoustic layout, that was a German Brunswick of Paul Godwin's
Orchestra playing "Dolly's Dancing".

All the Victor/HMV/Columbia commercial issues took immediate advantage
of the freedom of layout which electrical recording allowed. By
September 1926, acoustic recording techniques were dead and buried.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Mike Rivers]

Adrian Tuddenham wrote:

With a single mic recording (which this almost certainly was), the
balance is normally obtained by altering the distance of the various
instruments from the mic. In the case of the trumpet versus the
clarinet, this seems to be the case (and the trumpet plays side-on to
the mic when unmuted) but the trombone does not sound as far distant in
the acoustics of the room as its level would suggest.

The technician (they didn't have "engineers" in 1926 said "Hey, you with
the trombone! Don't play so loud."

If it's a fairly large room, which it probably was, he can step back a
few steps (which is about all it would take) without sounding distant.
Remember, nobody is really very close to the microphone (or horn) except
quiet things. And while I don't think I've heard the recording you're
asking about, I've heard some mighty loud clarinets.

How have they managed to balance the trombone and still keep it sounding
as if it is in the right perspective?

Good players who listen to each other.

--
If you e-mail me and it bounces, use your secret decoder ring and reach
me he
double-m-eleven-double-zero at yahoo -- I'm really Mike Rivers
)

[Laurence Payne[_2_]]

On Sat, 15 Nov 2008 14:54:50 +0000,
lid (Adrian Tuddenham) wrote:

The Victor recordings of Jelly-Roll-Morton made in the ballroom of the
Webster Hotel, Chicago, in September and December of 1926 have
incredibly good balance between the instruments.

In particular, the trombone is heard as an accompanying instrument; yet
a trombone can easily be 10dB louder than the rest of the band put
together. The unmuted trumpet and the clarinet sound equally loud.

With a single mic recording (which this almost certainly was), the
balance is normally obtained by altering the distance of the various
instruments from the mic. In the case of the trumpet versus the
clarinet, this seems to be the case (and the trumpet plays side-on to
the mic when unmuted) but the trombone does not sound as far distant in
the acoustics of the room as its level would suggest.

The mic was probably the nominally omnidirectional Wente/Thuras design
which went with the Western Electric recording kit, so there was no dead
side to it.

How have they managed to balance the trombone and still keep it sounding
as if it is in the right perspective?


Perhaps the musicians played well and someone put a mic in front of
it.

You're making the rather insulting assumption that no musical
performance could possibly be balanced unless an engineer comes in and
messes with it. Whereas, of course, exactly the opposite is the case!

[Adrian Tuddenham[_2_]]

Laurence Payne wrote:

On Sat, 15 Nov 2008 14:54:50 +0000,
lid (Adrian Tuddenham) wrote:

The Victor recordings of Jelly-Roll-Morton made in the ballroom of the
Webster Hotel, Chicago, in September and December of 1926 have
incredibly good balance between the instruments.

In particular, the trombone is heard as an accompanying instrument; yet
a trombone can easily be 10dB louder than the rest of the band put
together. The unmuted trumpet and the clarinet sound equally loud.

With a single mic recording (which this almost certainly was), the
balance is normally obtained by altering the distance of the various
instruments from the mic. In the case of the trumpet versus the
clarinet, this seems to be the case (and the trumpet plays side-on to
the mic when unmuted) but the trombone does not sound as far distant in
the acoustics of the room as its level would suggest.

The mic was probably the nominally omnidirectional Wente/Thuras design
which went with the Western Electric recording kit, so there was no dead
side to it.

How have they managed to balance the trombone and still keep it sounding
as if it is in the right perspective?


Perhaps the musicians played well and someone put a mic in front of
it.

That was certainly the case; but there are signs that this excellent
performance was additionally 'choreographed' in some way for the
recording.

You're making the rather insulting assumption that no musical
performance could possibly be balanced unless an engineer comes in and
messes with it. Whereas, of course, exactly the opposite is the case!

They were obviously a good band and well balanced internally; but the
small dynamic range of the recording equipment, without the help (?) of
compressors or limiters, meant that some additional balancing had to be
done by the recording engineer. Also, these recordings have a different
balance from later recordings of the same band.

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Adrian Tuddenham[_2_]]

Mike Rivers wrote:

Adrian Tuddenham wrote:

With a single mic recording (which this almost certainly was), the
balance is normally obtained by altering the distance of the various
instruments from the mic. In the case of the trumpet versus the
clarinet, this seems to be the case (and the trumpet plays side-on to
the mic when unmuted) but the trombone does not sound as far distant in
the acoustics of the room as its level would suggest.

The technician (they didn't have "engineers" in 1926 said "Hey, you with
the trombone! Don't play so loud."

This is a *jazz* band :-)


If it's a fairly large room, which it probably was, he can step back a
few steps (which is about all it would take) without sounding distant.
Remember, nobody is really very close to the microphone (or horn) except
quiet things. And while I don't think I've heard the recording you're
asking about, I've heard some mighty loud clarinets.

I have to record a band with a similar line-up and they want to achieve
a similar balance. In their natural habitat (a grotty pub with minimal
soft furnishings but a packed audience), the clarinet was peaking 6 dB
below the trumpet and the trombone was peaking more than 10 dB above it.
(Actually measured as 106 dBC at 15 ft - and that was the 'quieter'
stand-in trombonist, I have to make the recording with their usual one).

As far as possible I want to stick to single-mic technique; probably
using coincident crossed ribbons in stereo and a large, but acoustically
dead, hall.


How have they managed to balance the trombone and still keep it sounding
as if it is in the right perspective?

Good players who listen to each other.

These are they ... but there is a limit to how much I can ask an
exuberant trombonist to keep the noise down.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Laurence Payne[_2_]]

On Sat, 15 Nov 2008 17:30:10 +0000,
lid (Adrian Tuddenham) wrote:

The technician (they didn't have "engineers" in 1926 said "Hey, you with
the trombone! Don't play so loud."

This is a *jazz* band :-)

Sure. And a good one. They balanced, just like all well-played and
well-arranged acoustic performances always have and always will.
What's blocking you from this obvious fact?

[Scott Dorsey]

Adrian Tuddenham wrote:

I have to record a band with a similar line-up and they want to achieve
a similar balance. In their natural habitat (a grotty pub with minimal
soft furnishings but a packed audience), the clarinet was peaking 6 dB
below the trumpet and the trombone was peaking more than 10 dB above it.
(Actually measured as 106 dBC at 15 ft - and that was the 'quieter'
stand-in trombonist, I have to make the recording with their usual one).

The audience makes it much harder. I'd suggest putting the trombone
player behind a gobo, but you can't do that in front of an audience.
You can have the trombone player turn around and face the rear wall
or play into the piano, though. And you can certainly move the clarinet
very far forward so it is closer to the mike pair.

But what you need to realize here is that the PEAK level has nothing to
do with the perceived loudness. Brass instruments will have outrageously
high peak levels even when they are played quietly, because they have a
very sharp peaky waveform. You can crush it down with limiting, or you
can just live with it.

Balance everything so that the instruments sound balanced in the monitors,
and don't worry about the meters except to make sure they don't hit the
pegs. Don't try and mix with the meters. If you have to turn everything
way down in order to deal with the high peaks, go right ahead and do it.
You can use fast limiting after the fact to squash the peaks and you won't
hear much difference in sound when you do because of the nature of brass.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Adrian Tuddenham[_2_]]

Laurence Payne wrote:

On Sat, 15 Nov 2008 17:30:10 +0000,
lid (Adrian Tuddenham) wrote:

The technician (they didn't have "engineers" in 1926 said "Hey, you with
the trombone! Don't play so loud."

This is a *jazz* band :-)

Sure. And a good one. They balanced, just like all well-played and
well-arranged acoustic performances always have and always will.
What's blocking you from this obvious fact?

I've never heard a trombone played as quietly as that one sounded.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Adrian Tuddenham[_2_]]

Scott Dorsey wrote:

Adrian Tuddenham wrote:

I have to record a band with a similar line-up and they want to achieve
a similar balance. In their natural habitat (a grotty pub with minimal
soft furnishings but a packed audience), the clarinet was peaking 6 dB
below the trumpet and the trombone was peaking more than 10 dB above it.
(Actually measured as 106 dBC at 15 ft - and that was the 'quieter'
stand-in trombonist, I have to make the recording with their usual one).

The audience makes it much harder. I'd suggest putting the trombone
player behind a gobo, but you can't do that in front of an audience.

We won't have an audience at the recording session, so that is something
I could try.

You can have the trombone player turn around and face the rear wall

I had originally intended to do it that way; but the back wall of the
hall is 50ft away and not acoustically treated, so I was worried that
the sound would be distant. Perhaps with enough blankets, I could damp
the wall down or build a barrier and make that idea work.

With the mute on and when playing the lead, the trombone ought to face
the mic so as to give a clearer sound; the same applies to the trumpet.
When playing as lead instruments, the higher levels will be fine.

or play into the piano, though.

Not practical in this case because the location of the piano is fixed on
a stage with little room for any other instruments. Also, I want the
piano central and the trombone on the right hand side of the stereo
image.

And you can certainly move the clarinet
very far forward so it is closer to the mike pair.

That I will do.


But what you need to realize here is that the PEAK level has nothing to
do with the perceived loudness. Brass instruments will have outrageously
high peak levels even when they are played quietly, because they have a
very sharp peaky waveform. You can crush it down with limiting, or you
can just live with it.

The level I quoted was measured with an SPL meter, so it was the average
of the peaky waveform but the 'peak' (i.e. maximum) loudness of the
performance ...if you see what I mean.

I wouldn't consider a limiter because :

a) They weren't used in 1926
b) The customer doesn't want too 'pushy' a sound
c) I haven't got one.


Balance everything so that the instruments sound balanced in the monitors,
and don't worry about the meters except to make sure they don't hit the
pegs.

That's my usual technique, but I do find that a BBC-type PPM agrees with
my ears much more than a VU meter does.

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[hank alrich]

Adrian Tuddenham wrote:

I wouldn't consider a limiter because :

a) They weren't used in 1926

The transient response of the audio chain in 1926 was not what it is
today. Peaks were relatively limited compared to what is possible not
with 24 bit systems.

b) The customer doesn't want too 'pushy' a sound

Careful application of peak limiting can be undetectable, IME.

c) I haven't got one.

Not even in software, that you could apply in postproduction?

--
ha
shut up and play your guitar

[Adrian Tuddenham[_2_]]

hank alrich wrote:

Adrian Tuddenham wrote:

I wouldn't consider a limiter because :

a) They weren't used in 1926

The transient response of the audio chain in 1926 was not what it is
today. Peaks were relatively limited compared to what is possible not
with 24 bit systems.

The problem with a moving iron recorder system was that the response
could be very non-linear on peaks due to the inverse-square effect of
the air gaps in the magnetic circuit. Peaks became peakier.

Columbia discovered this when they started recording Japanese music with
a W.E. cutterhead. That was one of the reasons why the Blumlein
moving-coil cutter sounded a lot better - in effect, it had a virtually
unlimited undistorted transient response (over a much larger amplitude
than the grooves could handle).

b) The customer doesn't want too 'pushy' a sound

Careful application of peak limiting can be undetectable, IME.

....but if the customer doesn't want it to sound loud, what is the point
of a limiter? It is better to just keep the level down.

By the way, this outfit doesn't include percussion, so the peaks come
entirely from non-transient waveforms.


c) I haven't got one.

Not even in software, that you could apply in postproduction?

Nope. I've managed to get along without one so far.

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[hank alrich]

Adrian Tuddenham wrote:

hank alrich wrote:

Careful application of peak limiting can be undetectable, IME.

...but if the customer doesn't want it to sound loud, what is the point
of a limiter? It is better to just keep the level down.

One can employ peak limiting to make it smoother, not louder. It's just
a matter of choice how one uses the tool.

Peak limiting can be used to establish a more balanced sound in a
recording where some elements have extraordinary peaks relative to the
rest of the ensemble.

By the way, this outfit doesn't include percussion, so the peaks come
entirely from non-transient waveforms.

I'm not sure how a true peak can derive from non-transient waveforms.
Don't all dynamic waveforms have some rise time? The spikes from brass
are part of their signature, and look like peaks when one views them as
a waveform in the DAW.

--
ha
shut up and play your guitar

[Mike Rivers]

Adrian Tuddenham wrote:

I have to record a band with a similar line-up and they want to achieve
a similar balance. In their natural habitat (a grotty pub with minimal
soft furnishings but a packed audience), the clarinet was peaking 6 dB
below the trumpet and the trombone was peaking more than 10 dB above it.

That's not very close to a studio. You'll just have to say "Hey you with
the trombone! Don't play so loud." This isn't 1926 any more.

... but there is a limit to how much I can ask an
exuberant trombonist to keep the noise down.

Something's gotta give. Advances in microphones since 1926 haven't
helped musicians to balance themselves. Have they considered moving out
of the grotty pub and into a studio. They can bring in an audience. A
friend of mine with a studio in Berkeley CA has done that with jazz
bands. He has a big room that musicians like to play in. He'll put spot
mics on the bass and piano to help with balance, but the main pickup
will be a stereo pair. I think mostly he uses a Neumann SM69.


--
If you e-mail me and it bounces, use your secret decoder ring and reach
me he
double-m-eleven-double-zero at yahoo -- I'm really Mike Rivers
)

[Mike Rivers]

Adrian Tuddenham wrote:

...but if the customer doesn't want it to sound loud, what is the point
of a limiter? It is better to just keep the level down.

That's not the purpose of a limiter. The limiter will lower the peaks
that aren't really long enough to hear but will screw up your digital
recording.


--
If you e-mail me and it bounces, use your secret decoder ring and reach
me he
double-m-eleven-double-zero at yahoo -- I'm really Mike Rivers
)

[Mike Rivers]

Adrian Tuddenham wrote:

Perhaps the musicians played well and someone put a mic in front of
it.

That was certainly the case; but there are signs that this excellent
performance was additionally 'choreographed' in some way for the
recording.

This is part of sounding like an old time band. It's the latest thing
for bluegrass bands. It's OK for the clarinet player to take a step
closer to the mic when he's playing a solo, or for the trombone player
to step back when he's riffing. You can (and should) spend some time
with them in rehearsal and mark the floor where they're supposed to
stand. If they want to do it like old time, they have to party like 1926.

They were obviously a good band and well balanced internally; but the
small dynamic range of the recording equipment, without the help (?) of
compressors or limiters, meant that some additional balancing had to be
done by the recording engineer.

If it's dynamic balancing, it has to be done by the players, not the
engineer.

Also, these recordings have a different
balance from later recordings of the same band.

Nostalgia is never what it used to be.




--
If you e-mail me and it bounces, use your secret decoder ring and reach
me he
double-m-eleven-double-zero at yahoo -- I'm really Mike Rivers
)

[Adrian Tuddenham[_2_]]

Mike Rivers wrote:

Adrian Tuddenham wrote:

... In their natural habitat (a grotty pub with minimal
soft furnishings but a packed audience), the clarinet was peaking 6 dB
below the trumpet and the trombone was peaking more than 10 dB above it.

That's not very close to a studio. ...

...Have they considered moving out
of the grotty pub and into a studio.

Yes, we abandoned the idea of recording in the pub at an early stage;
their pianist complained about the (lack of) quality of the piano. I am
hiring a village hall with a superb piano and good acoustics (I recorded
a choir and instrumental trio there a couple of years ago). I'll take
along a few washing lines full of blankets and some pieces of carpet,
just in case we need to deaden it even more.

...They can bring in an audience

They prefer not to have an audience - and we can only get the hall and
the band together on a Sunday from 09:00 until 17:00, so very few people
would turn up even if they were able to find the village. There are
also all sorts of local regulations about events open to the public.

A
friend of mine with a studio in Berkeley CA has done that with jazz
bands. He has a big room that musicians like to play in. He'll put spot
mics on the bass and piano to help with balance, but the main pickup
will be a stereo pair. I think mostly he uses a Neumann SM69.

They want this to sound like an old recording (except in stereo and
without the surface noise), so I am going to try to do it with a single
coincident stereo pair of pseudo-ribbons.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Adrian Tuddenham[_2_]]

Mike Rivers wrote:

Adrian Tuddenham wrote:

Perhaps the musicians played well and someone put a mic in front of
it.

That was certainly the case; but there are signs that this excellent
performance was additionally 'choreographed' in some way for the
recording.

This is part of sounding like an old time band. It's the latest thing
for bluegrass bands. It's OK for the clarinet player to take a step
closer to the mic when he's playing a solo, or for the trombone player
to step back when he's riffing. You can (and should) spend some time
with them in rehearsal and mark the floor where they're supposed to
stand. If they want to do it like old time, they have to party like 1926.

That's exactly what they want to do.

I had even considered taking along some cardboard markers to place on
the floor and some targets to hang on the walls at the points that I
want them to aim at. Another useful trick is to form a barrier of
chairs to keep the players away from the mic.

They were obviously a good band and well balanced internally; but the
small dynamic range of the recording equipment, without the help (?) of
compressors or limiters, meant that some additional balancing had to be
done by the recording engineer.

If it's dynamic balancing, it has to be done by the players, not the
engineer.

Sorry, I meant that the band did it under the direction of the recording
engineer - they weren't just left to their own devices. There isn't
much 'balancing, in the modern sense, that the engineer could do with
just one omni mic - unless he is prepared to wave it around during the
take. (Yes, I have done that, when I had to record a six-piece folk
group with just an interview mic and a mono Uher. It worked very well)

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Adrian Tuddenham[_2_]]

Mike Rivers wrote:

Adrian Tuddenham wrote:

...but if the customer doesn't want it to sound loud, what is the point
of a limiter? It is better to just keep the level down.

That's not the purpose of a limiter. The limiter will lower the peaks
that aren't really long enough to hear but will screw up your digital
recording.

I normally have a BBC-type PPM set so that '4' = 0dBm = -12 dBFS and try
to keep the highest kicks of the meter below '5'. That should give me 8
dB headroom above any peak of 2 milliseconds or longer.

The scale on the DAT recorder will tell me if anything shorter than 2
milliseconds is causing overloading.

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Adrian Tuddenham[_2_]]

hank alrich wrote:

Adrian Tuddenham wrote:

hank alrich wrote:

I'm not sure how a true peak can derive from non-transient waveforms.
Don't all dynamic waveforms have some rise time? The spikes from brass
are part of their signature, and look like peaks when one views them as
a waveform in the DAW.

Perhaps I used the wrong terminology. I meant that I don't have to cope
with all the crashes and bangs of a drumkit, which are never the same
twice. The peaks in a trombone waveform (for instance) are a little
more repetitive and predictable; and the attack doesn't contain a huge
energy spike.



--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[hank alrich]

Adrian Tuddenham wrote:

hank alrich wrote:

Adrian Tuddenham wrote:

hank alrich wrote:

I'm not sure how a true peak can derive from non-transient waveforms.
Don't all dynamic waveforms have some rise time? The spikes from brass
are part of their signature, and look like peaks when one views them as
a waveform in the DAW.

Perhaps I used the wrong terminology. I meant that I don't have to cope
with all the crashes and bangs of a drumkit, which are never the same
twice. The peaks in a trombone waveform (for instance) are a little
more repetitive and predictable; and the attack doesn't contain a huge
energy spike.

Thanks, Adrian. That makes sense.

--
ha
shut up and play your guitar

[Scott Dorsey]

Adrian Tuddenham wrote:
Scott Dorsey wrote:
But what you need to realize here is that the PEAK level has nothing to
do with the perceived loudness. Brass instruments will have outrageously
high peak levels even when they are played quietly, because they have a
very sharp peaky waveform. You can crush it down with limiting, or you
can just live with it.

The level I quoted was measured with an SPL meter, so it was the average
of the peaky waveform but the 'peak' (i.e. maximum) loudness of the
performance ...if you see what I mean.

Ahh.. SPL meters aren't peak-reating meters at all. You don't want to
use the word 'peak' here.

I wouldn't consider a limiter because :

a) They weren't used in 1926
b) The customer doesn't want too 'pushy' a sound
c) I haven't got one.

In 1926, you had plenty of limiting built into the cutting head itself,
unfortunately. That didn't change until feedback cutters came in.

Limiting that is done carefully and cautiously won't change the sound
in any audible way. If you can tell it's kicking in, it's too much.
Limiting gets a bad reputation because it's abused but you don't have
to abuse it.

Balance everything so that the instruments sound balanced in the monitors,
and don't worry about the meters except to make sure they don't hit the
pegs.

That's my usual technique, but I do find that a BBC-type PPM agrees with
my ears much more than a VU meter does.

I'd generally agree with you on that, but brass is funny because it is just
SO spiky. For many years, tape limiting was the solution for brass issues.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Adrian Tuddenham[_2_]]

Scott Dorsey wrote:

Adrian Tuddenham wrote:
Scott Dorsey wrote:

[...]
In 1926, you had plenty of limiting built into the cutting head itself,
unfortunately. That didn't change until feedback cutters came in.

It depends on the cutterhead:

The Western Electric moving iron cutterhead overloaded in a
positive-feedback sense. The sensitivity was determined by the air gap
between the armature and the pole pieces. The force on the armature
obeyed the inverse-square law with relation to displacement (from the
central resting position) wheres the returning force of the springs was
more-or-less linear. This meant that a sinusoidal waveform became
peakier as the amplitude increased; the distortion was mainly third
harmonic.

The Holman moving iron cutterhead was similarly non-linear but in a
unidirectional way, so it generated mainly second-harmonic distortion.

The moving coil cutterhead (Blumlein or Voigt) could cut many times the
available groove amplitude before overloading set in. The limitation
was in the amplifier at high frequencies and in whether the record would
suffer inter-groove cutting or fail the wear test at lower frequencies.

The amplifiers used with the W.E. and Blumlein heads were generously
rated, so they did not begin overloading at the power levels used in
normal recordings. Much later, amateur and semi-pro set-ups used
smaller amplifiers which could suffer from peak-crushing overload
distortion - and they did give a limiting effect.

Listening to 78s from around 1926, there is no sign of limiting. On
worn copies, the waveform with the highest acceleration usually becomes
worn preferentially and this can often sound like bad peak clipping.
Sometime the wear is greater at certain frequencies than at others; this
corresponds to the high mechanical impedance of some previous acoustic
soundbox at its many resonances.


Balance everything so that the instruments sound balanced in the monitors,
and don't worry about the meters except to make sure they don't hit the
pegs.

That's my usual technique, but I do find that a BBC-type PPM agrees with
my ears much more than a VU meter does.

I'd generally agree with you on that, but brass is funny because it is just
SO spiky. For many years, tape limiting was the solution for brass issues.

....but the Ferrographs are behind the settee and they are so heavy to
drag out :-)

I'll use a combination of ears, PPM and the digital peak-reading
barmeter of the DAT machine to keep a check on things. (It might be
worth taking an analogue oscilloscope along too.)

We intend playing-back the morning's takes during the lunch break, so
any hidden problems should show up then.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Laurence Payne[_2_]]

On Sun, 16 Nov 2008 08:47:32 +0000,
lid (Adrian Tuddenham) wrote:

We intend playing-back the morning's takes during the lunch break, so
any hidden problems should show up then.

They'll be much more interested in musical issues. Woe betide you if
you need to veto the best performance 'cos you didn't get the
technicals right!

I suggest you play them back a take as early as possible in the day.
If you get approval of the balance, leave things there and let them
concentrate on getting their best performance. If it isn't good, fix
it. But do it quickly - the project is about their music, not about
your learning curve! You can't spend all day getting your side right
then say "OK lads, now give me your best one!"

Go for belt-and-braces wherever possible. You may be convinced a
mid-field stereo pair is the answer. Fine - but if you have enough
channels record individual mics as well.

[Scott Dorsey]

Adrian Tuddenham wrote:

The Western Electric moving iron cutterhead overloaded in a
positive-feedback sense. The sensitivity was determined by the air gap
between the armature and the pole pieces. The force on the armature
obeyed the inverse-square law with relation to displacement (from the
central resting position) wheres the returning force of the springs was
more-or-less linear. This meant that a sinusoidal waveform became
peakier as the amplitude increased; the distortion was mainly third
harmonic.

The Holman moving iron cutterhead was similarly non-linear but in a
unidirectional way, so it generated mainly second-harmonic distortion.

The moving coil cutterhead (Blumlein or Voigt) could cut many times the
available groove amplitude before overloading set in. The limitation
was in the amplifier at high frequencies and in whether the record would
suffer inter-groove cutting or fail the wear test at lower frequencies.

The amplifiers used with the W.E. and Blumlein heads were generously
rated, so they did not begin overloading at the power levels used in
normal recordings. Much later, amateur and semi-pro set-ups used
smaller amplifiers which could suffer from peak-crushing overload
distortion - and they did give a limiting effect.

All of this is true... but what you have to realize is that mild limiting
on very tall peaks is _not_ audible as distortion. Even the moving coil
heads had (and still have) mass... you can get quite effective limiting on
short small peaks from a modern Neumann head.

Listening to 78s from around 1926, there is no sign of limiting. On
worn copies, the waveform with the highest acceleration usually becomes
worn preferentially and this can often sound like bad peak clipping.
Sometime the wear is greater at certain frequencies than at others; this
corresponds to the high mechanical impedance of some previous acoustic
soundbox at its many resonances.

I'm not talking about over the top limiting that will be audible, I am talking
about gentle limiting on sharp peaks, the kind of thing that I think is
not going to show up except on an A/B test with the original.

As I said earlier, you can limit brass waveforms a LOT before they even
start to be noticed, let alone become unpleasant. Hell, take a look at
the grooves on a Herb Alpert record... that's a very close-miked horn and
the grooves are almost symmetric. That doesn't happen naturally.

After you have made your recording, take the digital file up on a workstation
and see what very moderate and light limiting will do for you. I think you
will be surprised how much you can control the brass/everything else balance
with very few side-effects.
--scott

--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Adrian Tuddenham[_2_]]

Scott Dorsey wrote:

Adrian Tuddenham wrote:

The Western Electric moving iron cutterhead overloaded in a
positive-feedback sense. The sensitivity was determined by the air gap
between the armature and the pole pieces. The force on the armature
obeyed the inverse-square law with relation to displacement (from the
central resting position) wheres the returning force of the springs was
more-or-less linear. This meant that a sinusoidal waveform became
peakier as the amplitude increased; the distortion was mainly third
harmonic.

The Holman moving iron cutterhead was similarly non-linear but in a
unidirectional way, so it generated mainly second-harmonic distortion.

The moving coil cutterhead (Blumlein or Voigt) could cut many times the
available groove amplitude before overloading set in. The limitation
was in the amplifier at high frequencies and in whether the record would
suffer inter-groove cutting or fail the wear test at lower frequencies.

The amplifiers used with the W.E. and Blumlein heads were generously
rated, so they did not begin overloading at the power levels used in
normal recordings. Much later, amateur and semi-pro set-ups used
smaller amplifiers which could suffer from peak-crushing overload
distortion - and they did give a limiting effect.

All of this is true... but what you have to realize is that mild limiting
on very tall peaks is _not_ audible as distortion.

That is interesting.

Even the moving coil
heads had (and still have) mass... you can get quite effective limiting on
short small peaks from a modern Neumann head.

I find that difficult to believe if you mean peaks in amplitude terms.
If the amplifiers are doing their job correctly, they will simply create
the force necessary to move the mass at whatever speed and to whatever
position the waveform demands.

If the amplifier output is being limited, either by intentional limiting
or by simply running out of power, the interaction with the cutterhead
mass will give a slew-rate limiting effect, not an amplitude one. A
loss of amplitude will result from slew-rate limiting, but it will be a
different kind of distortion from 'flat-topping'.

Is it possible that slew-rate limiting is the kind of distortion which
is inaudible on brass? It will still create intermodulation (as will
any kind of distortion) but the most prominent spurious products may
well coincide with the natural harmonics of the instrument and therefore
be inaudible.


Listening to 78s from around 1926, there is no sign of limiting. On
worn copies, the waveform with the highest acceleration usually becomes
worn preferentially and this can often sound like bad peak clipping.
Sometime the wear is greater at certain frequencies than at others; this
corresponds to the high mechanical impedance of some previous acoustic
soundbox at its many resonances.

I'm not talking about over the top limiting that will be audible, I am talking
about gentle limiting on sharp peaks, the kind of thing that I think is
not going to show up except on an A/B test with the original.

There is no limiting mechanism that I know of in the W.E. cutterhead
chain for signals within the working range of the microphone and
amplifier. They covered frequencies from 40c/s to 6 Kc/s and could cope
with amplitudes in excess of anything they dared cut onto the disc.
There is no audible evidence of slew-rate or amplitude limiting;
visually the tips of the waveform appear rounded, but this is more
likely to be due to the absence of high frequencies than to any form of
limiting (it occurs at all amplitude levels).

Perhaps with wider bandwidths in modern equipment there would be high
frequency peaks coming through which will cause the overloading effects
you are concerned about, but I need to deal with these at source by
acoustic means if I am to get the kind of sound the customer wants.
Pointing the trumpet and trombone sideways-on to the mic is a good
starting point, provided I can find a way of preventing them from then
sounding too distant in the perspective of the mix.


As I said earlier, you can limit brass waveforms a LOT before they even
start to be noticed, let alone become unpleasant. Hell, take a look at
the grooves on a Herb Alpert record... that's a very close-miked horn and
the grooves are almost symmetric. That doesn't happen naturally.

The customer doesn't want a 'modern' sounding record. If the brass
overloads when it is recorded, I'll take the whole level down. They
could do it in 1926 with a 30 dB S/N ratio, I ought to be able to do it
in 2008 with an 80 dB S/N ratio.


After you have made your recording, take the digital file up on a workstation
and see what very moderate and light limiting will do for you. I think you
will be surprised how much you can control the brass/everything else balance
with very few side-effects.

Good advice for a modern recordings, but I need to get the balance
correct at source for this one, not in post production.


--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Adrian Tuddenham[_2_]]

Laurence Payne wrote:

On Sun, 16 Nov 2008 08:47:32 +0000,
lid (Adrian Tuddenham) wrote:

We intend playing-back the morning's takes during the lunch break, so
any hidden problems should show up then.

They'll be much more interested in musical issues. Woe betide you if
you need to veto the best performance 'cos you didn't get the
technicals right!

I suggest you play them back a take as early as possible in the day.
If you get approval of the balance, leave things there and let them
concentrate on getting their best performance. If it isn't good, fix
it. But do it quickly - the project is about their music, not about
your learning curve! You can't spend all day getting your side right
then say "OK lads, now give me your best one!"

Excellent advice.

Because this is to be a single-mic recording, an awful lot will depend
on the musicians 'choreographing' the performance. They asked for it to
be done this way, so they realise that this is going to be a bit of an
experiment.

They are all mature and experienced musicians and will be playing their
standard repertoire under considerably better conditions than their
usual venue full of noisy punters, so they won't be struggling with the
musical side of things too much. I have allowed 8 hours of hall hire
time to complete a one-hour CD, so we shouldn't be under too much time
pressure either.


Go for belt-and-braces wherever possible. You may be convinced a
mid-field stereo pair is the answer. Fine - but if you have enough
channels record individual mics as well.

I only have two channels on a DAT recorder.

--
~ Adrian Tuddenham ~
(Remove the ".invalid"s and add ".co.uk" to reply)
www.poppyrecords.co.uk

[Scott Dorsey]

Adrian Tuddenham wrote:
That is interesting.

Even the moving coil
heads had (and still have) mass... you can get quite effective limiting on
short small peaks from a modern Neumann head.

I find that difficult to believe if you mean peaks in amplitude terms.
If the amplifiers are doing their job correctly, they will simply create
the force necessary to move the mass at whatever speed and to whatever
position the waveform demands.

Take a look at the waveform of the trumpet, for instance. There is a nice
even little waveform, and then there are these MONSTROUS periodic peaks
that jump 20 to 50 times higher than the rest of the waveform. They also
contain a whole lot of ultrasonic stuff because of the fast risetime.

Yes, it's true that given an infinite amount of amplifier bandwidth,
power and resistanceless coils that never heated up, you could make
the cutter head coil follow the input waveform perfectly, but in
reality even helium cooling won't let you follow those huge sharp peaks.

If the amplifier output is being limited, either by intentional limiting
or by simply running out of power, the interaction with the cutterhead
mass will give a slew-rate limiting effect, not an amplitude one. A
loss of amplitude will result from slew-rate limiting, but it will be a
different kind of distortion from 'flat-topping'.

Yes, the main effect IS slew-rate limiting.

Is it possible that slew-rate limiting is the kind of distortion which
is inaudible on brass? It will still create intermodulation (as will
any kind of distortion) but the most prominent spurious products may
well coincide with the natural harmonics of the instrument and therefore
be inaudible.

I don't know, that is entirely possible. But even a conventional limiter
can be cranked down very far on brass without changing the tone of the
instrument very much. This can be a very powerful and useful tool at times.

There is no limiting mechanism that I know of in the W.E. cutterhead
chain for signals within the working range of the microphone and
amplifier. They covered frequencies from 40c/s to 6 Kc/s and could cope
with amplitudes in excess of anything they dared cut onto the disc.
There is no audible evidence of slew-rate or amplitude limiting;
visually the tips of the waveform appear rounded, but this is more
likely to be due to the absence of high frequencies than to any form of
limiting (it occurs at all amplitude levels).

The waveform we are talking about is VERY spiky. There is a lot of sharp
stuff well above 6 KC... there is actually a lot of stuff well above 20 KC
that we would be better off doing without, too.

Perhaps with wider bandwidths in modern equipment there would be high
frequency peaks coming through which will cause the overloading effects
you are concerned about, but I need to deal with these at source by
acoustic means if I am to get the kind of sound the customer wants.
Pointing the trumpet and trombone sideways-on to the mic is a good
starting point, provided I can find a way of preventing them from then
sounding too distant in the perspective of the mix.

The good news with the trumpet and trombone is that, unlike with a lot of
instruments, most of the sound comes out of the bell. So if you stick
something in front of the bell (like a big gobo made from an office
partition), you can reduce sound evenly in all directions, including to
the sides where the sound that is striking the walls and turning into
ambience is coming from.

As I said earlier, you can limit brass waveforms a LOT before they even
start to be noticed, let alone become unpleasant. Hell, take a look at
the grooves on a Herb Alpert record... that's a very close-miked horn and
the grooves are almost symmetric. That doesn't happen naturally.

The customer doesn't want a 'modern' sounding record. If the brass
overloads when it is recorded, I'll take the whole level down. They
could do it in 1926 with a 30 dB S/N ratio, I ought to be able to do it
in 2008 with an 80 dB S/N ratio.

I'm just using it as an example... once you record, though, try and play
with a limiter, because you can limit an outrageous amount without any
real sonic effect and it's interesting just to see how it works. Also look
at the waveform on a scope and see it.... brass is just so neat.

I'm not suggesting you necessarily do this for the release, but you need
to do this so you'll see what the tool does for you.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."