Does the fact, that compressed music sounds louder than uncompressedaudio, mean that monitors should be calibrated to a lower level than 83dBspl?

[Nono]

I've been reading various articles about calibrating monitors and this
"magic" sound mixing level of 83dBspl.
This is apparently the level where we can hear all frequencies
equally, making it the best level for sound mixing.
After finding this 83dBspl with RMS measured pink noise played back at
0dB VU, it should be marked "0" on the monitor control.
From there on we can use 1kHz sine wave with a decibel meter on the
monitor out cable to find (especially) the -6dB and the -8dB positions
(according to Bob Katz).
This is because more compressed music like most pop music sounds
louder than uncompressed music and therefore should be monitored at
these lower positions.

Makes sense, but does it mean that this compressed music played at
these lower positions give the same spl level as uncompressed audio at
the "0" position?
So, it would be a wrong conclusion that as "0" =83dBspl, then "-6"
=77dBspl and "-8" =75dBspl
so that for music mixes one should calibrate their monitors to a lower
reference level like 78dBc, like this article
http://audiodesignlabs.com/wordpress/2008/05/professional-monitoring-system-calibration/
says, right?
In fact, these three positions all give 83dBspl, providing you play
the right type of audio at the right position.

I'd appreciate your help clarifying this before I start putting the
wrong marks on my console.

Best regards,
Norman.

[Richard Crowley]

Nono wrote:
> I've been reading various articles about calibrating monitors and this
> "magic" sound mixing level of 83dBspl.
> This is apparently the level where we can hear all frequencies
> equally, making it the best level for sound mixing.
> After finding this 83dBspl with RMS measured pink noise played back at
> 0dB VU, it should be marked "0" on the monitor control.
> From there on we can use 1kHz sine wave with a decibel meter on the
> monitor out cable to find (especially) the -6dB and the -8dB positions
> (according to Bob Katz).
> This is because more compressed music like most pop music sounds
> louder than uncompressed music and therefore should be monitored at
> these lower positions.
>
> Makes sense, but does it mean that this compressed music played at
> these lower positions give the same spl level as uncompressed audio at
> the "0" position?

Missing is the motivation for making compressed music sound similar
to unmolested recordings? You question seems more like a philosophical
question than a technical one.

Are you trying to master recordings to "compete" with other highly-
compressed recordings? Then perhaps the other compressed recordings
should be used to set the magic listening level (whatever it is)?

[Nono]

On 7 okt, 15:01, "Richard Crowley" > wrote:

> Missing is the motivation for making compressed music sound similar
> to unmolested recordings? You question seems more like a philosophical
> question than a technical one.

No, I'm not trying to make compressed music sound similar to
unmolested recordings, but let's say that I'm mixing a song that, for
whatever reason, is chosen to be compressed, I'd like to monitor it at
the appropriate level.
And when mixing those unmolested recordings I'd want to monitor them
at the level that is the most suited for this.
My only question is if both of these levels are the same 83dBspl, but
only on different monitor control positions, or are there different
"magic" SPL levels for different type of audio (in terms of
compression).
So, I may be wrong, but don't see anything philosophical about that.

> Are you trying to master recordings to "compete" with other highly-
> compressed recordings? *Then perhaps the other compressed recordings
> should be used to set the magic listening level (whatever it is)?

"WHATEVER IT IS?"
Well, THAT was my question:"WHAT IS IT?"
IS'NT IT THE SAME FOR ALL TYPES OF AUDIO?

[Scott Dorsey]

Nono > wrote:
>I've been reading various articles about calibrating monitors and this
>"magic" sound mixing level of 83dBspl.

There's nothing magic about it.

>This is apparently the level where we can hear all frequencies
>equally, making it the best level for sound mixing.

No, not at all. Look at the Fletcher-Munson curves. There are no levels
at which we hear all frequencies equally.

83dB is a reasonable level, because it represents about the average level
at which stuff will be played back. It's not too loud, it's not too soft.

>After finding this 83dBspl with RMS measured pink noise played back at
>0dB VU, it should be marked "0" on the monitor control.
>From there on we can use 1kHz sine wave with a decibel meter on the
>monitor out cable to find (especially) the -6dB and the -8dB positions
>(according to Bob Katz).
>This is because more compressed music like most pop music sounds
>louder than uncompressed music and therefore should be monitored at
>these lower positions.

I'll buy that.

>Makes sense, but does it mean that this compressed music played at
>these lower positions give the same spl level as uncompressed audio at
>the "0" position?

Yes, precisely. SPL is an average level. When you compress something, you're
bring the average levels up and keeping the peak levels the same.

>So, it would be a wrong conclusion that as "0" =83dBspl, then "-6"
>=77dBspl and "-8" =75dBspl
>so that for music mixes one should calibrate their monitors to a lower
>reference level like 78dBc, like this article
>http://audiodesignlabs.com/wordpress/2008/05/professional-monitoring-system-calibration/
>says, right?
>In fact, these three positions all give 83dBspl, providing you play
>the right type of audio at the right position.

I tend to like to monitor at higher levels than that, personally, because I
usually work with music that has a very high crest factor and isn't compressed
much at all. So it depends a lot on how you work.

>I'd appreciate your help clarifying this before I start putting the
>wrong marks on my console.

Put reasonable markings on your console, but keep an SPL meter near the
console and glance at it now and then.

There are some benefits to listening to some things at very high levels,
and benefits to listening to other things at very low levels. But somewhere
in that ballpark is about right.
--scott
--
"C'est un Nagra. C'est suisse, et tres, tres precis."

[Wecan do it]

"Nono" > wrote in message
...
> I've been reading various articles about calibrating
> monitors and this
> "magic" sound mixing level of 83dBspl.
> This is apparently the level where we can hear all
> frequencies
> equally, making it the best level for sound mixing.
> After finding this 83dBspl with RMS measured pink noise
> played back at
> 0dB VU, it should be marked "0" on the monitor control.
> From there on we can use 1kHz sine wave with a decibel meter
> on the
> monitor out cable to find (especially) the -6dB and the -8dB
> positions
> (according to Bob Katz).
> This is because more compressed music like most pop music
> sounds
> louder than uncompressed music and therefore should be
> monitored at
> these lower positions.
>
> Makes sense, but does it mean that this compressed music
> played at
> these lower positions give the same spl level as
> uncompressed audio at
> the "0" position?

Yes you are correct.

I have done this in my room years ago. I also have the SPL
meter permanantly integrated into my room and it has a very
large red LED dispaly and can easily be seen from every
listening place in the room. I also calibrated analoug meters
(an old compressor in bypass on the spare output of the DAC
and not in the signal chain) to the amp drive signal so I can
see how hard the signal is hitting against 0dBfsd.

1) After years of usage I can tell in a matter of a couple of
bars what the "k" level is on anything playing from a CD or
out of my DAW. It is kinda guessing since I am attempting to
measure a constantly changing music level not steady state
noise, but after looking at stuff that long it is not
difficult to see the average on most songs I work with.

2) If my analoug meters are lighting up all red, (the top red
light can not be made to light with unclipped sine input, but
it sure lights up a lot when you put a crushed to death CD in
the player) and the room level is 83 db with the "level" pot
at the k-18 position I know the level of the CD is K-18.

3) If I tickle the green led's and have to put the "level" pot
to K+6 to hear it at 80 dB in the room I know that I am having
at least 9dB of headroom. A lot of gain is nice to have on
your 'Level" control so you can max out your speakers with
high headroom digital signals.

I would say just cal and start thinking differently. This K
system is a way of thinking as much as a way of calibration of
hardware.

I had an engineer in here the other day and he kept reaching
for the level control and I chastized him for it and put it
back to K+18. Then he reaches for the master control in the
DAW and I told him that he mixes like a guy that has to pee
really bad, and instead of holding it in with his bladder he
is pinching the end of his you-know-what. He had to go back to
the tracks and subs to lower the room SPL and put some
headroom and crest in his music.

Think in K will help you control the bladder of your music and
have some headroom and some crest in your waveforms. And yes
the guy who puts your CD's or MP3's in the player may have to
turn up the volume on his earbuds, but it will sure sound
better and he may even play it again.

peace
dawg





> So, it would be a wrong conclusion that as "0" =83dBspl,
> then "-6"
> =77dBspl and "-8" =75dBspl
This is correct so long as you are not changing the signal
coming into the "level" control and just turning the "level"
control.


> so that for music mixes one should calibrate their monitors
> to a lower
> reference level like 78dBc, like this article
> http://audiodesignlabs.com/wordpress/2008/05/professional-monitoring-system-calibration/

If you set your level to K-6 and play a mix that has 6dB
headroom you will hear at 83 dBSPL.



> says, right?
> In fact, these three positions all give 83dBspl, providing
> you play
> the right type of audio at the right position.
>
> I'd appreciate your help clarifying this before I start
> putting the
> wrong marks on my console.
>
> Best regards,
> Norman.

[Nono]

On 7 okt, 15:41, (Scott Dorsey) wrote:
> Nono > wrote:
> >I've been reading various articles about calibrating monitors and this
> >"magic" sound mixing level of 83dBspl.
>
> There's nothing magic about it.
>
> >This is apparently the level where we can hear all frequencies
> >equally, making it the best level for sound mixing.
>
> No, not at all. *Look at the Fletcher-Munson curves. *There are no levels
> at which we hear all frequencies equally. *
>
> 83dB is a reasonable level, because it represents about the average level
> at which stuff will be played back. *It's not too loud, it's not too soft.
>
> >After finding this 83dBspl with RMS measured pink noise played back at
> >0dB VU, it should be marked "0" on the monitor control.
> >From there on we can use 1kHz sine wave with a decibel meter on the
> >monitor out cable to find (especially) the -6dB and the -8dB positions
> >(according to Bob Katz).
> >This is because more compressed music like most pop music sounds
> >louder than uncompressed music and therefore should be monitored at
> >these lower positions.
>
> I'll buy that.
>
> >Makes sense, but does it mean that this compressed music played at
> >these lower positions give the same spl level as uncompressed audio at
> >the "0" position?
>
> Yes, precisely. *SPL is an average level. *When you compress something, you're
> bring the average levels up and keeping the peak levels the same.
>
> >So, it would be a wrong conclusion that as "0" =83dBspl, then "-6"
> >=77dBspl and "-8" =75dBspl
> >so that for music mixes one should calibrate their monitors to a lower
> >reference level like 78dBc, like this article
> >http://audiodesignlabs.com/wordpress/2008/05/professional-monitoring-...
> >says, right?
> >In fact, these three positions all give 83dBspl, providing you play
> >the right type of audio at the right position.
>
> I tend to like to monitor at higher levels than that, personally, because I
> usually work with music that has a very high crest factor and isn't compressed
> much at all. *So it depends a lot on how you work.
>
> >I'd appreciate your help clarifying this before I start putting the
> >wrong marks on my console.
>
> Put reasonable markings on your console, but keep an SPL meter near the
> console and glance at it now and then.
>
> There are some benefits to listening to some things at very high levels,
> and benefits to listening to other things at very low levels. *But somewhere
> in that ballpark is about right.
> --scott
> --
> "C'est un Nagra. *C'est suisse, et tres, tres precis."

Thanks Scot,

One more question:
I guess I used the wrong terminology when I said that at 83dBspl we
hear all frequencies equally while I meant that at that level our ears
wouldn't accentuate certain frequencies over others (..something like
that), kind of a fairly flat frequency response.
Is the answer also "no" on this statement?

Thanks again,
Norman.

[Wecan do it]

"Nono" > wrote in message
...
On 7 okt, 15:01, "Richard Crowley" > wrote:

> Missing is the motivation for making compressed music sound
> similar
> to unmolested recordings? You question seems more like a
> philosophical
> question than a technical one.

No, I'm not trying to make compressed music sound similar to
unmolested recordings, but let's say that I'm mixing a song
that, for
whatever reason, is chosen to be compressed, I'd like to
monitor it at
the appropriate level.
And when mixing those unmolested recordings I'd want to
monitor them
at the level that is the most suited for this.
My only question is if both of these levels are the same
83dBspl, but
only on different monitor control positions, or are there
different
"magic" SPL levels for different type of audio (in terms of
compression).
So, I may be wrong, but don't see anything philosophical about
that.

> Are you trying to master recordings to "compete" with other
> highly-
> compressed recordings? Then perhaps the other compressed
> recordings
> should be used to set the magic listening level (whatever it
> is)?

"WHATEVER IT IS?"
Well, THAT was my question:"WHAT IS IT?"
IS'NT IT THE SAME FOR ALL TYPES OF AUDIO?


Calibrating your system and being aware of your K settings
will help you understand your absolute digital level
regardless of your monitoring level.

It is kind of like the problem a pilot has when determining
speed. Some instruments will tell him how fast he is traveling
with respect to the plane going through the wind. If he is
traveling with the wind the instrument will read low and if he
is traveling the same speed into the wind the instrument will
read high. If he measures with respect to the ground the speed
would be constant regardless of the wind.

Your ears and the music are the wind. K is the ground.

peace
dawg

[Wecan do it]

"Nono" > wrote in message One more
question:
I guess I used the wrong terminology when I said that at
83dBspl we
hear all frequencies equally while I meant that at that level
our ears
wouldn't accentuate certain frequencies over others
(..something like
that), kind of a fairly flat frequency response.
Is the answer also "no" on this statement?

Thanks again,
Norman.
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!
Just dont forget that the equal loudness contours were
derived, long ago, from a statistical compilation of hearing
tests from a limited number of subjects and does not
necessarily translate to any one human being, especially you.

Peace
dawg

[Nono]

On 7 okt, 16:23, "Wecan do it" > wrote:
> "Nono" > wrote in message One more
>
> question:
> I guess I used the wrong terminology when I said that at
> 83dBspl we
> hear all frequencies equally while I meant that at that level
> our ears
> wouldn't accentuate certain frequencies over others
> (..something like
> that), kind of a fairly flat frequency response.
> Is the answer also "no" on this statement?
>
> Thanks again,
> Norman.
> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!
> Just dont forget that the equal loudness contours were
> derived, long ago, from a statistical compilation of hearing
> tests from a limited number of subjects and does not
> necessarily translate to any one human being, especially you.
>
> Peace
> dawg

Thanks,
So, the K-system is to listen to a variety of music styles (in terms
of compression) at the same SPL level, and this level does not
necesary has to be 83dBSPL; you could choose any level that you're
comfortable with.
This means that the before mentioned article
http://audiodesignlabs.com/wordpress/2008/05/professional-monitoring-system-calibration/
is not wrong after all.
Anyway, I'll then be keeping my SPL meter out of the box for the time
being to be sure I'm putting my monitor control at the right position,
at least until I have more experience with it.
Funny Though that on my first attempts to calibrate, 85dBSPL showed to
be at the approximate position where I normally have my monitor
control when mixing............Of course I doubded that that could be
right...........

Regards,
Norman

[Wecan do it]

"Nono" > wrote in message
...
> On 7 okt, 16:23, "Wecan do it" > wrote:
>> "Nono" > wrote in message One
>> more
>>
>> question:
>> I guess I used the wrong terminology when I said that at
>> 83dBspl we
>> hear all frequencies equally while I meant that at that
>> level
>> our ears
>> wouldn't accentuate certain frequencies over others
>> (..something like
>> that), kind of a fairly flat frequency response.
>> Is the answer also "no" on this statement?
>>
>> Thanks again,
>> Norman.
>> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!
>> Just dont forget that the equal loudness contours were
>> derived, long ago, from a statistical compilation of
>> hearing
>> tests from a limited number of subjects and does not
>> necessarily translate to any one human being, especially
>> you.
>>
>> Peace
>> dawg
>
> Thanks,
> So, the K-system is to listen to a variety of music styles
> (in terms
> of compression) at the same SPL level, and this level does
> not
> necesary has to be 83dBSPL; you could choose any level that
> you're
> comfortable with.

For me, the K system helps me see (by looking at the Level
control) the amount of headroom in a digital signal hitting
the final DAC only when I am listening at 83dBspl c. Headroom
is then calibrated to how many dB below the constant max of
0dBfsd the signal is. And remember with music where it is
83dbspl is up to your own averaging of the meter you are
looking at.

My cobbled output-of-the-DAC meter made from an old compressor
on input monitor mode lets me get a feel for the amount of
compression no matter how loud or low I am monitoring in the
room.

Both tools work for ME because I am familiar with how they
look over years of listening to my reference music and other
music with them on all the time. I have to look at these
things often to keep my internal refference calibrated. These
tools help you keep some objectivity in you mix levels.

peace
dawg





> This means that the before mentioned article
> http://audiodesignlabs.com/wordpress/2008/05/professional-monitoring-system-calibration/
> is not wrong after all.
> Anyway, I'll then be keeping my SPL meter out of the box for
> the time
> being to be sure I'm putting my monitor control at the right
> position,
> at least until I have more experience with it.

Calibrate your system, park the control and use your spl meter
to understand how much headroom is in the signal. Varry the
signal going into the playback system / Level control to be
sure you are mixing at the right (you choose) level.


> Funny Though that on my first attempts to calibrate, 85dBSPL
> showed to
> be at the approximate position where I normally have my
> monitor
> control when mixing............Of course I doubded that that
> could be
> right...........
>
> Regards,
> Norman


peace
dawg

[Mike Rivers]

Nono wrote:
> I've been reading various articles about calibrating monitors and this
> "magic" sound mixing level of 83dBspl.
> This is apparently the level where we can hear all frequencies
> equally, making it the best level for sound mixing.

No, not really. It's a reasonably flat Fletcher-Munson curve, but the
reason for that level (actually that's 83 dB with one speaker, 85 dB
SPL with both speakers) is that's what SMPTE decided was the
standard level for mixing film sound. It's pretty loud, in my book.

> From there on we can use 1kHz sine wave with a decibel meter on the
> monitor out cable to find (especially) the -6dB and the -8dB positions
> (according to Bob Katz).
> This is because more compressed music like most pop music sounds
> louder than uncompressed music and therefore should be monitored at
> these lower positions.

Yes, that's the basis for Bob Katz' K-System of metering.

> Makes sense, but does it mean that this compressed music played at
> these lower positions give the same spl level as uncompressed audio at
> the "0" position?

Not necessarily the same SPL as measured on a meter, but it sounds
as loud. The idea behind the loudness wars is that if it sounds pretty
loud with that amount of monitor gain, it will sound even louder with
your reference monitor gain.

> So, it would be a wrong conclusion that as "0" =83dBspl, then "-6"
> =77dBspl and "-8" =75dBspl
> so that for music mixes one should calibrate their monitors to a lower
> reference level like 78dBc, like this article

> http://audiodesignlabs.com/wordpress/2008/05/professional-monitoring-system-calibration/
> says, right?

I didn't read the article, but the idea is that if you know how loud
uncompressed material sounds like at your reference (85 dB) level,
you should reduce the level by 6 or 8 dB, compress the **** out of
it, and mix it so it sounds a loud as your uncompressed reference.
Then, when you put back that extra 6 or 8 dB it'll be louder.

I once set up a K-System monitor calibration just out of curiosity,
but I didn't like mixing at 85 dB. It was too loud for me. And
besides, I don't mix for loudness.

[Nono]

On 7 okt, 18:40, Mike Rivers > wrote:

> I once set up a K-System monitor calibration just out of curiosity,
> but I didn't like mixing at 85 dB. It was too loud for me. And
> besides, I don't mix for loudness.

Couldn't one set up a K-system monitor calibration using a different
level than 85/83dBSPL?
I guess that if one always work with one style of music (in terms of
the amount / lack of compression) could suffice finding just that one
comfortable working level and mark that monitor position as reference?

What I'm trying right now is, with 83dBSPL as starting point, marking
the 0, -3, -6, -8 and -12 points for both the Genelec 1031A's and the
Auratones, as these points are different for both systems.
I think this will be sufficient for now, but time and experience will
tell.

[Wecan do it]

"Nono" > wrote in message
...
> On 7 okt, 18:40, Mike Rivers > wrote:
>
>> I once set up a K-System monitor calibration just out of
>> curiosity,
>> but I didn't like mixing at 85 dB. It was too loud for me.
>> And
>> besides, I don't mix for loudness.
>
> Couldn't one set up a K-system monitor calibration using a
> different
> level than 85/83dBSPL?
> I guess that if one always work with one style of music (in
> terms of
> the amount / lack of compression) could suffice finding just
> that one
> comfortable working level and mark that monitor position as
> reference?
>
> What I'm trying right now is, with 83dBSPL as starting
> point, marking
> the 0, -3, -6, -8 and -12 points for both the Genelec
> 1031A's and the
> Auratones, as these points are different for both systems.
> I think this will be sufficient for now, but time and
> experience will
> tell.

You can go over and under the "set point" in the Katz system
the set point is 83 dBspl for 0dBfsd. If you reach 83 dB spl
with your program and your "level knob" is K-3 you have only
3dB headroom. If you only reach 80dB with the same knob
position you have 6dB of headroom. This should be playing at
83dBspl with the Level Knob on K-6.

If you have the gain and cal up on the K system (K+3 - K+6
etc) then you can listen to music with any amount of headroom
at any SPL you want and still know how much headroom your
digital has.

With the K system you can get a feel for where you are with
respect to 0dBfsd, which is an absolute level of all ones and
not any higher, if you know the knob position and the SPL from
the speakers.

K system helps you set an objective acoustic reference.

peace
dawg

[Nono]

On 7 okt, 18:40, Mike Rivers > wrote:
> Nono wrote:
> > I've been reading various articles about calibrating monitors and this
> > "magic" sound mixing level of 83dBspl.
> > This is apparently the level where we can hear all frequencies
> > equally, making it the best level for sound mixing.
>
> No, not really. It's a reasonably flat Fletcher-Munson curve, but the
> reason for that level (actually that's 83 dB with one speaker, 85 dB
> SPL with both speakers) is that's what SMPTE decided was the
> standard level for mixing film sound. It's pretty loud, in my book.
>
> > From there on we can use 1kHz sine wave with a decibel meter on the
> > monitor out cable to find (especially) the -6dB and the -8dB positions
> > (according to Bob Katz).
> > This is because more compressed music like most pop music sounds
> > louder than uncompressed music and therefore should be monitored at
> > these lower positions.
>
> Yes, that's the basis for Bob Katz' K-System of metering.
>
> > Makes sense, but does it mean that this compressed music played at
> > these lower positions give the same spl level as uncompressed audio at
> > the "0" position?
>
> Not necessarily the same SPL as measured on a meter, but it sounds
> as loud. The idea behind the loudness wars is that if it sounds pretty
> loud with that amount of monitor gain, it will sound even louder with
> your reference monitor gain.


This is what I meant:
I understand that this flat Fletcher-Munson curve exists, but is not
necessarily 83dBSPL, and not the same for everybody.
Let's now assume that this Fletcher-Munson curve is flat for person A
at XXdBSPL playing an uncompressed recording.
Should we now compress the hell out of this song and play it back
again at XXdBSPL, I suppose that person A would perceive it to be
louder.
My question is then; would the Fletcher-Munson curve still be flat for
him?
Is the Fletcher-Munson curve related to a certain absolute SPL level
(per individual) or is it related to the person's perception of this
SPL level = loudness?

[Scott Dorsey]

Nono > wrote:
>
>One more question:
>I guess I used the wrong terminology when I said that at 83dBspl we
>hear all frequencies equally while I meant that at that level our ears
>wouldn't accentuate certain frequencies over others (..something like
>that), kind of a fairly flat frequency response.
>Is the answer also "no" on this statement?

Look at the curves.

As the level changes, the frequency response of your ears change. There is
no one "correct" level at which it is best.

We are used to hearing quiet sounds in one way, and loud sounds in another
way. When you play quiet sounds loudly, or loud sounds quietly, either way
the perceived tonality changes.

So the level at which you want to monitor has to do with the level at which
you expect the listener to be listening. In the classical and jazz world,
that's related also to the level of the original sounds. The the rock world
it often isn't.
--scott

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

[David Gravereaux]

Kind of off topic, but this reminds me of a time back around '92 when
the chief engineer at this place I was chief tech at, asked me to raise
the operating level (lower the output) of our DAT player from
-18dBFS=0VU to be -15dBFS=0VU.

I said no quite strongly. The producer was being "disturbed" that the
mixes were coming back "too hot". Ummm.. yeah, cause you're sending
too hot and trying to get full bit for each song. With the clipping
point of the 8086 desk really being around +30dBv, I should really be
dropping things instead.
--



-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)

iEYEARECAAYFAkrM5rgACgkQlZadkQh/RmGbnQCfXkTZV9eYqjoqTDODc27iLY5t
s70AoOWi3t+ksj4p2aKW8StcF7MD5GTD
=NK27
-----END PGP SIGNATURE-----

[Powell]

"Nono" wrote

> I've been reading various articles about calibrating monitors and this
> "magic" sound mixing level of 83dBspl.
> This is apparently the level where we can hear all frequencies
> equally, making it the best level for sound mixing.
>
Ok... 83 dB is a pretty good metric


> After finding this 83dBspl with RMS measured pink noise played back at
> 0dB VU, it should be marked "0" on the monitor control.
>
Good idea in theory, but pink noise is wideband noise that
maintains constant energy per octave. In reality only a very
few expensive monitor speakers are capable of a true (flat)
full range (20 - 20,000). So your ""0" on the monitor control"
using pink noise has no true calibration metric associated
with it. .


> This is because more compressed music like most pop music sounds
> louder than uncompressed music and therefore should be monitored at
> these lower positions.
>
Compression is an automated level control using the input
signal to determine the output level. So compression does
not make music sound louder. The reasons for compressing
"pop music" depends on the medium it is being played on...
radio, CD, MP3. While compression can sometime increase
fidelity when used on an individual channel as a whole it
decreases fidelity when used on the final mix. As a result
a single song, for example, will be produced in a number of
ways depending the final target market. In other words the
song on the radio is not the song on the CD, dynamically
speaking.


> Makes sense, but does it mean that this compressed music played at
> these lower positions give the same spl level as uncompressed audio at
> the "0" position?
> So, it would be a wrong conclusion that as "0" =83dBspl, then "-6"
> =77dBspl and "-8" =75dBspl
> so that for music mixes one should calibrate their monitors to a lower
> reference level like 78dBc, like this article
> http://audiodesignlabs.com/wordpress/2008/05/professional-monitoring-system-calibration/
> says, right?
> In fact, these three positions all give 83dBspl, providing you play
> the right type of audio at the right position.
>
> I'd appreciate your help clarifying this before I start putting the
> wrong marks on my console.
>
What do you need "marks on my console" for?

[Scott Dorsey]

Nono > wrote:
>
>This is what I meant:
>I understand that this flat Fletcher-Munson curve exists, but is not
>necessarily 83dBSPL, and not the same for everybody.

It does not exist. Look at the curves.
--scott


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

[Nono]

On 7 okt, 20:37, "Wecan do it" > wrote:
> "Nono" > wrote in message
>
> ...
>
>
>
>
>
> > On 7 okt, 18:40, Mike Rivers > wrote:
>
> >> I once set up a K-System monitor calibration just out of
> >> curiosity,
> >> but I didn't like mixing at 85 dB. It was too loud for me.
> >> And
> >> besides, I don't mix for loudness.
>
> > Couldn't one set up a K-system monitor calibration using a
> > different
> > level than 85/83dBSPL?
> > I guess that if one always work with one style of music (in
> > terms of
> > the amount / lack of compression) could suffice finding just
> > that one
> > comfortable working level and mark that monitor position as
> > reference?
>
> > What I'm trying right now is, with 83dBSPL as starting
> > point, marking
> > the 0, -3, -6, -8 and -12 points for both the Genelec
> > 1031A's and the
> > Auratones, as these points are different for both systems.
> > I think this will be sufficient for now, but time and
> > experience will
> > tell.
>
> You can go over and under the "set point" in the Katz system
> the set point is 83 dBspl for 0dBfsd. *If you reach 83 dB spl
> with your program and your "level knob" *is K-3 you have only
> 3dB headroom. If you only reach 80dB with the same knob
> position you have 6dB of headroom. This should be playing at
> 83dBspl with the Level Knob on K-6.
>
> If you have the gain and cal up on the K system (K+3 - K+6
> etc) *then you can listen to music with any amount of headroom
> at any *SPL you want and still know how much headroom your
> digital has.
>
> With the K system you can get a feel for where you are with
> respect to 0dBfsd, which is an absolute level of all ones and
> not any higher, if you know the knob position and the SPL from
> the speakers.
>
> K system helps you set an objective acoustic reference.
>
> peace
> dawg- Tekst uit oorspronkelijk bericht niet weergeven -
>
> - Tekst uit oorspronkelijk bericht weergeven -

Thanks for your patience and very good explanation; it's becoming
very, very clear to me.
The SPl meter is going to become an integral part of my system.
Now I'm thinking about taking apart one of those cheap SPL meters and
mounting the display on the meterbridge and the microphone somewhere
appropriate and see how I can make it work on something else than
batteries.
So, if I'm not heard from again, I've probably been electrocuted, LOL.

Regards,
Norman.

[Nono]

On 7 okt, 21:07, "Powell" > wrote:

> What do you need "marks on my console" for?

I'm trying to adopt, a somewhat simplified version of, Bob Katz's K-
system for monitoring.

[Mark]

On Oct 7, 3:22*pm, Nono > wrote:
> On 7 okt, 21:07, "Powell" > wrote:
>
> > What do you need "marks on my console" for?
>
> I'm trying to adopt, a somewhat simplified version of, Bob Katz's K-
> system for monitoring.

my take on this is:

if you play highly compressed material your SPL meter will only move a
little so you can set it so it sets around 83 dBSPL or whatever level
you like...

then wen you play uncompressed material, the meter will move a lot
more so you have to decide what that really means, my decision is that
I try to set the level so that for about 1/2 the time the meter is
above the set point and for 1/2 the time the meter is below the set
point..

this has worked for me to make two recordings with different amounts
of compression and dynamic range, sound about the same level on
average...

Mark

[Nono]

On 7 okt, 21:09, (Scott Dorsey) wrote:
> Nono > wrote:
>
> >This is what I meant:
> >I understand that this flat Fletcher-Munson curve exists, but is not
> >necessarily 83dBSPL, and not the same for everybody.
>
> It does not exist. *Look at the curves.
> --scott
>
> --
> "C'est un Nagra. *C'est suisse, et tres, tres precis."

My appologies for going on about something while you have already
posted an answer to my question.
The thing is that, English not being my native language, while I'm
formulating what I want to ask and by the time that I post it, the
answer has already been posted by you or one of the other guys.

Norman.

[Mike Rivers]

Nono wrote:

> Couldn't one set up a K-system monitor calibration using a different
> level than 85/83dBSPL?

I suppose, but I just turn the monitor level up to where it's comfortable
and use the meters to let me know if the level I'm running is in the
ballpark.

[Wecan do it]

"Nono" > wrote in message
...
>> No, not really. It's a reasonably flat Fletcher-Munson
>> curve, but the
>> reason for that level (actually that's 83 dB with one
>> speaker, 85 dB
>> SPL with both speakers) is that's what SMPTE decided was
>> the
>> standard level for mixing film sound. It's pretty loud, in
>> my book.

You dont have to mix at 83dbspl. But if something hits 0dBfsd
long enough it will be 83dBspl in your room. You can always
mix lower. Obsurdly if you had to mix at 83db how much gain
would you need to make the silent pause 83 db?


>> > Makes sense, but does it mean that this compressed music
>> > played at
>> > these lower positions give the same spl level as
>> > uncompressed audio at
>> > the "0" position?

Now you are getting it! Yes. To make high headroom recordings
play at 83 the gain must be turned to k-0 or higher. To make
crunchy k-3 recordings play at 83 you will have to turn the
gain on the level pot down 18db.

Where you will really notice the difference in the mix is when
you play the recording at 65 dB. On the high headroom
recording you will only hear the loud parts and on the
crunched one you will hear everything, perhaps with noticable
distortion, but you will hear it.

Same in the car or anywhere else with a lot of noise to drown
out the quiet.

>>
>> Not necessarily the same SPL as measured on a meter, but it
>> sounds
>> as loud. The idea behind the loudness wars is that if it
>> sounds pretty
>> loud with that amount of monitor gain, it will sound even
>> louder with
>> your reference monitor gain.

Dont forget that SPL does not conviently translate to loudness
when the signal is music. SPL is an objective measured
quantity, loudness is a subjective quantity with no direct way
of measurement as far as I know.
>
>
> This is what I meant:
> I understand that this flat Fletcher-Munson curve exists,
> but is not
> necessarily 83dBSPL, and not the same for everybody.
> Let's now assume that this Fletcher-Munson curve is flat for
> person A
> at XXdBSPL playing an uncompressed recording.
> Should we now compress the hell out of this song and play it
> back
> again at XXdBSPL, I suppose that person A would perceive it
> to be
> louder.

If done right the person would hear it at the same perceived
loudness but would perceive the damage done (or not done) to
the music by the compressor.


> My question is then; would the Fletcher-Munson curve still
> be flat for
> him?
> Is the Fletcher-Munson curve related to a certain absolute
> SPL level
> (per individual) or is it related to the person's perception
> of this
> SPL level = loudness?

FM is a curve measured by playing a tone of SPLx at 1kHz then
playing the other tones till the listener thinks the level of
other tone is same as 1K. This happens at fixed spl's.

peace
dawg

[Wecan do it]

"Powell" > wrote in message
...
> Good idea in theory, but pink noise is wideband noise that
> maintains constant energy per octave. In reality only a very
> few expensive monitor speakers are capable of a true (flat)
> full range (20 - 20,000). So your ""0" on the monitor
> control"
> using pink noise has no true calibration metric associated
> with it. .


The SMPTE method referenced in Katz book requires the levels
to be calibrated using sub octave band limited noise (or maybe
it was warble tones) to get around the bandwidth and room mode
issues.

Since we are dealing in "loudness" there will never be a true
calibration but being in the ball park on the right base can
sometimes get you everything you want as far as a reference
level.

The real benefit of the K system can be that one can listen to
many CD's now and see the level that they were printed at and
make a very good estimate of the headroom in the recording one
is making with respect to those recordings.

peace
dawg

[Wecan do it]

"Nono" > wrote in message
...
> On 7 okt, 21:07, "Powell" > wrote:
>
>> What do you need "marks on my console" for?
>
> I'm trying to adopt, a somewhat simplified version of, Bob
> Katz's K-
> system for monitoring.

Get ready for every one here who does not understand what Bob
has said to rain on your parade and make you doubt if you made
a good decision to adopt the K system. I think it is about 10
years in the making now.

Give it 10 more years and everyone here poo pooing it today
will have been an early adopter, adopting it 30 years ago,
even before Bob Katz thought it up.

Go for it Norm, the K-system will help you get a handle on
your work.

Peace
dawg

[Scott Dorsey]

Wecan do it > wrote:
>
>The real benefit of the K system can be that one can listen to
>many CD's now and see the level that they were printed at and
>make a very good estimate of the headroom in the recording one
>is making with respect to those recordings.

I always liked Gabe Wiener's trick of indicating the original SPL level of the
opening bars of the CD in the liner notes, so you could make sure the
approximate playback level matched. More labels should do this.
--scott

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

[Nono]

On 8 okt, 01:51, "Wecan do it" > wrote:

> Go for it Norm, the K-system will help you get a handle on
> your work.
>
> Peace
> dawg

THANK YOU, and also the other contributers, very much.
I have been in contact with Bob Katz himself too, and I must say that
he has been very patient, friendly and helpful as well.

Best regards,
Norman.

[Wecan do it]

"Scott Dorsey" > wrote in message
...
> Wecan do it > wrote:
>>
>>The real benefit of the K system can be that one can listen
>>to
>>many CD's now and see the level that they were printed at
>>and
>>make a very good estimate of the headroom in the recording
>>one
>>is making with respect to those recordings.
>
> I always liked Gabe Wiener's trick of indicating the
> original SPL level of the
> opening bars of the CD in the liner notes, so you could make
> sure the
> approximate playback level matched. More labels should do
> this.
> --scott
>
> --
> "C'est un Nagra. C'est suisse, et tres, tres precis."

I did that for a while when I got the K- system up. By putting
the K-n (n= the amount of headroom below 0dbfsd) next to the
song I could set the "level" control at the appropriate level
and have it play at 83dBspl or any KNOWN spl I wanted without
hearing any music.

What a breakthrough if the level could be encoded into the
header of a song at manufacture and the player could read it
and set the SPL to a target the listener prefers.

I bought a CD changer that could remember the level for a
hundred CD's. It was a very important feature in my buying
decision. I wore that CD player out over a period of 15 years
and I never used the feature once. It was a pain to program
it.


peace
dawg

[Richard Crowley]

Wecan do it wrote:
> What a breakthrough if the level could be encoded into the
> header of a song at manufacture and the player could read it
> and set the SPL to a target the listener prefers.
>
> I bought a CD changer that could remember the level for a
> hundred CD's. It was a very important feature in my buying
> decision. I wore that CD player out over a period of 15 years
> and I never used the feature once. It was a pain to program
> it.

The way modern CDs are mastered, it hardly seems like it
really matters. But apparently you proved that yourself.

[Eric B[_2_]]

An interesting discussion.
Bottom line- your ears will last longer if you monitor at a lower
level.
However, some things want to be listened to at higher SPL because they
will be played back at high SPL
For tracking, it hardly matters.
For mixing, if it feels good and works for you, OK but it is not
critical.
For mastering, perhaps you have a point, but I suspect that if you
were to poll mastering engineers that they may concede to your point,
but will fall back on whatever works for them.
And sometimes it just feels good to break out of the box you can
create for yourself by trying to be (perhaps) too standardized.
....interesting discussion
Eric

[Paul Bissell[_2_]]

>You can go over and under the "set point" in the Katz system
>the set point is 83 dBspl for 0dBfsd.

I think some of the confusion is that you can look at this two ways. In
the above way, you are using 0dBfs as the reference which would lead to
an low initial postion of the monitoring knob. The way other implement
this (and what Katz has written) is to use K-20 as the initial reference
point. That is, set your tone or whatever to -20dBfs and mark the
monitor level knob where it makes 83dBspl - this would be the high point
on the monitoring knob. Then, all of your monitor knob settings will be
down from there.
It is half empty / half full since in both cases they work out the same
with the same dBfs and dbspl references.
I am just stating this as if the original poster of this thread picks up
the Katz book or reads the online materials the level set up would seem
'backwards' to using 0dBfs as the initial reference.
My question is how would one create a 0dBfs 'average level' output? Even
a sine wave would clip the DAC and thus the 83dBspl measured in the room
wouldn't be accurate. I figured that was the reason to start -20dB below
full scale digital.
All good...

Paul

[Wecan do it]

"Paul Bissell" > wrote in
message > My question is how would one create a 0dBfs 'average
level' output? >Even a sine wave would clip the DAC and thus
the 83dBspl measured in > the room wouldn't be accurate. I
figured that was the reason to start ->20dB below full scale
digital. All good...

> Paul


+ and - OdBfsd is the maximum single digit(s) that can be
enumerated from a waveform feeding an ADC. An unclipped
sinusoidal waveform at its maximum level will have one of
these and only one per alternation. If there are two in a row,
the resulting waveform will be doing something that only
occurs with electronic signals or pressure vessels, holding
pressure at a constant.

You can feed clipped waveform data, data having multiple
simultaneous 0dBfsd data in a row into a DAC and the DAC will
produce an electrical signal that will try to make the speaker
system hold the pressure but this may not sound natural, or
undistorted. This may also be the sound you desire.

So it is up to the mixer to decide at what SPL they would like
the signal that represents "nowhere else to go but down" set.
Katz chose 83dB because of the results of a "study".

The study he cites (IIRC) was one of a room full of SMPTE
convention goers. They listened to a program averaging SPL
around 83dB and much like Goldilocks, and her soup, when
asked it was to hot a couple of them raised their hands, when
asked if it was to low a couple of them raised their hands,
when asked if it was just right almost everyone raised their
hands.

Perhaps the -20dB starting reference is because that is the
level where it becomes difficult to have a crest factor
greater than with live music. A sine wave has a crest factor
of 3dB, random noise somewhere around 10dB and a gunshots have
been reported somewhere around 30dB.

Therefore if you calibrate your "monitoring level" control to
K-20 it is almost assured that no matter how un-compressed a
signal being fed to your speakers is if you have your knob at
K-20 you will be able to monitor at 83dB SPL.

Leaving you knob here and running a sine wave with one sample
per alternation touching 0dBfsd you should get 83dB with the
level control dialed down to the lower level at K-3.

~~~~~~~~~~~~~~~~~~~~~~~~```
Correcting earlier posts, if one desires even more gain that
k-20 one would have to calibrate the dial to K-23 K-26 etc as
the gain on the dial is measured above the K-20 point.

Program Level Control SPL
K-3 K-3 83dBspl
K-20 K-20 83dBspl
K-20 K-3 (83-17) 66 dBspl
K-20 K-24 (83+4) 87dBspl

Why? Cause no matter what your in room monitoring level you
can know what your headroom is on your program.

Hope this helps.

Peace
dawg