Can you help me calibrate my monitors according to the K-System?

[AlexRoudos]

Hello everyone, my apologies for this post since i know that this issue has been discussed in depth, but i'm very new to this and besides the fact that i've read everything i could find here, i still have the feeling that i miss a lot.

My basic problem is that i don't know how to start with the levels of my monitors(K+H 0300) and my DAC-1. I have downloaded the pink noise file and i also have the appropriate db meter but.......

Should i turn the monitor levels on the back all the way down or leave them where they are(at maximum) and just raise the volume from the DAC-1?

I'm definitely confused about it so any help would be greatly appreciated.

Thank you in advance.

[bleak orange]

yes,

it would be nice to have a simple step by step description on how to set it up.
K-systems for dummies sort of

I see the new psp xenon have pink noise generator to ease the setup( k-12, 14 and 20 ), but their manual says nothing on actually setting it up.

the Bob Katz site is full of info, but it can be a bit confusing

[bob katz]

Originally Posted by bleak orange

yes,

it would be nice to have a simple step by step description on how to set it up.
K-systems for dummies sort of

I see the new psp xenon have pink noise generator to ease the setup( k-12, 14 and 20 ), but their manual says nothing on actually setting it up.

the Bob Katz site is full of info, but it can be a bit confusing

The basic level settings are VERY basic and quite easy. But if you want to be fastidious about it you can get down to barometric pressure measurements and need a calibrated microphone! Let's just say that if you follow the following, you'll be within a dB or so of correct, which is more than enough for most people:

1--- download the -20 dBFS RMS pink noise signal from our website. This signal is accurate to 0.1 dB, wideband, but since it is at 44.1 kHz, avoids any information above 22.05 kHz. I have not measured the Xenon pink noise signal to know if it is accurate. Need to do that!

2---Play ONE channel at a time (use the mute button or disconnect the signal from all but one speaker

3---Using an SPL meter set to C-weighted, SLOW, adjust the gain of your amplifiers or DAC to yield 83 dB SPL at the 0 dB position of your monitor gain control. That's right, your monitor gain control needs to be marked in dB, or this whole idea falls apart. That's because the 0 dB position will be used very rarely.

Any more details than that and it starts getting into ways to be even more accurate, issue of bandwidth measurement errors, issues of fine-tuning the woofer level, and other little nitpicks, all of which are covered in chapter 15 of the second edition of my book, I think much easier to read and understand than the first edition. Someday I'll put that on the website, as well.

BK

[AlexRoudos]

Thank you very much Mr. Katz, this is exactly what i was looking for, and many others i suppose.

[bleak orange]

I don't know if its the language ( English is not my first language ) or that I am a complete idiot...but I still don't quite get it....

1) Load the pink noise file into the daw.
2) don't touch the volume in the daw. i.e. leave it at 0, no gain up or down
3) take out the spl meter and start gaining your monitor controller (dac ) untill the spl meter reads 83 db at each channel ( left - right )
4) then, whatever the dac reads, when the spl meter reads 83 db is 0 ??
5 ) and then ??

umm ??

[bob katz]

Originally Posted by bleak orange

I don't know if its the language ( English is not my first language ) or that I am a complete idiot...but I still don't quite get it....

1) Load the pink noise file into the daw.
2) don't touch the volume in the daw. i.e. leave it at 0, no gain up or down
3) take out the spl meter and start gaining your monitor controller (dac ) untill the spl meter reads 83 db at each channel ( left - right )
4) then, whatever the dac reads, when the spl meter reads 83 db is 0 ??
5 ) and then ??

umm ??

The whole thing is based on having a monitor gain control marked in decibels. In step 2 you also need to set that monitor control to 0 dB position.

Then step 3 is pretty obvious.

Then step 4, when the SPL meter reads 83 dB at 0 dB on your monitor control, you're calibrated. To the get an idea of what this calibration means, you might visit the honor roll at Digido.com. If your monitor gain correlates with, say, 2 dB of what's listed there, then you are probably listening to the same material at similar listening level as many other mastering engineers whose "sensitivity" for forte falls around 83-86 dB SPL.

At that point what does this mean? Well, think of the monitor control like a water faucet. If the water pressure in the city is low today, you have to open the faucet more to get the same presssure out. If you could mark the position of the faucet (monitor control) , then you can get an idea of how much "pressure" (loudness) is coming into the faucet by the pressure (loudness) coming out and the position of the faucet (control).

Notice how the most squashed material requires you to turn your monitor level control DOWN in order to reach a sane loudness at your ears. In our studio somewhere below -8 or -9 position is the beginning of no return, where transients are obviously degraded or sacrificed for a higher RMS level.

Much more about this in my book :-)

BK

[dahkter]

Hi Bob,
I'm also interested in the K system. I have a rather simple question, how far from the front of the speaker should the test microphone / spl measurement device be?
Thank you,
Jason
PS - I attempted this over the weekend and with the spl meter next to my ear, 83 db was loud.

Thanks.

[Thor]

I'm not Bob, nor do I play him on TV, but let me attempt to answer your question.

The 83dB value (for one speaker, both speakers should give you about 86dB) is measured at the listening position.

So if you have the famous RatShack meter, try mounting it on a mic stand and set it up so it's where your head normally is when you listen/work.

One question this raises in my mind, seeing as this standard originally stems from film, is this: where in (often quite large) movie theatres does one measure when calibrating the playback system? Down front? Middle? Back? Balcony? Average of readings? Or do they have custom calibration routines for that kind of situation?

And one Bonus question for Bob: should a line source be approached any differently than a point source for purposes of establishing a fixed monitor reference point (i.e. K-system calibration)?


Cheers,
Thor

Originally Posted by dahkter

Hi Bob,
I'm also interested in the K system. I have a rather simple question, how far from the front of the speaker should the test microphone / spl measurement device be?
Thank you,
Jason
PS - I attempted this over the weekend and with the spl meter next to my ear, 83 db was loud.

Thanks.

[dahkter]

Thanks Thor,
For someone not Bob, good answer!

I still find 83 loud as hell, especially with the sub turned on to get the lower frequencies. My neighbors agree.

[Thor]

That 83dB reference is supposed to be 'as loud as it gets', i.e. on ffff passages only, which normally only happens occasionally. Of course, with any modern hypercompressed material (i.e., everything at ffff all the time), it can get a bit annoying. Thus, the -6, -9 and even -12 monitor settings.

Sub calibration is separate from monitor calibration. You might want to bandwidth limit your pink noise (from memory I though Bob's samples were already bandpassed...?) and try without the sub. Sub calibration comes afterwards. No wonder your neighbors noticed!

Cheers,
Thor

Originally Posted by dahkter

Thanks Thor,
For someone not Bob, good answer!

I still find 83 loud as hell, especially with the sub turned on to get the lower frequencies. My neighbors agree.

[AlexRoudos]

Well, instead of starting a new post, maybe you could be so kind to help me with a couple of questions i have after calibrating.

When mixing, am i supposed to put a k-meter on the main bus? I produce electronic music(from ambient to trance) so i think i should mix according to the k-14, right?

And if that's the case, what should i see on the k-mixer in order for everything to be correct?

I also do my own ITB "mastering" so the question is quite obvious. In which value do i set the k-meter and what do i go for? I use the UAD Precision Limiter.

Looking forward to your help, again.

[Thor]

Hi Alex,

I think it would be a good idea to get Bob's book, it goes through all of these issues in depth.

In summary: yes, put a K meter at the end of your chain, I think (again from memory, hopefully Bob will jump in and correct any misinformation) you might want to aim at K-20 actually when mixing, giving the mastering engineer room to work. We can always smash it flat if needed, but it's pretty hard removing limiting/compression that's already there....

We mastered several electronica records last year, I think one ended up around K-12, one much louder, but the others ended up around K-14 and sounded amazing (good mixes, obviously). So I would aim conservatively while mixing, but it's entirely up to you. Make it sound good ! The meter is set up so that the loudest passages (ffff - 83dB on one monitor) are in the red, most of the time you'll be in the green or yellow areas, with the chosen 0 reference point being the nominal average.

Hope this helps.

Thor

[AlexRoudos]

Hi Thor and thank you for your reply, it's very helpful.

I've read Bob's book which is amazing but somehow it wasn't all that clear to me, but again i'm very new to the whole K-system theory and probably that's why i didn't get it.

To summarize, you suggest mixing with the k-20 and mastering with the k-14. Does this mean also that when using the k-14 i will have to set the volume of my DAC-1 to the -6db or 79db or not?

Thanx again.

[Thor]

It depends what you want. The nice thing with the K system is that it provides a standardized reference (that works out to be pretty consistent with how we hear).

If you want a 'hot' master to blow people off the dance floor, K-20 or K-14 probably won't do it for you as a master level (unless you have a DJ who can make large level adjustments on the fly). But you never know. A lot of modern stuff is so compressed and limited (so little headroom/crest factor) that it would be about a K-6 (even K-3 with some release), basically a pancake.

If you're looking for something that breathes, has dynamics with punch and clarity, I wouldn't advise pushing the master level much beyond a K-12. IMHO that's pretty much around where things start breaking up. To paraphrase Chris, "your mix has reached it's loudness potential." Every mix has it's potential, beyond which things just get worse and worse. You want to take it to the sweet spot, but no further. It's something that takes experience and time.

So while there are certain guidelines and standards, there are no set and fast rules. I guess that's why there's still a need for mastering .

To answer your last question, yes, you attenuate your monitor controller by -6dB to arrive at K-14, by -8dB to get to K-12 (both as referenced from your K-20 monitor setting, or "0"). Since you're aiming for *consistent loudness* at any K setting, the mixes will have to be processed to give higher levels (and lower headroom) as compared to K-20 material. So it won't be -79dB, it will still be -83dB, you have to compress/limit/whatever to make up for the difference. That's the whole idea with the K-system, consistent levels at different settings for different use (audiophile/classical/film, regular listening, broadcast). The only difference between the scales is the crest factor (headroom). The 0 point is always -83dB on one speaker.

I sometimes explain it to clients as water in the room. The lower the water level, the bigger the waves you can have (headroom). Big waves (transients) are one of the things that give the music punch and also can fool us into believing a recording is real. We can also add more water to the room, raising the level, but the height of the waves will be reduced accordingly, because of the ceiling - it's fixed. If the room is filled almost to the top with water, how big can the waves be? Not very big. Meaning no real punch, no dynamics, no room to breath. Digital audio works in pretty much the same way.

The thing people seem to forget is that with a low average level (water/audio), you can always *turn up the volume* on your stereo, and the dynamics in the audio will be preserved! Try that with Pink Floyd DSOTM and with Ricky Martin "Living the Vida Loca". Turning up the volume on your stereo does not raise the water level in the room, it raises the whole room (water, waves and all), preserving the relationship between the peaks (wave tops) and average (water level). Using compression and limiting in mixing/mastering is the same as raising the water level in the room, which allows only smaller waves on top, i.e. it decreases the dynamics. You give less room for punch and clarity (less room between the water level & ceiling = less room for big punchy waves). The choice is yours, as it's your mixes and your music. Just know what it is that's happening when you choose one thing over another .

Hope this helps.

Cheers,
Thor

Originally Posted by AlexRoudos

Hi Thor and thank you for your reply, it's very helpful.

I've read Bob's book which is amazing but somehow it wasn't all that clear to me, but again i'm very new to the whole K-system theory and probably that's why i didn't get it.

To summarize, you suggest mixing with the k-20 and mastering with the k-14. Does this mean also that when using the k-14 i will have to set the volume of my DAC-1 to the -6db or 79db or not?

Thanx again.

[bob katz]

Originally Posted by dahkter

Hi Bob,
I'm also interested in the K system. I have a rather simple question, how far from the front of the speaker should the test microphone / spl measurement device be?
Thank you,
Jason
PS - I attempted this over the weekend and with the spl meter next to my ear, 83 db was loud.

Thanks.

Normally the microphone should be where your ears are going to be!

I'm not surprised it sounded LOUD to you. There are many factors to consider, and in addition to the type of music you were listening to and how compressed it is:

It's important to distinguish the meanings of "SPL" and "Loudness". It is incorrect to say that something is "83 dB loud" when you are looking at an SPL meter. Loudness is a perception of the ear. That said, the FURTHER the loudspeakers from you and the more diffuse the room, the LOWER the perceived loudness for the SAME MUSIC played at the same sound pressure level. In other words, the higher the transient to average ratio, the greater the transient response of the loudspeakers, and to some extent, the brighter the loudspeaker, the louder it sounds. Nearfields sound louder at the same SPL as mid and far fields. There is a known approximately 2 dB difference in loudness between the large theatre and the home theatre at the same SPL. This is ONE of the reasons why I chose 83 dB instead of 85 for the small room (and I didn't do this in a vaccuum, it was discussed with experts in the field).

So what this means is that "your mileage may vary" and there will be a "fudge factor" of as much as 2 dB or more in where your monitor control ends up in order to match someone else's reproduction of the same music... you may need a slightly lower monitor control position (and thus a slightly lower SPL) to yield the same loudness. Still, the closer we engineers begin to speak the same language, the more we will become aware of what we are doing, resulting in:

1---more consistency in mixes on a project even when done by different mix engineers. They can all agree to mix at, say, K-14, and come a lot closer to each other

2---early awareness that you may be overcompressing simply by observing the position of your monitor gain control, without having to look at a meter at all.

and other nice benefits :-)

BK

[dahkter]

Originally Posted by bob katz

Normally the microphone should be where your ears are going to be!

I'm not surprised it sounded LOUD to you. There are many factors to consider, and in addition to the type of music you were listening to and how compressed it is:

It's important to distinguish the meanings of "SPL" and "Loudness". It is incorrect to say that something is "83 dB loud" when you are looking at an SPL meter. Loudness is a perception of the ear. That said, the FURTHER the loudspeakers from you and the more diffuse the room, the LOWER the perceived loudness for the SAME MUSIC played at the same sound pressure level. In other words, the higher the transient to average ratio, the greater the transient response of the loudspeakers, and to some extent, the brighter the loudspeaker, the louder it sounds. Nearfields sound louder at the same SPL as mid and far fields. There is a known approximately 2 dB difference in loudness between the large theatre and the home theatre at the same SPL. This is ONE of the reasons why I chose 83 dB instead of 85 for the small room (and I didn't do this in a vaccuum, it was discussed with experts in the field).

So what this means is that "your mileage may vary" and there will be a "fudge factor" of as much as 2 dB or more in where your monitor control ends up in order to match someone else's reproduction of the same music... you may need a slightly lower monitor control position (and thus a slightly lower SPL) to yield the same loudness. Still, the closer we engineers begin to speak the same language, the more we will become aware of what we are doing, resulting in:

1---more consistency in mixes on a project even when done by different mix engineers. They can all agree to mix at, say, K-14, and come a lot closer to each other

2---early awareness that you may be overcompressing simply by observing the position of your monitor gain control, without having to look at a meter at all.

and other nice benefits :-)

BK

Hi Bob,
Many thanks for your reply, have a good day.
Thank you,
Jason

[AlexRoudos]

Thank you Thor once again for your invaluable help. Now i get it a lot better.

If you don't mind i have a last question.

When "mastering"(until now without k-14) i use a sequence of UAD plugs which are the Fairchild 670, the Helios 69 EQ and the Precision Limiter.

I don't compress much, usually i apply around 4-6dbs of compression and the make up gain i apply is exactly the value needed for the mix to sound the same when i turn the comp on and off. I prefer to add gain using the Helios 69 EQ instead of the 670's make up gain. So, this brings us to the Precision Limiter which from now and on will be set to K-14.

And here's my question :

What should i read on the meter and what should be the settings on the limiter? Where should i set the output ceiling and what's the peak and average attenuation i should read?

I hope these questions make some sense. I have to to mention also that i'm using this sequence for quite long now for "mastering" my music and the results were always great. I do have some descent experience and good ears i hope, but now with the K-System i feel totally like a newbie.

Thanx again.

[PatrikT]

Originally Posted by AlexRoudos

I don't compress much, usually i apply around 4-6dbs of compression <SNIP>

Then you are probably not mastering either...


Best Regards
Patrik

[AlexRoudos]

I totally agree with you Patrik,

I never said that i'm mastering but that i'm "mastering".

[PatrikT]

Because that kind of compression is crazy much even with the lowest ratios...

[AlexRoudos]

Originally Posted by PatrikT

Because that kind of compression is crazy much even with the lowest ratios...

For some reason, i always thought(from what i was gathering) that 4-6dbs of compression is on the low side and i also never thought to search and ask to see what the pros think. Surprisingly the results were always sounded good.

Anyway, i have a lot to learn in this matter. BTW, what is the recommended average amount of compression for electronic music in the mastering stage?

Thanx Patrik.

[bleak orange]

Ok,...I think I've got it now

But I ordered the second edition of Mr. Katz book anyway. I could probably learn a thing or two...

Cheers !

[bleak orange]

Ok, So i got the book. 2nd edition. read through it...got more confused...

correct me where I am wrong

1) put a spl meter with a microphone in your listening position. Have it facing up ^ towards the ceiling. and set it to "slow" and "C"

2) Put on a pink noise file in your daw, or in my instance, the PSP Xenon limiter ( which generates pink noise )....It has 3 meter switches : k -20, k - 14 and k-12. The meters change according to which meter is engaged. to the right of these switches it has a "left" knob, a "right" knob and a "l+r" knob. Lets say I want to monitor at k-14, I select the K-14 meter, and I press the "left" button. Pink noise comes out of my left monitor. etc...

3) I have a Grace design 904 monitor controller and I start cranking the volume upwards until it reads 83 on the spl meter ( 84 on my Grace design, which goes from 0 ( silence ) to 100 (ear bleed) )The Grace is calibrated in 0.5 db steps, and it may be the cheapo SPL meter that is a db off, or it might be the room

4) Now I repeat the procedure for the right channel, and the spl meter reads the same value. i.e. 83, when the Grace design is in the 84 position. when the pink noise comes out from both the monitors, the spl meter reads about 86

5 ) now what ? should I yank the volume knob down on my Grace design until the spl meter reads 76 ? ( 8 db below 83 ) or should I leave it at the 84 position ( which is uncomfortable to listen to ) and turn down the volume in my daw ??

6) and when I work, should I aim to have no peaks above the "0" in the K-12 meter ?? Even if it goes up to + 14 ? Or may I allow the most violent passages go up in the red ( from + 4 to +14 ) ??

7) What about the sub woofer in all this ? I had it turned off when measuring, is there anything to consider when turning it on ? IS there a special way to measure a sub woofer to fit the k-14 system ??

8) I still find 73 db ( when measured with the spl meter ) a bit too loud to mix \ monitor on...the pitfalls of monitoring at a lower volume is that one overcompress right ?

9) How much lower then 0 ( on the k 14 meter) can you go ? ride the faders to taste ? or is there a general idea that the lowest part of a song should be no more than x db lower than 0 on the k-14 scale \ meter ?

Finally, am I right in assuming that:
If you have a digital "K-meter", a meter that is calibrated to either -12, k-14 or k-20 and stick with it when working, it doesent matter how many db you're monitoring at ? The position of your volume knob is irrelevant to the metering, as long as you rigidly follow the meters on your "K-meter". set your monitor controller to a comfort level and let the meter be your guide ??

and secondly:
The reason for calibrating your monitors to a 83 db reference has to do with how the ear operates ?

tia

[bob katz]

The BIG (only) problem with the Grace is it encourages people (like you) to get very confused between the concepts of SPL and GAIN!

Get Mike Grace to relabel his unit so that the label reads 0 dB when the monitor gain yields 83 dB SPL with the standard source. This should greatly reduce your confusion.

Becky and Fred were having an argument. Becky said she always listens at 83 dB on the controller, but Fred measured her SPL and it turned out to be 90! Now, who's right?

The answer is that monitor controllers marked in SPL are downright misleading and confusing!

This is like labelling master faders in a mixer in dBu, with +4 written on there somewhere.

Originally Posted by bleak orange

Ok, So i got the book. 2nd edition. read through it...got more confused...

correct me where I am wrong

1) put a spl meter with a microphone in your listening position. Have it facing up ^ towards the ceiling. and set it to "slow" and "C"


SNIP



5 ) now what ? should I yank the volume knob down on my Grace design until the spl meter reads 76 ? ( 8 db below 83 ) or should I leave it at the 84 position ( which is uncomfortable to listen to ) and turn down the volume in my daw ??

As I said above, the labels on the Grace are very confusing. For a K-14 recording, you will need to turn your monitor gain DOWN closer to about 6 to 8 dB below the original calibration point. NOT the "volume" in your DAW, if you are mixing to a true K-14.

6) and when I work, should I aim to have no peaks above the "0" in the K-12 meter ?? Even if it goes up to + 14 ? Or may I allow the most violent passages go up in the red ( from + 4 to +14 ) ??

When you say "peaks" are you referring to the peak level of the material as opposed to its RMS level? The answer is to let peaks fall where they may, there is NO obligation to peak to full scale, but more importantly, use the meter as a guide to the crest factor. No obligation to "normalize" the peak to full scale if you are mixing.

Perhaps your crest factor will be no more than 10-11 dB... if your mix is good sounding that's it. In that case, your average level on forte passages will come around 0 on the K-14 meter and your peaks will fall, say, around +10-11 dB, not hit the top. The whole idea is that it is an artificial concept to expect "all" your peaks to hit the top of the scale. That could only be achieved obtained with a peak limiter unless you are clipping the crap out of the mix.

7) What about the sub woofer in all this ? I had it turned off when measuring, is there anything to consider when turning it on ? IS there a special way to measure a sub woofer to fit the k-14 system ??

I really do go into far more detail about this in the 2nd edition of the book than this questioner implies. A good re-read of chapter 14 and 15 will answer ALL of his questions.

But to summarize:

First adjust the subwoofer for flattest response and best splice to the mains.
Then note (as described in my book) that using Wideband pink noise for monitor level calibration and a C-weighted meter can cause some measurement errors, up to a dB or so. If you turn your subwoofer on and off at that point you will see a minute difference in the measured level. At that point you can either ignore the error and proceed to getting good mixes without any serious worries, or you can get more serious about being even more accurate!

Which means to be more accurate, you'd need to get a better, more calibrated SPL meter, and also use a narrow-band pink noise (500-2 kHz) RMS calibrated to -20 dBFS. Then the level of the subwoofer is totally irrelevant to your reading.

8) I still find 73 db ( when measured with the spl meter ) a bit too loud to mix \ monitor on...the pitfalls of monitoring at a lower volume is that one overcompress right ?

73 dB setting on the Grace? Play some of the CDs mentioned in the Honor Roll at Digido.com and see how closely you come to the numbers there (making a mental subtraction, that is, 83 position on the Grace is 0 dB on your monitor gain). Are you within, say, 2 dB? If you are, then you are probably overcompressing, because 73 dB position is -10 dB on your monitor. If your mixes are LOUD at that position, you are producing mixes that are already at MASTERED level for many current contemporary and overcompressed mastered CDs.

Take a look at page 175 in the second edition to see what I'm getting at.

9) How much lower then 0 ( on the k 14 meter) can you go ? ride the faders to taste ? or is there a general idea that the lowest part of a song should be no more than x db lower than 0 on the k-14 scale \ meter ?

Again, this is covered very well in chapter 14.

Every song is different, generally you need to find your monitor position, which will be somewhere around -6 to -8 (or higher) if you are trying to produce good sounding, open mixes that will later be mastered. You can really ignore the meter at this point, that's the whole point, a fixed monitor gain is liberating! Use your ears at that point and ignore the meter.

This corresponds with about a K-14 or lower level mix.

For a secondary reference on the meter, the forte passages will fall around 0 dB, fortissimo as much as (no more than, recommended) +4 dB, and softer passages as far below 0 dB as your ears tell you you should go!

Finally, am I right in assuming that:
If you have a digital "K-meter", a meter that is calibrated to either -12, k-14 or k-20 and stick with it when working, it doesent matter how many db you're monitoring at ? The position of your volume knob is irrelevant to the metering, as long as you rigidly follow the meters on your "K-meter". set your monitor controller to a comfort level and let the meter be your guide ??

Your comfort level IS your guide, but I'd be concerned if your mixes sound loud to you at a -10 dB monitor position. Which is why I recommended you check out the Honor roll and see if you're within about 2 dB of that. If you are, then I think you may be mixing too hot, or overcompressed. But if the honor roll universally sounds too loud to you, then one of three things can be happening:

1) either you miscalibrated your monitors

2) OR your ears like to listen at lower levels Or

3) You're listening nearfield with the monitors too close to your ears.

and secondly:
The reason for calibrating your monitors to a 83 db reference has to do with how the ear operates ?

Yes, that's also covered in the book :-). 83 is approximately the level where the ear is most linear. If you listen within about 4-5 dB of that SPL either way, you'll probably make the most balanced mixes with a bass response that will translate elsewhere. The softer you listen, the more likely your bass will be too heavy elsewhere, and vice versa.


I hope all this helps. A good re-read of chapter 14 should help also. But much of this becomes a simply abstract concept for a lot of people until they have tried it, used it a bit, and then the lightbulb comes on, duh!

When I first discovered the strong relationship between degree of compression and monitor gain it was a very enlightening moment. In the 20th century, we set our VU meters to 0 and adjusted our monitor gain second. In the 21st century, in a calibrated room, we should adjust our monitor controls first to the position we are going for, and then mix, virtually ignoring the meters! Note that good film mixers have been following this practice for years!

Hope this helps,


Bob
BK

[Thor]

Hi Alex,

sorry I didn't see your post until now. See what happens when I spend all my time working instead of hanging out on forums?

Sounds to me like your 'not compressing much' is actually quite a bit. But YMMV, if you're happy and your clients are happy, more power to ya.

I'm not familiar with the UAD plugs, can you set the limiter output to a K scale? I thought only the PSP Xenon had that.

On to your question....

Output ceiling has nothing to do with K system, set it to whatever you usually use. Some (me) use -0.3dB, others -0.01, still others 0.0.

What you should read on the meter is explained in Bob's book. The green area up to 0 is where the bulk of the music should be. Crescendo's work up into the yellow and the loudest passages hit the red now and then. However it will always depend on your source material. If you're mastering a black metal band, you might find yourself in the red from start to finish. If you have a quiet acoustic jazz guitar piece you might never hit the yellow. It all depends. The peak and average attenuation will also vary, depending again on the source levels.

The advantage of the K system is that it allows you to work against a known reference - actually several. 83dB SPL at listening position for the loudest passages combined with a known and predefined amount of headroom (crest factor). This puts the music we're working on in the loudness 'sweet spot' for our hearing (auditory perception is not linear), not too loud and not too soft, so we can make accurate judgments and adjustments. It also gives us a known amount of headroom that has been found appropriate for a wide range of musical styles as well as applications (home theatre, audiophile audio, broadcast, muzak, etc).

The more you work with it, the easier it is to judge loudness (and quality, at least as related to headroom and dynamics) and to dial in something good. At least in my experience. As Bob notes in his book, a similar system has been in use for decades in the film industry. When was the last time you heard a film that had really poor sound, or was too loud/too soft? Exactly!

We're all newbies at times in different areas, no problems at all.

Hope this helps.

Cheers,
Thor

Originally Posted by AlexRoudos

Thank you Thor once again for your invaluable help. Now i get it a lot better.

If you don't mind i have a last question.

When "mastering"(until now without k-14) i use a sequence of UAD plugs which are the Fairchild 670, the Helios 69 EQ and the Precision Limiter.

I don't compress much, usually i apply around 4-6dbs of compression and the make up gain i apply is exactly the value needed for the mix to sound the same when i turn the comp on and off. I prefer to add gain using the Helios 69 EQ instead of the 670's make up gain. So, this brings us to the Precision Limiter which from now and on will be set to K-14.

And here's my question :

What should i read on the meter and what should be the settings on the limiter? Where should i set the output ceiling and what's the peak and average attenuation i should read?

I hope these questions make some sense. I have to to mention also that i'm using this sequence for quite long now for "mastering" my music and the results were always great. I do have some descent experience and good ears i hope, but now with the K-System i feel totally like a newbie.

Thanx again.

[compusic]

Originally Posted by bob katz

3---Using an SPL meter set to C-weighted, SLOW, adjust the gain of your amplifiers or DAC to yield 83 dB SPL at the 0 dB position of your monitor gain control. That's right, your monitor gain control needs to be marked in dB, or this whole idea falls apart. That's because the 0 dB position will be used very rarely.


BK

Hi, may I have one more question in regard to the step 3?

I use self-powered monitors, and their input gain control is set at +4db position, as recommended by the manufacturer. between the Pro Tools 96 I/O and monitors, there's a monitor gain controller.

The controller works at +4db with 18 db of headroom. The 96 I/O's output ports work at +4db level with 14 db of headroom. but I don't know the monitors' headroom. in fact, I did e-mail to the manufacturer for this, but no reply.

And as for the monitor gain controller, there's no labels such as 0db, -6 db ,etc. only 0% to 100% in steps of 5%. I have asked the manufacturer about the controller's unity gain position, kind of like the 0 db position of a mixer fader, also asked its input nominal level. below is their reply:

------------------------
"the volume control cannot be reagrded as a mix fader control. It does
not have a unity gain position. As with the large consoles like SSL
or Neve the control room monitor level potentiometer is calibrated to
about -6dB fully clockwise. That is to have a more usable
potentiometer range at normal listening levels.

The controller's nominal input level is +4dB.

You may find the following steps useful to set-up the controller:
1. turn the input sensitivity of the speakers or the amplifier max
(fully clockwise).

2. Check the position of the controller's volume control at e.g. 48dB, 64dB and 84dB (using an SPL meter at the listening position).

3. If the control range of the controller's volume control at low listening
level is close to the fully counter clockwise position, reduce the
input sensitivity of the spaekers or amplifier until the volume
control ranges between 9-10h at low monitoring levels and 3-4h at
high monitoring levels."
-----------------------------

I have done a loop test, and find that when the controller's volume control is at about 85% position, the controller will output the same level as the signal enters its input port. at 100% position, output level is 2.5 db higher.

the situations above confuse me a lot. what should I reach for to get the caliberated level? monitor's input gain control? or the controller's volume control? or both?

and btw, any more info in regard to the manufacturer's reply above?

thanks in advance.

[bob katz]

Originally Posted by compusic

Hi, may I have one more question in regard to the step 3?

I use self-powered monitors, and their input gain control is set at +4db position, as recommended by the manufacturer. between the Pro Tools 96 I/O and monitors, there's a monitor gain controller.

Hi, Compmusic, ask away! The "+4 dB" position of the self-powered monitors refers to their input sensitivity in dBu. In other words, they will produce a given SPL when fed a certain voltage level. The marks on this fader are largely irrelevant and you should use the input gain of the self-powered monitors to calibrate your SPL when your monitor controller is sitting at the calibration position.

The controller works at +4db with 18 db of headroom.

That's not enough headroom in my opinion. Which controller was that? I think you should run the controller LOWER and the input gain of the powered monitors higher.

The 96 I/O's output ports work at +4db level with 14 db of headroom. but I don't know the monitors' headroom. in fact, I did e-mail to the manufacturer for this, but no reply.

This is the Pro Tools I/O? I strongly doubt they only have 14 dB of headroom. If connected balanced in a fully-balanced environment they should have about 20 dB or more of headroom if running at +4 dBu with a sine wave signal at -20 dBFS.

And as for the monitor gain controller, there's no labels such as 0db, -6 db ,etc. only 0% to 100% in steps of 5%.

That puts you back in the 20th century in my parlance. Monitor gain controls marked in anything but dB are just passé :-)

As for the rest of the discussion from the manufacturer, I would appraoch it instead from the point of view of headroom and signal to noise ratio. With the DAC connected and locked, and no audio signal going through, and with the monitor controller at a nominal position, say, at 80% on the control, turn the gain of your monitor speakers up ALL the way. If you hear hiss, then that's too much. Turn the monitor speakers gain down until you cease to hear the hiss. Then move the monitor controller up and down and if you don't hear any hiss at any of its positions, you're "golden". This will maximize the headroom of the monitor controller and optimize signal to noise ratio of the system.

Then try the standard -20 dBFS per channel pink noise test signal and the SPL meter and if at some position of the monitor near its top you can get the 83 dB SPL, then you have optimized your system. This position on the monitor controller you call 0 dB.

Then comes the fun. Get a Dymo labelmaker. Mark this monitor position as 0 dB. Disconnect the louspeakers and connect a calibrated VTVM instead. Send a 1 kHz tone out of the DAC and adjust its digital level until you read 0 dBu analog. Then reduce the monitor gain a dB at a time until you have 1 dB marks down to about -14. And now, presto, you have a calibrated monitor control!

BK

[flatfinger]

Thanks for your incredibly helpfull contribution to audio!!

[flatfinger]

Thanks for your incredibly helpfull contribution to audio!!

[compusic]

Thanks for reply. but, I need to make somethings clear.

Originally Posted by bob katz

The "+4 dB" position of the self-powered monitors refers to their input sensitivity in dBu. In other words, they will produce a given SPL when fed a certain voltage level. The marks on this fader are largely irrelevant and you should use the input gain of the self-powered monitors to calibrate your SPL when your monitor controller is sitting at the calibration position.

as for the controller's "calibration position", do you mean unity gain position?

Originally Posted by bob katz

That's not enough headroom in my opinion. Which controller was that?

that's a surround monitor controller, model 2489, from Sound Performance Lab. Its manual says clearly that its maximum input level is +22 db.

Originally Posted by bob katz

This is the Pro Tools I/O? I strongly doubt they only have 14 dB of headroom. If connected balanced in a fully-balanced environment they should have about 20 dB or more of headroom if running at +4 dBu with a sine wave signal at -20 dBFS.

Yes, 96 I/O is Pro Tools HD's audio interface, also a ADC and DAC. and its manual also tells clearly that it has only 14 db headroom above working level +4 db, balance connected.

Originally Posted by bob katz

If you hear hiss, then that's too much. Turn the monitor speakers gain down until you cease to hear the hiss.

but at what listening distance? I can hear the hiss when I put my ears very close to the speakers, even when I turn the input control on monitor all the way down.

Originally Posted by bob katz

Then move the monitor controller up and down and if you don't hear any hiss at any of its positions, you're "golden".

though the monitor has hiss just as I said above, the surround controller's control does not effect the hiss level. well, here I can suppose that the hiss is not due to the output of the controller.

Originally Posted by bob katz

Then try the standard -20 dBFS per channel pink noise test signal and the SPL meter and if at some position of the monitor near its top you can get the 83 dB SPL, then you have optimized your system. This position on the monitor controller you call 0 dB.

as for the last sentence "This position on the monitor controller you call 0 dB.", do you mean try to find the monitor's input control position without changing the surround controller's position, in other words, leaving the surround controller at unity gain position, and find monitor input control's position which gives the 83 dB SPL?

Originally Posted by bob katz

Then comes the fun. Get a Dymo labelmaker. Mark this monitor position as 0 dB. Disconnect the louspeakers and connect a calibrated VTVM instead. Send a 1 kHz tone out of the DAC and adjust its digital level until you read 0 dBu analog. Then reduce the monitor gain a dB at a time until you have 1 dB marks down to about -14. And now, presto, you have a calibrated monitor control!

above is what I have difficulty to understand.

as for "Mark this monitor position", do you refer to the surround controller's position? then it's the unity gain position.

why need to find a position -14 db below? well, let me describe the scene just before this step, the surround controller is at unity gain position, and monitor's input control gives the 83 dB SPL, and what the monitor receives is a -20 dBFS pink noise. It seems that the noise signal has already reserved 20 db headroom. why need another 14 db?


thanks in advance.