Hey, I found some interesting information on K-Metering http://www.meterplugs.com/kmeter  (the information is at the bottom right of the page) I've been researching this a little bit in the last week because I put up a master that was a little hot on the song forum.  What I found most interesting was the different calibration levels and corosponding meters. IE K20 was for broad dynamic music like classical and jazz.  K14 was for rock and K12 was for broadcast and pop.  After watching one of the videos on the page which I've posted a link to, the studio engineer said a master should touch 0 on the K-Meter.  I'm wondering if anyone else is this ridgid when mastering with the K-Meter.  I've pulled some tracks by Bowie into Audio Creator and they are all way over and are crushing the +6 mark on the KMeter, The Beatles are also hitting the +6 mark but not crushing it.  I've always felt as long as your track was between 0&+6 on the Kmeter you had an industry standerd track in terms of loudness.  Whats everyone elses opinion on this??
 
Ben

Ben I find that while you are tracking and mixing yes you can work with the standard K metering reference levels eg K -20, K-14 and K-12 for example.

Mastering is whole different ballgame however. A typical loud Hip hop CD for example might be averaging around K-7 or K-6 as you say. When I am in mastering mode I use a special preset in my digital mixer that alters all the calibration so that the music needs to be say up around K -7 to make the meter read 0 dB VU.

This means that an unmastered mix now will fall short and only read to some lower value. eg If I have worked at K -14 for example during mixing and then I switch over to my K-7 mastering setup the meters will only make it to say -7 dB on the scale hence I have to pull things up using the usual mastering techniques like EQ, compression and limiting to get the music back up to 0 dB VU. But then I know when it does get there I am reaching K-7.

I usually get the client to give me their fave ref CD as I am mastering. Then I load it onto a track while the system is set for K-14 say. It will slam the meter of course but then I pull the track down by the required amount to get the meter to read nicely at 0dB VU. The amount I have to pull the track down is what you have to strive for in mastering. So I if I have to pull a track down by -7db to make that ref CD read nice while the system is at K-14 then I know I have to add that much to K -14 for the client to be happy.

Beatles don't count. Their stuff was recorded pre-digital age and later remastered, in some cases brutally crushed (see the abomination that used to be George Harrison's masterpiece "All Things Must Pass"). The best references, IMO, are digital recordings from the 90's. That's about when they hit the sweet spot, I think. 

Saying you MUST hit 0 on the k-meter is overly dogmatic. In all likelihood, if your music has any dynamics in it at all, you're gonna hit it and blow right past it on the loudest parts, and be well below it during the quiet bits. 

Dave is a bit incorrect there. With well mixed and mastered music there will be an average level that is present for most of the time (or the loudest parts) and this should just reach 0 dB VU on the meter, everything else falls below that but if its nicely mastered it won't fall that much below at all. The music should rarely go above 0 dB VU and if it does it is only very briefly and it might reach +2 to +3 dB at the most.

And this applies whether the ref level is K-14, K-12 or a higher mastered ref level such as K-7. Mastering is going to tame the dynamics down a bit and prevent the meter dropping too low anyway.

I am also referring to either real VU meters or VU meter plugins. I am not a fan of the type of meter Ben has pointed to in his first post. I am not sure they exhibit the same ballistics as a real VU meter or VU meter plugin. The reason why real VU's or good VU plugins are good is that they are inherently slower (300ms to reach 0 dB VU) and that is a very important quality. Because of this they tend not to swing around so wildly.

If a real VU does swing around wildly it means there is something wrong and one needs to investigate it as to why. There is usually a good reason.

Thanks guys, I was trying to post some sort of screen shot but to no avail.  I think I am in the ball park, I use the span meter and I've always had the meter pushing +3 to +6 depending on the type of music.  Of course the average sits around about 0 but as Jeff suggests, it doesn't just sit there though, there will be peaks and troughs of -3 +6.
 
Ben