K-System - Normalize before entering Sonar ?

Be wary the K-System refers to the average volume level at 83dB SPL at the 0 decibel point. The K System was developed by Bob Katz as an alternative to the varying RMS measurement standards.
 
Let me explain, the K-20 reference was designed for classical,soundtrack and music without a lot of loud transients like you get in dance music or rock music. So the -20dB refers to the average volume level at any given time rather than the instantaneous VU reading. 
 
When you refer to the meters using the K-system when the bars cross the 0 point, this is where it's exceedig -20dB RMS. However if your VU meter may be peaking at -3dB which is a tell-tale sign of lack of dynamic boost in your overall mix. It also allows you to judge the VU volume and RMS in the same place.  So adjusting the EQ or volume of a certain track might be enough to fix that. Limiting should be last thing on the agenda.
 
I'm rubbish at explaining things however I hope that helps.
 
Personally I never normalise sounds unless it's a final product not mid-project. 
 

I don't normalize tracks outside of Sonar.
I prefer the gain nobs to strict 'normalization' using process > audio > Normalize.
 
I do use process > audio > normalize sometimes. This does not normalize to RMS or VU. It normalizes to peak.
 
 
 

all it does is move the waves up to where the peaks all hit the same spot (that you dictate, all the way up to 0db)
 
it does not provide any advantage at all,
except to bring up the noise floor; if there is noise on your tracks at low level,
it brings that up too.

"Normalization" has a broader definition than just raising everything until the highest peak is at some target level. Broadly, it means to make things conform to some standard. You can normalize to any target: peak, RMS, LUFS, K-20 or any other reference you like. 
 
Whether it's worthwhile pre-normalizing files prior to mixing is another question. It could be a time-saver as Jeff suggests, but it depends on how wildly off they are. If they're all reasonably in the ballpark, I wouldn't bother. It's not that hard to normalize within the DAW if that turns out to be necessary.

interpolated
http://www.meterplugs.com/kmeter
 
Watch the Freemasons video. Makes interesting viewing.
 

Watch out using bar graph meters for showing you rms levels. A VU gives you something else apart from the rms level.  And that is the ballistic.  ie how the needle dances to the music.  There is much to be gained from learning how to read that.  Bar graph meters don't show it anywhere nearly as well.
 
Real VU's do it the best in terms of ballistics but I must say the VST's are getting very good.  The Klanghelm meter has got a very good ballistic.  I have devised some special tests for showing VU ballistics and some VST's just dont cut it especially in how they fall back.  But the Klanghelm meter is excellent.  Not perfect but damn close.  Also PSP has released a nice VST as well. (Triple meter) and the ballistics have improved in the VU meter I am sure of it.
 
 

Sonar show rms levels as being 3 db lower than what they should be which is a bit weird.  Normally rms is 3db down from peak and Sonar is doing that correctly.  But the standard for rms readings is that the rms reading of a continuous sine wave should be the actual peak value of that sine wave, not 3 db down.
 
The main issue is that the DAW rms meter reading is way down on the scale so you are not getting the full scale reading that a VU meter has around 0 dB VU.  Some DAW's (eg Studio One) allow the bar graph meters to go into K System mode so then they jump up to around 0 db VU and all the headroom is above that now.  It is not a bad indicator but it still misses out on the ballistics concept though.
 
Yes I use real VU's all the time.  But that is not an option for many.  It requires the purchase of decent VU meters and they are not cheap.  The proper electronics to drive them too.  And you have to mount them in a box etc so it is all a bit fiddly.  You can buy them too fully made up in a box etc but they will be expensive.
 
The real VU's are something else again though.  Hard to explain it.  They dance in the ultimate way.
 
Hey this is cheap:
 
http://www.ebay.com.au/itm/Custom-VU-Meters-3-Large-Simpson-VU-Meters-/111650225948?pt=LH_DefaultDomain_15&hash=item19fedf3f1c
 
Problem is it is in Australia just around the corner from me. I should check this out myself! Except it is US voltage eg 120V so it would be perfect for you guys.

 
I don't see the size of those meters, but I wouldn't expect much.  And you'll need the circuitry.
 
I'd go with something more like these :  Simpson VU meters
 
I've caught some pretty nice NOS meters for great prices if I waited for the right moment.
 
You can also remove them from old equipment that you can get for cheap.
 
EDIT :  Hey, did one of us just grab those meters Jeff linked to ?  Jeff ?
 

Thanks Jeff & Paul

I have the Blue Cat meter and it is great to have and tells you a lot. I don't use it much now though. I have found it is good to work in conjunction with this too:
 
http://www.pspaudioware.com/plugins/tools_and_meters/psp_triplemeter/
 
This covers all the VU movement that the PSP meter does not have.
 
The free Orban Loudness meter is good too.
 
http://www.orban.com/meter/
 
 

I like to change K ref levels.  I like working between K-14 and K-20.  I think there is an art to reading a VU meter.  The real ones seem to make a fraction easier.
 
When I am doing a mixdown of a pre master I will go for a total reading on my VU's so the needle is just reaching 0 dB VU for most of the mix.  Right now I am doing a mix for guy and I have decided to make K-14 my ref level on this.
 
The VU needle will read low from  0 dB VU and sometimes goes a little over but not wildly go over at anytime. There is some point where the needle tends to hover and be a lot of the time.  That is the point to get around
0 dB VU.  After I replay this I get an identical reading back and there should be no reason why it should not.
 
A bit of the skill is in getting the needle to averaging around this 0 dB point.  You may be setting it a little low if there is wild swinging.  You tend to limit the upper level of wild swinging to around 0 dB VU instead.  But then the average level ends up a little low.
 
When a mix is doing well the ballistics of the needle seem to be a certain way where it just hits 0 dB VU and not go much over but the transients are still present and punchy.  Then you can creep the rms level of that needle movement to being right on 0 dB VU.  (you can practice looking at great mixes on VU's such as Steely Dan's 'Everything Must Go' CD)
 
Wild swinging of the VU at mixdown means something is responsible and it usually only one thing as well.  Just have to track it down.  Do some editing, reset any dynamics processing on that track and you are away.  The wild swinging in the main mix has gone now.  Now the VU is showing much more of average level and you can adjust it now to sit right up at 0 dB VU.  You can easily add 3 to 4 dB of rms gain once wild swinging of the VU is under control in a mix.