yeah, what's up with green?

Jeff Evans

If you play a CD that has only 5 db of dynamic range on a K system that is calibrated for -20 db FS to give 85 db SPL then the track will be around 15 db louder than a normal K -20 level would be. If you have VU's calibrated to show K -20 at 0 db VU then the VU's will go off the scale and the track will be loud yes.

But if you have imported a track like this onto a Sonar track then you have to pull it down by 15 db for the CD to show 0 db VU again and you are back to K -20 and the level is 85 db SPL again back in the room.

So no matter what you play the level will always be 85 db SPL in the room. 85 db is 85 db and the level will always be the same. What varies is how much one has to pull the track fader down in order to get back to K -20. If its 15 db then you know the track is averaging K -5 which is loud on the CD. 

That's not the case. If that were the case, then the whole K-System would be irrelevant. What matters is the average level, since that's what we sense as loudness. You can have a song with a snappy snare that's hitting at +5dB on the k-system peak meter, but everything else is quite low. That won't be very apparently loud. But if you compress that snare and then bring up the level to keep the peaks the same, the average level goes up and it gets apparently louder.
 
That's the whole point of the K-system. The most you compress, while keeping peaks at the same level, the louder it is apparently, which drives you to back off on the compression because it gets uncomfortably loud. Well, given that all music today is grossly over-compressed, by definition, they should all be very uncomfortably loud or the whole system is not going to do what it's designed to do. If you put on Jesus of Suburbia and it doesn't horribly loud, then nothing that is that over compressed is going to sound horribly loud, and therefore you'll have no real incentive to reduce the compression. 
  
If you import a commercial track and drop the fader to -14dBFS (assuming you do k-14), then by definition the peaks will never go above 0dBFS k-system, but that track (if it's a typical modern track) will be extremely loud because it has a very high average level.
 

You have to first distinguish between music that has been produced (prior) to mastering using the K system and what happens to that after. (eg after mastering) Many people don't fully understand it (K system) and it takes a bit of digesting. Bob's book explains it well though.

So if we work at say K -14 and track and mix we will end up with a track that has an rms (or average) level if -14 db FS. One could master as well and not change that average level at all and just leave it. (As I pointed out String has done in many parts of Brand New Day, not all though he does go above it at times but things often come back to it)

But as we know people are not prepared to leave a track at that level and they want it louder. But what Dean is talking about is how the SPL level in the room will effect your decisions to compress limit etc and that is the better way to look at it. So if I want my track now to get say 6 db louder (Average K level is now - 8 db which is getting up there) then the corrcet thing to do is lower the monitor level in the room by 6 db (before mastering) so we are hearing an SPL level of say 79 db SPL. But through compression etc we can get that music 6 db louder and hence things start to come back to normal in the room although it will start to sound a little compressed. If you want to achieve the same level as say a Black Eyed Peas CD then you first need to put it into the K system to see how far above it the track is, eg it might be 10 db louder than a K -14 track. So we need to lower the monitor level in the room by the same amount. You can go the other way too. When I mix I tend to push the levels in the room another 3 db to make it louder. It makes you do less to your mix. Big mistake many people make is having the monitor level in the room too low. That is why they push things into the red and over compress. By monitoring louder you end up backing everything off and there are no overs and you use less compression in your mix too. (or you use it more for the effect rather than getting a volume boost)

You do need to mix at various levels in the room but what levels? At least K gives us a starting point of 85 db SPL and the easiest way to achieve it is through a cheap SPL meter. Also K gives some standard and uniformity to work with regarding track and buss levels and this is where people can get confused.  The concept is to keep the average (or rms) levels constant on tracks and busses and allow peaks to vary above that. On some tracks peaks might only be a few db higher than the rms level, on others it could be as high as 10 db.
With analog we had a frame of reference and it was called 0 dbu. We used it everywhere in maintaining constant levels going down through channel strips and going in and out of multitracks etc. But with digital we are left a bit stranded with our levels and there are so many ways of getting completely different levels on our tracks, peak and rms. And there is not even a standard if I was to get a track from one of you for example it could be anywhere! But at least with the K system there is some frame of reference. eg If I was working with Dave or Dean on a track and we agreed to work at say K -14 then if I imported an overdub or something from them it would least be at the right level. Peaking 0 db VU on my VU's and I would not have to do much to it in terms of making it match everything else on my other tracks.

Loud rms levels and great transients cannot really exist at the same time and you have to compromise with one or the other. (Although there are some plugins coming out that claim to make tracks loud and keep the transients intact) It will be good to check some of those out and see how they stack up)

Anyway sorry to get into a big K systemn rave but it is well worth looking into. The OP was about how loud we should be mixing and I think we all agree that 85 db SPL is a good place to start. But if you cannot monitor safely at 85 db for any reason you are making slight comprimises and having to adjust a bit and I am confident that people like Dean can do it. But it is better to have a wider range of volume levels at your disposal including 105 db SPL if you need to.

Jonbouy
And do I want to be monitoring to some standard SPL when working with a string quartet one day and a full-on 'banger' of a dance tune the next? 

There are three different levels for that reason. And remember, the actual volume in the room is related to the average level of the mix, which I keep trying to get across. The average level of a string quartet will be considerably lower, so the effective level in the room, other than in the biggest swells of the music, will be considerably lower. But that full SPL is there for those big swells, so that they do sound big and impressive as they should. The string quartet just won't use that full level constantly, whereas the dance number may be close to doing so most of the time.
 

I remember the very first time I pushed a set of woofers out of their frames and got them to go plop.

The crowd went wild. Do it again! Do it again!

still boils down to common sense, guys.

and they don't teach that at harvard!

WHAT??!!! SPEAK UP!!!!  I CAN'T HEAR YOU........

still boils down to common sense, guys. 

 

 
The problem is that common sense is not very common.......
 
 
Just some OT general thoughts.
People are different in many aspects of life, some rely on hard facts/numbers, others use their intuition.
 
BTW an interesting thread.
 
Best, JH
 

compression..but not to the point of loss of dynamics.its a big game of trial and error my friend..you can try and absorb all this scientific masters degree stuff..but when it comes down to it..you need to raise the floor while keeping the noise low and the signal high..you need devices

Well, the first thing to consider is that 'loud as possible' isn't even a goal anyone should want to aspire to. It's pretty much guaranteed to degrade the quality of whatever it is you are doing.

The very first thing to look at is to just bounce the mix to a stereo track. Look at that track and see if you have any particularly bad peaks. It's not uncommon that you'll find that in a specific place a number of things just happen to fall right on the money together and it makes for a particular big peak. If you run the song and set the volume so that the highest peak is just below 0dBFS, then that one peak will set the maximum level.

If you have that kind of issue just use a little automation to bring that one peak down in the original tracks. It can be done very transparently without using any compression. That in and of itself can mean you can bring the average level up some number of dB in a lot of cases with really zero side effects.

Then bounce it again and go back and see what else is contributing to the highest peaks. It's often the snare, or kick or something like that. Look if you can get the level of that one thing down without losing the impact you want. That might let you get the average level up a few more dB without affecting anything but one track. If it just has a few overly high peaks here and there, again try to some automation on just those peaks. If it's all the way through, try a compressor.

Then repeat the above if there are other things that you can apply the same. Try those types of things before you slap a limiter on the master bus and crush it to death. You can often get the average level up pretty nicely while keeping a reasonably natural sound.

Either way, trying to get your songs stupidly loud is nothing to really strive for. It's a sickness of our time really and hasn't done music any good. It's something driven by music PR people that has been going on so long now that a lot of kids have never heard anything else, so they think it's normal. And, as many people have pointed out, it's not really more impactful, because you can't have loud without quiet. If it's just a wall of sound it stops being loud really, because either your ears just adjust to it or they turn down the volume. It's just obnoxious and fatiguing. Really punchy tracks, with loud/soft contrast are the ones that really sound loud.

 Low cut filters ( also known as  HPF) are your freind!! Sub sonic stuff can take allot of headroom and contribut nothing to the mix .............


Here's a sweet freebee

http://www.brainworx-music.de/en/plugins/bx_cleansweep_v2