Hitting the red: I was wrong about this

I have a brand new, for me, Kontakt 5 patch, Upright Piano Modern Preacher, that hits the "red" with moderate velocity mapping in the MIDI source. I came upon the circumstance just yesterday.
 
Yuck.
 
 
My suggestion: mix with the powered monitors switched to "on". 
 
:-)

In a dense mix, one of the single most challenging things to get right can be gain staging.Especially when you have the occasional peak transient on an otherwise great track.Let it go or lower the whole track ? Edit midi velocity? Apply some clip compression in only a few places? Volume automation? Lots of approaches, but it isn't always as easy as it seems.
 
Pianos can be some of the toughest IMO. But they aren't alone. 24 bit seems to be a bit less forgiving, but that might be a misconception on my part.

Starise
24 bit seems to be a bit less forgiving, but that might be a misconception on my part.

You're right, because when we say "24-bit" we're implying integer values. For integer (as opposed to floating-point) numbers, there simply isn't anything beyond 0 dB. That's why it's so important to not drive your audio interface into the red, because it knows nothing of decimal points and therefore has an absolute upper limit that literally can't be crossed.
 
Floating-point, OTOH, automatically adjusts to whatever size number you want to stick in there. It just shifts the decimal point as needed. It can therefore accommodate very large (or very small) numbers without running out of bits. You do lose some accuracy as some of the least-significant decimal places are chopped off, but this loss of accuracy is irrelevant in audio, so in theory 32-bit floats can go over zero with impunity.
 
It's a wonderful innovation, using floating-point data in DAWs. However, it breeds sloppiness because it allows us to get away with practices that wouldn't have been acceptable in the pre-digital age. And presuming everything's OK can occasionally bite ya in the arse.
 

My suggestion: mix with the powered monitors switched to "on".

Good advice!
However, it was because I had the speakers off (listening on cans) that I initially noticed the crackling that headed me to this discovery.

bitflipper
in theory 32-bit floats can go over zero with impunity.

You can exceed zero and never clip - this much is always true (as long as you appropriately attenuate before a conversion back to fixed point format happens, of course).
 
On the other hand, I think I'd feel cheated if a plug-in that purported to emulate analog gear didn't behave differently when processing signals that would be considered too hot in the analog world.  If that plug-in doesn't do something different when being fed too-hot data like the real gear, then it's not behaving as advertised.

I guess this is why the ProChannel modules have clip lights? 

gswitz
Bit, some plugs are designed to start distorting before you reach zero. They model overloading before the signal is clipping.

Yes, that's intentional harmonic distortion. Lots of plugins do that, on purpose, and it usually sounds OK. They're emulating the way tubes and transistors behave as they're driven outside their linear range. But that's not the kind of distortion I'm talking about.
 
Digital clipping and foldback are just ugly. They generate tones that are not harmonically related to the signal. They don't emulate anything in the real world. They just sound bad. Fingernails on a chalkboard bad. Technically, neither should happen in a floating-point system. I can only speculate about what's going wrong within those plugins.
 
Values are passed into and out of the VST interface as floating-point data. However, the spec calls for audio to be represented using values between -1 and 1. That means 1.000 is valid, but 1.001 may not be.
 
Now, there is no reason 1.001 can't be stored in a floating-point variable and manipulated exactly the same way as 1.000. But if the developer assumed that no value would ever exceed 1.0, it's possible that his math could be thrown off, or some logic circumvented when values unexpectedly went over 1 or under -1. You'd have to examine their code to know what's happening, but I think it's reasonable to say that it's either a mistake or an oversight.
 
 

If you have to turn things down too far on track faders before you mix you are still recording a little hot. All the more reason for K system and VU metering.  The use of the K system for me extends right back to setting input signal levels arrivinga at the DAW.  It is easy to get a K system ref level recorded on a track for any input source.  Apply it at the time of turning soft synths into audio ensures the audio is also at the ref level.
 
If you track at your reference eg K-14, then later when you are mixing, the faders are much closer to unity to still achieve a K system ref levels on any buss those tracks are sent to.  I prefer the amount of gain  adjustment around unity compared to anywhere else on a fader.  Tracks only have to be shifted downward slightly from unity in order to create a perfect balance and the right level on that buss. A buss master can also easily fine tune your K ref level on any buss.  That is what it is for.
 
With plug-ins if you make sure your levels are K-14 or 0 dB VU just prior to going in, you know you have got a nice level going into any plugin without clipping.  Then make a slight adjustment with the VU level on the way out too.  You can get those two levels (IN and OUT) very well matched and you can be sure nothing bad is happening inside the plugin itself. (things getting near 0 dB FS for example)  If you are still concerned, drop down to K-20 and you will never have an issue with internal distortion. (with plug-in chains you can always insert VU meters between all the plug-ins in a chain to ensure they are ALL working at a perfect level too. Not all plug-ins match the OUT level exactly to the INPUT level, you need to keep an eye on that. Output levels vary from many plug-ins and need to tuned back to your ref level sometimes.) Matching plug-in input and output levels with a VU also means when you bypass or test for what a plug-in may be doing in your mix, the level won't change and that sound sits perfectly in your mix, in or out.
 
With soft synths, they all vary by a lot.  Some from very quiet to very loud.  All you have to do is take a VU meter reading straight out of any soft synth and just adjust that for your K ref or 0 dB VU.  You will never hear distortion again from any virtual instrument when you get into this habit.  Some of mine are quiet and I have to add gain to them to achieve the desired K ref level.  Others need to be turned down quite noticeably.  As much as 9 or 10 dB.  Most offer output variable level adjustments.
 
By doing this all your soft synths will be at the right level prior to mixing too.  It is good to render them at the correct level, one less step to do in your pre mix work prep.
 
Reading a VU meter also takes a little practice too.  Signals vary over a wide dynamic range sometimes and it is a matter to understand what the VU meter is showing you.  We are looking at the loudest bits and over longer times as well.  After a while it is easy to see where that average point is with the meter ballistic and line that up as best you can around 0 dB VU.
 

Epiphanies are defining moments. You never know to much to have one every now and then. 
 
I've noticed that Melodyne can distort and clip in a really nasty way and you don't have to be in the red to do it. I forget the exact circumstances but the wave before was nowhere close to clipping, yet in the processing, Melodyne slammed it to the flat top..... I went back and used undo to get back to the pre-processed wave, then reduced the gain and tried again..... waaa laaa.
 
I always try to look to be sure I'm at the proper levels and check the processed output before moving on to the next thing.

MY TYPICAL TRACKING LEVELS:
 
Peak = -6 to -8
RMS = -22
 
that's a big crest factor.
 
 
i don't feel the need to EVER get into even the yellow with my track levels,
and most of you have heard the results.
 
happiness,
is clean tracks.
 

tremojem
 
i'm talking strictly about TRACKING.
 
during the tracking phase, i like to try to have my individual tracks never exceed -6 on PEAK.
 
typically, it's actually -12.
that is the sweet spot.
 
when i combine all of my tracks together to do a mix,
i try to keep the MASTER bus with peaks of no more than -8, and RMS of no more than -20.
 
this has to do with matching analog scale of 0dbu at digital scale -20 dBFS
 
so when i pull my final mixes into WAVELAB to do my mastering,
i have a file that usually has a Peak level of about -6, and a RMS of about -22, if i did my job right.
 
 
then i master it, and typically bring my levels up to peak of -0.2 and RMS of about -10-12,
aiming for a dynamic range of about 10.
 
i can push my masters louder, to about DR8, and NOT get into trouble,
but i like the sound of it less squished.
 
 
if i could put out my mixes WITHOUT mastering, i think those sound the best of all.
 

TremoJem,

I do the same thing as you. I've found that when mastering tools are applied, it inevitably changes the relationship between the various parts of the mix, so it's necessary to go back and do some tweaking of the levels of individual channels or busses. In other words, there may be an optimum mix, but you don't know what it is until you've done the mastering.

I'm sure that in someone's eyes this will be "wrong", but it's what works for me.

trem,
sounds like you are asking 'mastering' specific questions....
 
that is an entirely different thing.
 
i have a 3rd party software i use (Wavelab) dedicated specifically to mastering.
i use a 3rd party set of Plugins (WAVES) to use for mastering.
 
 
when i am going to master a song, i bounce it to a stereo file, 24 bit, no dither.
it is a 'what you hear is what you get' bounce directly thru the master bus, with all plugins and automation included.
 
then i "EXPORT" that 24 bit wav file out of sonar completely and work in Wavelab.
 
if you are going to try to master inside of Sonar (not my choice method) then i would suggest bouncing your song down to a final mixed stereo file, export that, and bring up a brand new clean 'song' file and import it into that, putting your plugins on the master bus and processing directly thru that, avoiding all the unnecessary processing in your working song. (even the channel strip eq's hog memory)
 

I find when you are sending a group of tracks to a buss it is not hard to keep an eye on buss levels.  When you are VU metering there it is easy to get an overall reading on that buss.  You can almost mix with the VU meter itself.  On a drum buss as you bring in the big hitters such as kick and snare then you can get those VU's moving up to -3 dB VU or so.  As you add extra things into that drum buss the final levels will start reaching 0 dB VU.
 
As you solo into a buss and listen to what is just on there, the VU will tell you if your buss is starting to run hot or not.  I might group all those tracks that are feeding that buss and pull them down a dB or two during the process of getting buss mixes correct.  You can fine tune the buss master to keep things on track.  Any slight compression over a buss will create the effect of raising its rms level up slightly and reducing slightly the dynmic range on that buss too.  (usually to good effect)  The ballisitics of the meter changes immediately you insert a compressor over a buss.  You can use it to fine tune your compressor settings. Light compression over some Virtual Instruments is almost essential.
 
Buses then feed the master buss and it is not hard to juggle those buses to get the mix perfect and the main stereo buss ends up at the K ref level too for the entire mix.  I have the real VU's over the main stereo mix at all times right in front of me and that keeps you in check right back at the point of bringing tracks up one by one into their buses.  You have always got your final target level in front of you as you mix.  The virtual VU's are good for the spot positions within your mix.
 
Do this and the whole session drops down to the point of no clip lights on anywhere at any point in your signal flow.  With plug-ins it is easy to VU meter on the way IN and OUT to make sure everything is running at the right level.  It is possible to send too high levels to a buss and then you are struggling to pull things down.  Mixing into a VU is more like you gradually build up the VU level as you bring things in and then to the point where everything is in and it is just hitting its final target level of 0 dB VU. And not swinging wildly over either.

Jeff E., what VU meters are you referring too?
 
This option does not exist in X3e...to my knowledge, but then again I have only scratched the surface of this application.

No, the VU meters are not standard.  They are extra plug-ins that you can get.  But they don't cost much though.  A few companies make them.  The Klanghelm meter is pretty good:
 
http://klanghelm.com/VUMT.php
 
The rms or VU meter can take different forms.  eg bar graph displays etc but I like the standard VU needle.  It offers a lot of information in a short space of time.
 
I have encountered soft synths that produce too much volume and the sound can actually leave those VST's almost clipping.  Metering their output level will enable you to fine tune their output to match your system reference level.  Make the VU meter the first effect in the effects bin.
 
Some soft synths have limiter features built in and I usually prefer to switch any of that off and let them send out a full uncompressed or unlimited signal.  By turning the output down to match your system level the VST is now not creating any signal that is close to clipping.  The signal level is right at the source.  Synth sounds vary in level a lot because of the nature of the envelopes etc.. The time to turn the output level right up on a soft synth is when the sound is a very short envelope that does not reach much level and dies quickly. Then you may even need the VST output level at full just to get a K system ref level. A really loud bass sound or brass patch on the other hand might require the VST output level way down to get that same K system ref level coming out. Hardware synths behave the same way wth patches. They vary so much in level.
 
VST Presets can sometimes take over the setting of the output level knob as well.  Changing it will often only work with your current selected preset.  Selecting a new preset will often return the output level to the setting locked in with that preset.  (It is not a bad idea to resave presets used in any production with the right output level setting for that preset and your system)
 
Messy sounding instruments can also mean messy sounding arrangements.  It is good to look at which synths are playing what all the time and seeing how many are playing at once.  Ideas can interweave rather than try to exist all at once. But that is all another story.