Combo Remapper - use score text and symbols to switch articulations
  1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Logic 8 To normalize?

Discussion in 'Logic 8' started by adamdrastic, Mar 22, 2009.

  1. adamdrastic

    adamdrastic New Member

    Hi guys,

    I'm working on a project by my rockabilly band. We've tracked drums and stand up bass simultaneously in separate rooms... but the studio recorded everything with ridiculously low gain...

    So, what's the best way to salvage these tracks? Should I normalize them, and if so can I do that within LE 7? Or is compression better?

    Thanks !!
     
  2.  
  3. Orren Merton

    Orren Merton Logic Samurai / Administrator Staff Member

    First of all, one of the joys of digital recording is that "ridiculously low gain" isn't actually that ridiculous anymore. In the days of analog tape, the gain had to be at a certain level or else the recording wouldn't be louder than the tape hiss. Since digital has no tape hiss, you can record at very low levels without problems. However, there is still room ambiance, the noise floor of your equipment, mic preamps, etc.

    So the first question is, are your individual tracks ruined by hiss/noise, or is the problem simply one of the volume being low? If the instruments are at the same level of the background noise, that's a problem, but if the instruments themselves came through without noise, even if the levels are at -30dB or so, that's fine.

    The best way to handle mixing the song is to simply mix the song at lower levels. If half your tracks are peaking at -30dB and the other half at -3dB, pull down the faders of the tracks at -3dB and mix the song at a softer level. You can always turn up the volume at the level of your studio monitors so you can hear it all loud and clear as your mixing.

    As for the final bounce, you have many options to bring it up to the level of other songs. I would not recommend normalizing in LE7, this is a "fixed point" bounce that operates on the file itself and has very limited headroom. However, if you upgrade to LE8, it comes with an incredibly efficient Normalize option in the bounce dialog. In LE8, the Normalize in the bounce dialog is a "floating point" process, that maps each bit of your audio onto the final 24-bit track at optimal level without ruining your dynamics. That alone may be a reason for you to move to LE8 or Logic Studio.

    You can also use effects such as a compressor or limiter on the output to raise the level of the final bounce. Logic Studio comes with a wonderful adaptive limiter (AdLimiter) but that's not part of Logic Express, I believe.

    Anyway, low level tracks in a digital recording are not necessarily a deal-breaker, that's the main point. :)

    Hope that helps,
    Orren
     
  4. Eli

    Eli Senior member

    Hi there,

    Every thing Orren says is spot on. To add to it from personal experience- I have recently began tracking live drums at my place after years of either using samples or tracking drums (at other studios) with and without outboard compression going in to the sound card. Maybe my tastes and aesthetics have changed / evolved / developed over the years, but I am complete believer at this point in tracking live drums with no compression, and at safe low levels.

    I record at 24 bits so the lower levels are not a problem in terms of loss of resolution. And I like doing this to allow myself (when I am playing drums) maximum peace of mind about not worrying about hitting the drums too hard and overloading the converters. The LAST thing I want on my mind when I am laying down a track is worrying about my playing volume. I set the levels low enough to be COMPLETLY at ease with hitting my drums absolutely as hard as I could possibly want to.

    But back to your question about mixing them afterwards: compression is your friend. But use a combination of serial and parallel compression. The more compressors you have in line, the less you need to push any individual one of them, and therefore the less overt and overstated the compressed aspect of the drum sound will be.

    For example, I routinely put a compressor on each of my overhead tracks, kik track,and snare track. I then output them all to a bus to be used as a drum submaster. I then put another compressor on this bus. If this isn't enough, put two compressors on this bus! I'm willing to bet that if you follow the above recipe and play with the faders, you'll get what you need.

    If you still need more, play wiht the make up gain on the compressors. Still more? Slap a limiter at the end of the whole thing.

    Seriously, low level drums are GREAT to work with! I have come to love it. They give you maximum flexibility with getting creative when you mix them. And maximum freedom from the performers of things.
     
  5. HKC

    HKC Senior member

    Orren Merton wrote: First of all, one of the joys of digital recording is that "ridiculously low gain" isn't actually that ridiculous anymore.


    Except you will lose bit depth which is not so good.
    Normalising won´t salvage that though (that deal was done at the recording stage) so mixing the tracks and making sure that the final mix is in 0 is the way to go. If recording levels are really, really low you could consider adding the gain plugin to the tracks.
    If the tracks are recorded in 24 bit (which they probably are if it was recorded on equipment that is under 10 years old) it´s probably no problem at all.
     
  6. bambony

    bambony Administrator Staff Member

    serial and parallel compression

    Eli's suggestions are excellent and helpful but I'd like to clarify that the compression techniques Eli suggests are all serial. This is where a number of compressors are used 'inline'. Eli does suggest using compressors on parallel tracks on a multi-mic drum recording but this isn't parallel compression in its purest form.

    'Real' parallel compression involves mixing together a compressed signal and the same signal uncompressed. Some plug-ins now allow you do this internally without the need to use multiple tracks. This technique can sometimes introduce phase anomalies but with modern DAWs and delay compensation this should be a thing of the past.

    Tony
     
  7. Peter Ostry

    Peter Ostry Administrator Staff Member

    Why not pushing up the low tracks instead, for example by Gain plugins? I personally prefer to have the mix at a level where my monitoring system is calibrated to.

    ---

    But I would like to know what adamdrastic meant by "incredible low level". What is this in the particular case, in peak and/or RMS values?
     
  8. Orren Merton

    Orren Merton Logic Samurai / Administrator Staff Member

    That works too, absolutely!

    In my example above, Logic's Gain plug-in doesn't have 30dB of gain AFAIR. Of course you could always use multiple gain plug-ins—in a floating point system, there's no real headroom loss if you use four gain plug-ins on each track, if you want to.

    For my way of working, I've got nice loud monitors with built-in 120W amplifiers (I normally have to attenuate them down -20dB to keep from blowing my ears out!) so I'd rather just mix everything at a lower level, and turn up my monitors.

    Orren
     
  9. Peter Ostry

    Peter Ostry Administrator Staff Member

    Aha!
    You did not calibrate your system and you have no marks on the controller knob.
    Got you!

    ;)
     
  10. Eli

    Eli Senior member

    Hi Tony,

    You're of course right; it's not true parallel compression. So, to elaborate on my previous suggestions, let me add this:

    Create two subgroup aux's for the drums by assigning the individual drum outputs to a unique bus and then calling up two separate stereo aux's with the same bus input on each. Then squash the hell out of one of them with a ton of compression. Leave the other one uncompressed (except for the individual track compression reaching both aux's). Then play back and slowly bring up the squashed aux fader to taste to blend it in with the unsquashed signal.

    Also, an interesting side note - Logic's compressor, as of v.8, now also has an internal mix slider in the interface. So, you can in effect establish these types of effects all from within a single instance of the plug in now. But thats' another thread :cool:
     
  11. zerobeat

    zerobeat Senior member

    Every bit gives you 6dB of dynamic range. Therefore, even if a 24bit recording is recorded at, say, -48dB (that's 8 bits worth of "lost" resolution: 48dB/6dB = 8bits) then you still have 16bits left over. So fretting over recordings done at -20 or -30dB are pointless.

    What's so great about 24dB isn't that it's better than 16dB as a DELIVERY medium. The great thing is that we can use that enormous headroom (theoretically 144dB instead of 16bit's theoretical 96dB which is still pretty large) to our advantage when recording and mixing.

    There are many reasons why a signal could be too low, and many of those might cause the sound to suffer a lot (like plugging a microphone into a line input - and getting a horrible signal/noise ratio). But being a couple dozen dB shy of 0dBfs (full scale) in the 24bit domain when all the other parts of the gain stage are OK doesn't mean a whole lot in the grand scheme of things.
     
  12. zerobeat

    zerobeat Senior member

    This is exactly how I mix. I start with all track faders at -10dB.

    This ensures that I don't hit the ceiling with fader creep.

    Its kind of how I used to mix in the analog days. More noise is infinitely better than even a microscopic sacrifice of clarity and non-distortion.

    And now in the digital days that "more noise" translates to "I still can't hear the noise when the music's playing loud and even if I could it's no louder than the room tone in this very quiet room I'm in".
     

Share This Page