This site is supported by the advertisements on it, please disable your AdBlocker so we can continue to provide you with the quality content you expect.

Welcome to Our Community

Wanting to join the rest of our members? Feel free to sign up today.

Gain staging / trim plug / printing to wav across board

Discussion in 'F.O.H.' started by LeSedna, Dec 30, 2014.

  1. Studdy

    Studdy Member

    Joined:
    Jan 24, 2012
    Messages:
    947
    Likes Received:
    0
    Trophy Points:
    16
    Doesn't normalizing to peak raise the noise floor and kill your headroom? I learned early on that normalizing to peak was bad practice. I guess it was a myth. Can you tell me the advantage of normalizng to peak first? Dont you just pull everything down again after the fact? Thanks man.
     
  2. Jordon

    Jordon Member

    Joined:
    Sep 14, 2008
    Messages:
    1,008
    Likes Received:
    3
    Trophy Points:
    38
    Location:
    Chicago
    I don't normalize to 0dbfs. I looks like lately I've been setting the AudioSuite Normalize plugin to -8db peak, which usually gets me in the ballpark. Like I said, it completely nulls compared to clip gain or a printed trim plugin. I was always told normalizing was bad practice, as well, but if there's no change to the audio other than a volume increase, then I'm not going to worry about it. I'm positive that this isn't the "right" way to do it, but it streamlines my workflow and allows me to focus on actually mixing, instead of fiddling around with shit.
     
  3. LeSedna

    LeSedna Mat or Mateo

    Joined:
    Jan 20, 2008
    Messages:
    5,404
    Likes Received:
    1
    Trophy Points:
    38
    Location:
    Montpellier, France
    That was probably what i was thinking about earlier. By normalizing up you're raising your previous noise, by normalizing down you're deducing it but you are bringing yourself closer to the noise floor of your (usually) 24bit. I mean, when you're printing. Before you are, supposedly the internal engine keeps everything in programmation variables so nothing should happen.

    What I am guessing is the overall effect is such an non-event people wouldn't care about it. Or amiwrong ?
     
  4. Studdy

    Studdy Member

    Joined:
    Jan 24, 2012
    Messages:
    947
    Likes Received:
    0
    Trophy Points:
    16
    Obviously proper tracking levels eliminates the need for any of this. I think I will stick to a clean eq first in chain to get levels proper plus deal with any frequency issues. Cheers.
     
  5. Kohugaly

    Kohugaly Member

    Joined:
    Oct 15, 2011
    Messages:
    258
    Likes Received:
    0
    Trophy Points:
    16
    +1 on that!

    As for normalizing versus manual trimming - there is no difference between the two, except that normalizing is automatic (they both just boost/cut the volume). It finds the highest peak and then trims the entire file so that the peak is at xy dB ceiling of your choosing. The downside is, that the highest peak in the file tends to be quite random. Two identically sounding transients may easily peak 5dB apart. Normalization may give you this massive offset from optimal gain level, just for the sake of a single few sample long transient. Nobody who's sane would re-track perfectly sounding performance just because he "doesn't like where this single transient peaks in the wave form".

    As for printing the trim... 24bit-depth is a massive overshoot in precision. Last few (2-3) bits are just plain random noise. By trimming the file -15dB you loose no "audio" part of the information in wave file. With 32bit float format (which is used by most daws internally) there is no lose of data at all. Unless you are printing into 16bit file, printing definitely shouldn't affect the sound-quality at all.
    The only real question is, does it give you advantages that are worth the time and effort? (obviously opinions on this differ, so you'll have to give the answer to yourself)
     
  6. Jordon

    Jordon Member

    Joined:
    Sep 14, 2008
    Messages:
    1,008
    Likes Received:
    3
    Trophy Points:
    38
    Location:
    Chicago
    Great point about the normalizing with very dynamic tracks. I generally only use this approach on guitars, bass, synths, and vocals. sometimes drum rooms and overheads, but kick, snare, and toms, it's just too unpredictable.
     
  7. LeSedna

    LeSedna Mat or Mateo

    Joined:
    Jan 20, 2008
    Messages:
    5,404
    Likes Received:
    1
    Trophy Points:
    38
    Location:
    Montpellier, France
    My own understanding is that every single time you adjust the volume you are loosing information, by quantization. I just think this is so irrelevant for a normal day to day mixing we shouldn't even care, in fact. In the analogue domain, everything coming from the tape, is then processed by your board which by definition is not digital thereofe it doesn't even matter.

    So in fact I think it is only a matter of session set up and your own preferences. I personnally like now to have all my tracks to -18dbfs or OVU from an analogue-like VU plugin with a 300ms averaging time, it takes s couple minutes and from that point I know all tracks are analogue emulation-ready
     
  8. rapucore

    rapucore i really hate spiders

    Joined:
    Feb 25, 2013
    Messages:
    786
    Likes Received:
    22
    Trophy Points:
    18
    Location:
    FIN
    +1

    Mixing feels so much better when you can just start doing your thing and rely on your ears and instinct. Proper gain staging is a must IMO.
    Like LeSedna said, it all boils down to your own preferences. I vouch for printing the gain staged tracks to new .wavs and work from there. :)
     
  9. TheWinterSnow

    TheWinterSnow Den Mørke Natt

    Joined:
    Oct 22, 2008
    Messages:
    3,090
    Likes Received:
    2
    Trophy Points:
    38
    Location:
    Sacramento, CA
    You don't loose information when you adjust the volume in the DAW, it is rendering the source audio every time you playback. Adjust the volume back up and all it does is doesn't attenuate as much.

    Printing is another story though. If you dither on the way out, the noise floor stays the same, if your tracks were too hot, you get your signal closer to the noise floor. At 24-bit however, the noise floor is almost non-existent, that is because 24-bit has a dynamic range about 3dB short of the human ear.

    My opinion is that it is quicker in Reaper to use the pre-fx volume automation as a trimmer. That way it is completely lossless to the audio data and you can always alter or remove it if you wish without the need to re-render.
     
  10. LeSedna

    LeSedna Mat or Mateo

    Joined:
    Jan 20, 2008
    Messages:
    5,404
    Likes Received:
    1
    Trophy Points:
    38
    Location:
    Montpellier, France
    ^ I was referring printing, not playback, as playback is driven by the internal engine which is usually oversampled anyway, but indeed my sentence felt like the other way.

    I have actually discovered recently the prefx volume setting in reaper, although I think I will just get in the habit of printing a track to a more acceptable level whenever it seems off to me.
     

Share This Page