High or low volume into interface for DAW?

  • You want to use the monitor DAW outputs being the overall master volume. The theory is that lowering volume digitally removes data and therefore you should only adjust output volume at the analog stage, ie the signal sent to the speakers, not before. In fact my DAW system sounds MUCH better when I use this type of setup vs using the Kemper as the Master vol or even the internal mixer.


    [edit] for recording SPDIF at max is the correct way.. what is this 'leave headroom' thing? you lower individual tracks with the mixer

    Edited once, last by mikeb ().

  • What exactly is this 'headroom' for later on?


    If you record your signal at 0dB you cannot make your signal louder later without clipping, for example using a compressor or other loudness shaping tools.
    So leaving a bit of headroom is always recommended.
    When recording in 24 bit or 32 bit floating point you can track hotter though than -10 dB.
    When I bounce Toontrack Superior Drummer I aim for -6 dB. Later on I can make use of my UAD compressors then.

  • If you record your signal at 0dB you cannot make your signal louder later without clipping, for example using a compressor or other loudness shaping tools.
    So leaving a bit of headroom is always recommended.
    When recording in 24 bit or 32 bit floating point you can track hotter though than -10 dB.
    When I bounce Toontrack Superior Drummer I aim for -6 dB. Later on I can make use of my UAD compressors then.

    Actually the way I understood it is that you can always remove volume(information) later by simply lowering the volume on the particular track (digitally). Then you can add effects as you need while adjusting track volume. The internal mixing is 64bit on most DAWs which allows as much 'headroom' as you like as long as you mix down to the proper levels. :?: :?:


    Quote

    If you clip your master bus, you've clipped the playback, and the audio if you render/save. Clipping will not occur in the internal engine, but once you leave the app (to a soundcard, etc) . . . it will clip.

    [edit] reading more, seems that DAWs have infinite headroom 64bit float mixing. So basically I see no reason to leave headroom while tracking.

    Edited once, last by mikeb ().

  • So basically I see no reason to leave headroom while tracking.


    Whats your source for this? In fairness I don't understand all this float stuff your talking about but i've never read or heard from any good engineers which have seconded not leaving headroom when talking about proper gain staging.

  • 64bit floating point simply means a value to a certain number of decimal places e.g 64.2358987 as opposed to integer values which are whole numbers e.g1024.

  • Leaving headroom is a good idea especially if your recording 16 or 24 bit so recording at around -6db to -3db or lower is fine. If you recorded at close to 0db you would need to attenuate the input volume of additional post fx later within the plugins you use. Simply turning down the fader will not do this as your post fx will likley be pre fader by default. That said 32 bit floating point and higher will not clip in this situation but its important to think about how little headroom your leaving for your mix and faders by having such hot signals. There is no benefit recording at a very high level it wont sound better especially if your recording via SPDIF it's not like we are fighting background tape hiss any more

  • Guys, you are recording way too loud. I've noticed this in a lot of Kemper soundcloud and youtube examples that those are actually clipped.


    Do not aim for that 0dBFS, it is the best way to achieve what you're not looking for (terrible sound quality)
    You should record ANY signal (Kemper or else) at approx -18dBFS as your Average Level (not peaks)
    Also record in 24bit (plenty of dynamics/no need to "sacrifice" headroom)


    Your interface/soundcard preamps and converters are designed to work at those values, and recording a hotter signal might start to distort the sound in the pres and then in the converters.


    And don't forget to check your profile, it may actually be clipping in the Kemper itself (lots of downloaded user profiles are too ******* loud even "pro" profiles like TAF)


    Cheers

    ;)

    Edited once, last by fak0u ().


  • Yes to this. I said -10db for the peaks but average of -18 is as close to a recommended standard as there is gonna be for gain staging


  • Whats your source for this? In fairness I don't understand all this float stuff your talking about but i've never read or heard from any good engineers which have seconded not leaving headroom when talking about proper gain staging.


    Those engineers were probably coming from analog recording and then switched to Protools for digital. In the analog world gainstaging was very important and with the early Protools nearly as much. Protools was 24-bit fixed point for the better part of it's existance and that format DOES clip if you are not watching your levels. And it can do that during internal operations inside the audio-engine.


    Native DAWs like Cubase were mostly 32-bit floating point from the very beginning or adopted that format pretty quickly. With 32-bit float it's impossible to clip during internal editing and mixing. You can only clip at two points: The analog input and when printing down to 16-bit fixed point for your master. The last plugin in the chain (usually a multiband comp) can produce intersample clipping that can hurt the master.


    But other than these two points gain staging is not important anymore - that is if you are not using an older Protools version ;)