The "wiser" way of outputscaling, building a mixer
Hallo all! I'm building an audio (ableton style) mixer in max, and want a quick advice. Let's say I have 10 channels, and therefore want my LEFT/RIGHT output scaled by *0.1, so it won't overload my dac~.. I get that, but - if I mute two channels, then I will only have "9 channels". Should i then multiply my signal with 0.125.. or stay with the *0.1 ? (I wonder what "real" DAW makers do - do they scale their audio every time a channels is muted, or just whenever a new track is created, or is there something else i don't get/see?)
you would not scale at all. just take the channels as they are and leave it to
the user that the sum does not exceed 1.0 .
thats also how analog mixers do work - and it works in practice.
so depending on how many channels you use on a analoge mixer, you have to turn every channel down accordingly?
Roman is 100% correct....But to elaborate a little:
On an analogue desk, if you have one strip set to unity gain, and set to 0dB, and the master bus is set to 0dB, you'll have no amplification or attenuation at all. The signal level you put in is the same as what you get out. Therefore if you have two channels with identical inputs panned central with the same settings the master bus will be kicking out +6dB i.e. twice as loud. When different signals are used on each input the output will be less but still above 0dB because of how the two signals sum together in terms of transients and frequency content.
Getting your sources to run as loud as possible without peaking at the output has been the holy grail of pop music and dancefloor producers since forever and is all about compression, eq, space within the mix, and generally not cluttering the soundscape. As a general rule, if it sounds cluttered or distinctly sounds considerably louder or quieter when you mute/unmute channels then you probably need to address your sources or signal chain :)
So, yes.... Roman is right - leave the mixer as a straight 0-1. The mixing itself is a thread for another forum....probably. Apologies if any of my rambling is preaching to the converted at all!
"to turn down" is not exactly what you have to do.
the theorie behind it is even simpler: it is not the normal case that you
have signals on your mixer channels which are hitting 0 dB, in a realistic
situation you have a lot of signals which are not that loud anyway.
you use the faders on a mixer mostly to mix the gains against each other.
you start with all fader set to 0db and lower things which are too loud.
so when you mix pop music with 48 channels most faders are between -6 and 0,
and most signal you run in the mixer are between -20 and -3 db.
now if you sum this stuff, your LED meters or VU meters might end up
somewhere at +3, and you audio will have single peaks of maybe +12.
you would now lower the master volume to -4db and use a peak limiter on it
before you send the output to a PA or record it digital.
there is only one little difference between analog and digital mixers and that is
that in the digital world you should (even when you work with 32 bit format files)
always try to stay at -0.0 in groups and in the master.
btw, if you would mix hundreds of digital channels which are all at +12db you
will end up with a lesser quality sum compared to not so loud channels in most
audio programs, so itis good advise to avoid that anyway.
-110
cool thanks a lot guys!