Set individual channels lower than the subgroup fader?

[Sundown]

Hey all,

 

I love Bobby Owsinski's books (I love Craig's too, don't worry ). I have a couple of editions of the Mixing Engineer's Handbook and it's one of those books I re-read every couple of years.

 

Bobby threw me for a loop though while reading the fourth edition, which says that individual channel faders feeding a subgroup should be set lower than the subgroup fader.

 

For a moment I thought he might be referring to long-gone analog days, but he shows DAW screenshots.

 

I know that subgroups or busses serve multiple functions... You can apply effects to the group, you can adjust the stereo width collectively, etc.

 

But if I can't control the overall level of the group with a single fader (and I have to go back and pull down all the individual channel faders first), what's the point?

 

If it is a good recommendation for better, cleaner mixes, I'll certainly do it (and I can always gang the individual channel faders so that moving one retains the relative balance to others), but that just seems like a wasted step in software.

 

What does the forum think?

 

P.S. Not that it matters, but I run Cubase 6.5 right now.

 

Todd

[KuruPrionz]

It has been my experience miixing tracks in my DAW that I can have several tracks that show meter readings well below zero when played seperately but when summed the meter for the final output will be up in the red.

 

They are cumulative, they add up. At least that's what I have seen.

 

This is certainly a factor in a common complaint I've read regarding mixing/mastering - that the original recording was recorded way too hot. Many recommend individual tracks not go higher than -12 or even -18 and t seems this is the reason. 32 tracks of -12 is probably too loud.

 

It would depend on what your peak levels are when you record?

Just my 2 cents, cheers Kuru

[Anderton]

This is a good example of why a VCA channel is useful. Assuming there's no level-dependent effect in the bus (e.g., distortion or dynamics), assigning the individual channel faders to VCA control fader gives you a way to change the level gong to the bus fader, without having to change the channel faders or bus fader.

 

I wrote a blog post about Four VCA Channel Applications for Studio One, but the same principles apply to the VCA faders in Cubase, which I also have.

[Mike Rivers]

Hey all,

 

Bobby threw me for a loop though while reading the fourth edition, which says that individual channel faders feeding a subgroup should be set lower than the subgroup fader.

 

For a moment I thought he might be referring to long-gone analog days, but he shows DAW screenshots.

 

In the analog days (not long-gone for some of us), this was true. It was an issue of the amount of headroom available on the subgroup bus. The Mackie 1604 and first VLZ series of mixers had this issue - too many signals going on to the bus with their peak levels occurring at the same time summed to more volts than the bus output could handle.

 

The same was true in the long-gone digital days. The subgroup bus could handle 16 bits, as could the channels going into it, and when they tried to sum to greater than 16 bits, you got distortion. But with 32-bit, and now 64-bit wide buses, there's plenty of headroom. However, eventually you need to get the mix level down and when you pulled the subgroup down, you could get some truncation distortion. However that problem has been pretty much solved with today's algorithms and better converters.

 

[edit] Oh, another thing - plug-ins that come ahead of of the group master might not be happy with that many bits coming at them.

 

I remember being floored when I was reviewing the first PreSonus StudioLive mixer and wanted to see what happened when I overloaded the mix bus. I fed all 16 channels with a 0 dBFS sine wave of the same frequency, set all the faders to unity gain, and expected to see a huge amount of distortion at the analog output. Sure enough, the master bus meters were at the virtual pin, and the analog output was solidly clipped. However, pulling the master fader down so that the master meters dropped below full scale brought the digital level driving the D/A converter on the output down to where the analog stage was comfortable, and I got a clean sine wave out at the maximum analog output level of the mixer, somewhere around +18 dBu. I didn't have a really good distortion analyzer at the time to see how clean that sine wave really was, but it looked fine on the oscilloscope.

 

 

But, still, good digital gain structure is important and you shouldn't be sloppy just because you can.

[Sundown]

Thanks guys -

 

I don"t record my individual channels very hot. I record and render at 32bit floating (with 24bit AD/DA), but my signal peaks are usually around -12. I never let them exceed -6. On a plugin VU meter calibrated to -18 (Klanghelm VUMT), those peaks will tend to hover around 0 depending on the material. I just like the convenience of being able to pull up or down the entire drum kit with a fader, etc.