lowering channels faders or master fader when mixing?

[djanogil]

When mixing in DAW, to avoid clipping and to keep the best audio quality and low distortion, is it recommendable to lower the master bus level, or keeping it at unity and lowering all the channels of the mix?

[Bishbashbosh]

There will probably be a lot of debate on this..... BUT, in my opinion, MY mixes sound better when I give loads of headroom at all stages- into and out of plugs, on each channel, and at the master fader.
On a 24bit system there's no real need to push towards zero at all. Let your ME push the level up for loudness, but for mixing keep levels way down. I usually aim for MAXIMUM peaks at -8 ish.

There was an interesting thread on this recently: http://gearslutz.com/board/showthread.php3?t=54452

Very different approach than analogue, but then it's a totally different beast.

[funkynuts]

I completely agree.

There is absolutely nothing to be gained from going into the red at any stage of a completely digital enviroment such as your daw mixer.
Allow adequate headroom on each track incuding the master fader and increase your monitoring level to compensate.

Going into the red at any point of an all digital mixing enviroment will not add pleasent artefacts like the analogue world.

Plus you will increase your chances of overs etc.

You are also not having to drive high signals through a digital system in order to contend with the self noise apparent in analogue circuitry etc.

Leave plenty of headroom and aim for an rms output of -12 db with peak no higher than -7or-8 db.

[Matthew Murray]

I can definitely see the benefit of keeping lots of headroom when mixing -- but how about when tracking? It's always been my logic when tracking digitally to get the hottest signal possible without clipping at the peaks ... is that not as important as I thought?

[djanogil]

Thanks for your response,
I'll put it differently;

In a situation where I have a good balance in my mix and some headroom on my individual channels, but with a master fader clipping at unity,
should I leave the master at unity or put down the levels of all the channels
until I get good headroom on the master fader?

[gsharp]

Originally Posted by djanogil

Thanks for your response,
I'll put it differently;

In a situation where I have a good balance in my mix and some headroom on my individual channels, but with a master fader clipping at unity,
should I leave the master at unity or put down the levels of all the channels
until I get good headroom on the master fader?

Do whichever you prefer. Pulling down the master just rescales your output so its no longer clipping. That's what I do when I get all dialed in and happen to notice a few peaks going over. Easier than trimming everything down.

The posts above make a great point about leaving plenty of headroom. Lately I've started working with my volume knob up higher to start so I can hear what's going on when I'm first getting sounds while leaving LOTS of room on the master. Before I was constantly trimming my mix down as it progressed.

[djanogil]

Originally Posted by gsharp

Lately I've started working with my volume knob up higher to start so I can hear what's going on when I'm first getting sounds while leaving LOTS of room on the master. Before I was constantly trimming my mix down as it progressed.

Exactly, the same happens to me most times, so I'll try the same route.
BTW questlove rules

[gsharp]

Originally Posted by djanogil

Exactly, the same happens to me most times, so I'll try the same route.
BTW questlove rules

yes, he pretty much does.

[seriousfun]

Originally Posted by Matthew Murray

I can definitely see the benefit of keeping lots of headroom when mixing -- but how about when tracking? It's always been my logic when tracking digitally to get the hottest signal possible without clipping at the peaks ... is that not as important as I thought?

This is true when recording to analog tape, but absolutely not true when recording digitally.

If you are tracking 16-bit, you have to be careful, but if you are recording at 24-bit or better, give yourself lots of room. You will retain full dynamic range if you are peaking at 12 or even 24 dB below 0 dBFS, and have more room when adding channels together at the mix bus or master fader. As always, do what your ears tell you to do.

[funkynuts]

Originally Posted by Matthew Murray

I can definitely see the benefit of keeping lots of headroom when mixing -- but how about when tracking? It's always been my logic when tracking digitally to get the hottest signal possible without clipping at the peaks ... is that not as important as I thought?

This is still valid.

As long as you are still allowing sufficient headroom before overs.

Your a/d convertors may perform better when fed higher level signals and reproduce the signal more accurately.

This is especially important with low to medium end converters.

It also allows for a better signal to ambient noise ratio in a less than pro recording enviroment.

Just ensure that there is always plenty of headroom.

[John The Cut]

This is the way I see it, and correct me if I'm wrong - but resolution on the channel faders in a DAW decreases as you pull the fader down (bit depth resolution and also resolution of the fader (its harder to mix when the fader is down)).

I've always understood it to mix with the channel faders near to 0 as possible to maintain resolution and pull the master fader down to avoid clipping.

[Bishbashbosh]

Originally Posted by mogWai

This is the way I see it, and correct me if I'm wrong - but resolution on the channel faders in a DAW decreases as you pull the fader down (bit depth resolution and also resolution of the fader (its harder to mix when the fader is down)).

I've always understood it to mix with the channel faders near to 0 as possible to maintain resolution and pull the master fader down to avoid clipping.

http://akwww.digidesign.com/support/...48BitMixer.pdf

This is worth a read.... Clearly not all the information pertains to all DAW's, but it'll do a little to dispel some of the myths surrounding digital audio. Particularly the one about losing resolution by pulling down faders.

There is much incorrect belief surrounding digital audio, which is probably left over from older 16bit systems. The dynamic range of 24bit, and the nature of digital summing mean that there is pretty much a consensus that keeping levels DOWN at all points in the chain is a good thing.

The only times I think that keeping levels UP is an improvement, is if you are using crappy converters, or recording 16bit, where in both cases using the full bit depth is probably going to make improvements.

There is also the issue of hitting DA's with a reasonable level if using a summing box, but that's probably for another thread.

[XSergeantD]

Here's my post from last September (had to go find it as I already forgot this info), not exaclty what you're looking for but gives you some info you may want:
A few things about the master on Profools, Master Faders do not use up additional DSP, they're ALWAYS there, just not always shown. The master fader red light is not showing clipping on input, it's clipping on output only, so don't pull individual faders down, just the master fdr.
The mixer plugs that come w/ HD got 48dB of headroom, meaning that the 'input' side of the bus, signals can never clip - even if chnl faders are set at full+6dB o gain. The 'output' side of the summing mixer can clip. You can use a master fader to drop the output level of any mix summing piont (bus or physical out). This is where the red light on the master fader will tell you 'pull me down' without loss of quality.

[djui5]

It depends on if you have a compressor on the master fader or not. If you do, bringing down the individual faders can affect the input to the compressor and change your sound. Bringing down the master fader won't affect this.

Also, if you need to you can place a trim plug-in on the first insert of the master fader and adjust as necessary.

It can be a pain to bring down all your tracks when deep into a mix. It can really change the sound of things.

[gsharp]

Originally Posted by djui5

It depends on if you have a compressor on the master fader or not. If you do, bringing down the individual faders can affect the input to the compressor and change your sound. Bringing down the master fader won't affect this.

Actually, in PT at least, the plugs on the master are post fader. Pulling down the master pulls down the level into them. That's the one valid complaint (I disproved the voodoo for myslef) I see against BTD. If you do a fade on the master your compression will change on the way down.

[Improv]

Originally Posted by gsharp

Actually, in PT at least, the plugs on the master are post fader. Pulling down the master pulls down the level into them. That's the one valid complaint (I disproved the voodoo for myslef) I see against BTD. If you do a fade on the master your compression will change on the way down.

That's why a lot of PT guys buss to an aux first, where they insert their 2-bus processing. Then this is fed to a seperate bus or main output, with a master fader, feeding whatever you mix down to. Fades on the master fader at this point are post-fx and compression (the way it often should be).

Post-fader processing on the master fader is designed with scaling in mind, not fades or rides. The idea is to scale the signal feeding the first plugin to avoid overs, and it's nice to have for this purpose. a pre/post toggle for plugins on a master fader in PT would be a wonderful idea, though.

I often do most production work with the master fader at -6 (or even lower for heavier stuff) and an L2 on the master fader at -6 (or whatever, to compensate). I also track at "low levels" (i.e. at 0VU, or about -18dB on a stock PT interface... no need to track hotter), and tend to use dynamics processing on the channels so that a good mix looks like straight faders at 0 (this gives me a reference point for rides... 12 dB of gain on a channel is usually plenty of headroom for rides). Using this setup, I always have plenty of headroom, and usually end up pushing up the master fader (and raise the L2 accordingly) as the song gets "full" and I know how much headroom I'll really need.

[John The Cut]

Originally Posted by Bishbashbosh

http://akwww.digidesign.com/support/...48BitMixer.pdf

This is worth a read.... Clearly not all the information pertains to all DAW's, but it'll do a little to dispel some of the myths surrounding digital audio. Particularly the one about losing resolution by pulling down faders.

There is much incorrect belief surrounding digital audio, which is probably left over from older 16bit systems. The dynamic range of 24bit, and the nature of digital summing mean that there is pretty much a consensus that keeping levels DOWN at all points in the chain is a good thing.

The only times I think that keeping levels UP is an improvement, is if you are using crappy converters, or recording 16bit, where in both cases using the full bit depth is probably going to make improvements.

There is also the issue of hitting DA's with a reasonable level if using a summing box, but that's probably for another thread.

erm... thanks for the info, but now I'm confused. On page 6 of that article it says that lowering channel faders does lose resolution and you should use the master fader to pull down a mix, because you wont lose resolution there.

This is what I understood to be right in the first place, and its what I do. Are you saying channel faders dont lose resolution?

[gsharp]

Originally Posted by mogWai

erm... thanks for the info, but now I'm confused. On page 6 of that article it says that lowering channel faders does lose resolution and you should use the master fader to pull down a mix, because you wont lose resolution there.

This is what I understood to be right in the first place, and its what I do. Are you saying channel faders dont lose resolution?

you lose fader resolution (just like on an analog console) but you don't lose any data resolution. you lose the .1 dB increments once you get down to -40 or so. it goes to .3 or .4 dB But all your audio is still there.

[Lawrence]

Originally Posted by gsharp

Actually, in PT at least, the plugs on the master are post fader. Pulling down the master pulls down the level into them. That's the one valid complaint (I disproved the voodoo for myslef) I see against BTD. If you do a fade on the master your compression will change on the way down.

Really? I have PTLE but (obviously) don't use it a lot. I didn't know that.

That's an oversight. SX/Nuendo has permanent pre and post fader insert slots on it's channels including the master. You put the plug(s) wherever you want it, pre or post. I think it's 6 pre and 2 post per channel.

Lawrence

[Crushed]

Yes. In Nuendo, the first 6 inserts on each channel, including the master are pre-fader, and the last two are post-fader. The rationale is 2buss compression, flowed by dithering on output. Makes sense to me. Can Pro Tools not do this? Seems wrong to me, but what do I know?

Allen

[jdunn]

I think it's better to lower the individual channels and try to leave your master fader at zero. This way you're not overloading the mix buss of the DAW. Here's a quote from the "Summing Philosophy" thread which seems to support this method.

Originally posted by Paul Frindle:

You can avoid the whole darned problem by simply reducing levels at the begining of every channel (each track into the mixer) and aiming for -6 to -10dB on the meters rather than flat out - and putting back the level again at the final output of the mixer. You can afford to do this as a 24bit system has around 140dB of dynamic range - loads more than you will ever need. Give it a try you might be really surprised - and you might avoid the analogue summing which is less accurate and has problems of it's own :-)

I try to track 24-bit at conservative levels. It's inevitable that some peak will get close to -0dbfs during a tracking session anyway, so it's better to be safe than sorry. Digital clipping sounds horrible to me.

Then for the mixing process, I do as Paul suggests. To me, it seems better to use the gain control on each channel, leaving the mix buss room to breathe. This way you don't have to attenuate the signal again at the master fader, which may sound better too. The less stages of digital gain adjustment the better, I think. So I trim the channels, then the busses, and leave the master at zero.

Also keep in mind that some of the tracks you're working with might have a lower level than some of the other tracks. So if your DAW channel faders are at zero, you can only make the quiet channel 6dB higher than the rest of the tracks. This may not be enough, which is why it might be a good idea to start your DAW mix with channels down 5dB or so.

I guess YMMV, and it does somewhat depend on the level(s) of the tracks you're mixing.

[gsharp]

Originally Posted by Crushed

Yes. In Nuendo, the first 6 inserts on each channel, including the master are pre-fader, and the last two are post-fader. The rationale is 2buss compression, flowed by dithering on output. Makes sense to me. Can Pro Tools not do this? Seems wrong to me, but what do I know?

Allen

It's probably coming int PT8. In the meantime I just refuse to mix songs that fade.

[maxy]

Great thread I was just asking this question in another thread - glad I looked over here. Seems there is no 1 answer. Guess I'll just keep doing what I've been doing - record with plenty of head room and adjust faders or master as I see fit.

Thanks

[laughingbear]

Yup ....very interesting!

It appears as if it alos greatly depends on the Host, such as Nuendo or PT. I wonder about you Logic Guys and what approach you have....

[djanogil]

Thanks for all these info,
just to clarify;

A master fader set at unity and going in the red indicates that the summing of all the channels creates a level that's overloading the mix bus,
so even if we back down the master fader, we're not tackling the problem at source, we're just disguising it , right?

So that's why I understand it makes sense not to overload the mix buss in the first place, by leaving the necessary headroom for the individual channels.

Am I making sense?

By the way, I'm on Cubase SX

[gsharp]

Originally Posted by djanogil

Thanks for all these info,
just to clarify;

A master fader set at unity and going in the red indicates that the summing of all the channels creates a level that's overloading the mix bus,
so even if we back down the master fader, we're not tackling the problem at source, we're just disguising it , right?

So that's why I understand it makes sense not to overload the mix buss in the first place, by leaving the necessary headroom for the individual channels.

Am I making sense?

By the way, I'm on Cubase SX

You're overloading the output of the mix bus, not the input. Big difference. Set up a mix so that it's distoring like crazy at the mix bus, pull your master way down so you don't get any more reds, then turn your volume up to compensate. If the distortion is not there when you turn the volume up then you can see that pulling the master down does indeed fix it.

[djanogil]

Cool,

I thought distortion (or overloading) existed independently of the mix buss..
Well I've actually experienced what you refer too, loud overloaded mix that didn't sound distorded once the master fader was lowered.
But then I had to back it down a lot to get a good level and that didn't feel quite right either..

In other words what's the best scenario;

Hi levels on the channels and low Master out

or

Low channels levels and hi master out.

??

[gsharp]

you're right in that it doesn't always sound right after lowering the fader, but I've found that happen either way (Lowering the channel faders or the master). My theory is that the distortion/compression at the d-a converter has become part of the sound as you mix into an overloaded 2-bus. When you pull faders down to compensate for the overs the mix can sort of fall apart sometimes.

I still think the best bet is to pay attention to how you hit the 2-bus as you start your mix, just like the console guys do. you know some guys start with the kcik and bass hitting at -10 or whatever. Turn your volume knob up to start. I start loud when I'm getting sounds so I can hear what's going on. If your kick is hitting at -3 you have nowhere to go. I just turn the volume knob up and let the kick hit at -15 or 20 but at the same apparent loudness to me. Start with your 1st elements hitting way below max and you'll have plenty of room and can always turn the whole mix up a bit at the end. Much better than having to turn it down after you've got it slamming.

[djanogil]

Ok ,

I'll try that very soon.
Thanks

[jdunn]

Originally Posted by laughingbear

Yup ....very interesting!

It appears as if it alos greatly depends on the Host, such as Nuendo or PT. I wonder about you Logic Guys and what approach you have....

I'm using Logic, but it's stemmed out to a Phoenix Nicerizer 16. So I'm using 16 busses out of Logic, not quite the same as doing it all ITB.