Gain Staging & Mixing Pre-Mixed/Mastered Stems

[isawsasquatch]

Originally Posted by Sampire

when mixing, I typically as a rule, gain stage every track to around -17 db...sometimes I question this process though when considering that some of these stems may be for example, drum breaks that have already been mixed and mastered....

what do you guys do?

Make it sound good without any clipping anywhere. That's all you need to know.

[SNVP]

Originally Posted by Sampire

when mixing, I typically as a rule, gain stage every track to around -17 db...sometimes I question this process though when considering that some of these stems may be for example, drum breaks that have already been mixed and mastered....

what do you guys do?

It's all about headroom, the more the better....!

Signals are signals regardless of their source, and so they all must follow the gain structure of the track,

level is the first point of EQ and so it is important to gain stage each signal appropriately.

Its perfectly normal to sample mastered material, it just becomes another signal in your project.

When mixing I like to leave as much headroom as possible !

You may find this interesting

The Reason Most ITB mixes don’t Sound as good as Analog mixes (restored)

[Red Black]

For steady-state signals I level them with a VU meter to about -18dBFS RMS. For transient signals I usually use peak metering and let them peak no higher than -6dBFS. This is a great starting point for me. If I was dealing with a pre-mixed source including drums then I'd most likely use peak metering as previously mentioned. If it had already been mixed and mastered then I may use subtle de-compression or expansion before levelling to -6dBFS.

I always use clip gain for these level adjustments as well.. for a lead vocal I would have already done a fair bit of manual dynamic range management just using clip gain and a VU which leaves less work for the compressors.

As I start to process the tracks I make sure the perceived level stays the same on the tracks before and after any processors.

For mixing I tend to use K-20 metering and I also use an auto-mute plug-in on the mix bus. I never have to worry about peak level using K-20 but if I did hit 0dBFS somehow the auto-mute plug-in would kick in and I would instantly know my mix needs to be adjusted. It's basically like a little virtual fuse that will blow when the signal level reaches a pre-defined threshold and then it repairs itself a few seconds later! Awesome! Infinitely better than using a limiter to 'catch' your errors in my humble opinion..

Cheers..

RB

[Red Black]

Originally Posted by isawsasquatch

Make it sound good without any clipping anywhere. That's all you need to know.

I used to share that attitude but there's a lot of workflow benefits that come from gain-staging your levels in my opinion.. especially when it comes to things like compression and also being able to easily repeat sounds you've achieved in the past. Another benefit is that I can send signals straight into analogue-modelled plug-ins or external hardware at the level they were designed to operate at and I don't even have to think about it. Everything is more controlled and more consistent which leaves you more time and energy to focus on the music!

I should also mention that if I've recorded something myself I've already set the proper levels at the recording stage so there's no extra effort required.. it's already done, the tracks are already at the nominal level. It's only when I get tracks recorded elsewhere by other engineers that I need to adjust the levels.

RB

[isawsasquatch]

Originally Posted by Red Black

I used to share that attitude but there's a lot of workflow benefits that come from gain-staging your levels in my opinion.. especially when it comes to things like compression and also being able to easily repeat sounds you've achieved in the past. Another benefit is that I can send signals straight into analogue-modelled plug-ins or external hardware at the level they were designed to operate at and I don't even have to think about it. Everything is more controlled and more consistent which leaves you more time and energy to focus on the music!

I should also mention that if I've recorded something myself I've already set the proper levels at the recording stage so there's no extra effort required.. it's already done, the tracks are already at the nominal level. It's only when I get tracks recorded elsewhere by other engineers that I need to adjust the levels.

RB

There are indeed a number of benefits to gain-staging, but at the end of the day, the object of it all is to avoid digital clipping. Rather than overcomplicate things for the OP (who, from the sound of it, has little engineering experience and/or technical understanding) by throwing a bunch of random values and K System mumbo jumbo his way, it's more than enough to simply say, "don't clip anything anywhere".

The only time such technical detail is necessary is when interfacing with analog gear, but even then, why not simplify it and suggest the user gain stage everything to 0VU. That would serve an identical function, plus it would remove any convertor calibration issues that may render your digital values meaningless.

[Sampire]

so far I've understood all "jargon" thrown around in this thread....lol....it makes me sad that apparently I come across as ignorant?

my mixes sound great imo and others who pay me to do so must think so also, so....it's not like I don't know what I'm doing - I just from time to time question standard practices and would like to discuss my thoughts on here - I have no official schooling in audio engineering like some of you may, but I get by

thanks for the input fellas!

[Red Black]

Originally Posted by isawsasquatch

There are indeed a number of benefits to gain-staging, but at the end of the day, the object of it all is to avoid digital clipping. Rather than overcomplicate things for the OP (who, from the sound of it, has little engineering experience and/or technical understanding) by throwing a bunch of random values and K System mumbo jumbo his way, it's more than enough to simply say, "don't clip anything anywhere".

I guess for me personally avoiding clipping is really the least of my concerns; I practise gain-staging specifically for workflow and consistency reasons.. I also use a few analogue modelling plug-ins and like hardware they (usually) prefer to be hit at analogue levels.

The level I specified wasn't random either, it was the exact level that every analogue-modelling plug-in I know of is designed to be operated at and the exact level that every VU meter plug-in I know of is calibrated to by default. I wish someone had given me similar advice years ago but instead people told me to just "not clip anything" and that's exactly how I got some bad habits.

Not trying to be a dick or a contrarian either, I just genuinely disagree with that attitude and I believe it's technically wrong and perhaps even a little lazy! :]

Originally Posted by isawsasquatch

The only time such technical detail is necessary is when interfacing with analog gear, but even then, why not simplify it and suggest the user gain stage everything to 0VU. That would serve an identical function, plus it would remove any convertor calibration issues that may render your digital values meaningless.

For me that's not true.. it's all about workflow and consistency of levels, not interfacing with external gear.

I would have specified 0VU but I didn't think that wasn't specific enough, in that it might invite the question "What's 0VU?". Specifying -18dBFS RMS is a lot clearer to most people..

Cheers..

RB

[jrhager84]

-18dbfs RMS, -6 peak here.

Sent from my SAMSUNG-SGH-I777 using Tapatalk

[bgrotto]

Originally Posted by Red Black

I guess for me personally avoiding clipping is really the least of my concerns; I practise gain-staging specifically for workflow and consistency reasons.. I also use a few analogue modelling plug-ins and like hardware they (usually) prefer to be hit at analogue levels.

The level I specified wasn't random either, it was the exact level that every analogue-modelling plug-in I know of is designed to be operated at and the exact level that every VU meter plug-in I know of is calibrated to by default. I wish someone had given me similar advice years ago but instead people told me to just "not clip anything" and that's exactly how I got some bad habits.

Not trying to be a dick or a contrarian either, I just genuinely disagree with that attitude and I believe it's technically wrong and perhaps even a little lazy! :]




For me that's not true.. it's all about workflow and consistency of levels, not interfacing with external gear.

I would have specified 0VU but I didn't think that wasn't specific enough, in that it might invite the question "What's 0VU?". Specifying -18dBFS RMS is a lot clearer to most people..

Cheers..

RB

By and large I agree with everything you've said in principle, but I'm curious how you know the exact digital levels your analog modeling plugins are meant to be operating at.

[Red Black]

Originally Posted by bgrotto

By and large I agree with everything you've said in principle, but I'm curious how you know the exact digital levels your analog modeling plugins are meant to be operating at.

Just from reading the manuals! :]

RB

[bgrotto]

Originally Posted by Red Black

Just from reading the manuals! :]

RB

Heh. Imagine that! Such a novel idea