The No.1 Website for Pro Audio
Gain staging for dummies?
Old 5th October 2018
  #1
Gear Maniac
 

Arrow Gain staging for dummies?

I watched about 25 videos and read dozens of forum posts about gain staging but it seems like no one agrees with each other.

Maybe there's no definitive answer but I'm looking for the Keep It Simple answer.

FYI:
I'm talking about digital gain staging and gain staging recorded audio and midi clips.


Now, I never gain staged at all because I simply wasn't aware of what gain staging is and my DAW has no pre fader metering.


Researching the topic I've found conflicting answers.

Some say you should get ALL tracks to -18dBFS before you start your mixdown.

Others say you should gain stage before arranging (which imo is the most logical approach cause I want to EQ and tailor sounds from the start).

Then there's a group of people that say input signal clipping in DAWs doesn't matter at all. Only the master channel is important.


My theory:
Input gain doesn't matter when it's below 0 and has enough headroom for EQ adjustments and other effects that add gain.


Is my theory right or am I missing something?
Old 5th October 2018
  #2
Quote:
Now, I never gain staged at all because I simply wasn't aware of what gain staging is and my DAW has no pre fader metering.
You actually did, but you just didnt know you were doing it.
Gain staging happens pre and post.
Simply put, Gain staging happens at every path that can alter the gain in audio or MIDI. This goes for DAW;s and analog

Gain Staging (structure) refers to the signal level as it moves form its source to its final destination. Along this path, you can have points where signal level changes can be made. Monitoring the strength in each point is a must! You do not want to clip or over saturate a gain stage.


An example of a gain stage:
Voice to Microphone to Preamp to Compressor to Sound Card. With this normal vocal recording chain, you have 5 possible ways to change and alter the signal strength. The strength of your voice and the positioning of the mic to your mouth is a gain stage. The mic is also a gain stage, because it can have pad settings on it. All these components affect the signal strength.

You have to experiment with many approaches to find what works best for you and your set-up. Note, that not every approach will work for every situation. You need to trust your ears. If you're getting a great sound and your settings don't look right, that's OK, as there are no rules to gain staging. Over time you will build confidence when setting up gain structures. Most important is to use your ears and do not clip your audio signal.


Things To Watch Out For:

If your input level is too high, you track fader may have to be very low. This can make it too low to have control in the upper part. When your track father is low, its very difficult to adjust and fine tune the audio levels.
If your preamp level is too high, the signal can overdrive the sound card's input or the next gain stage in the signal path. Preamp settings are the upmost important. A bad preamp setting will result in failure
If your preamp level is too low, your track fader will have to be too high and you can get a bad signal to noise ratio.
Old 5th October 2018
  #3
Gear Maniac
 

Quote:
Originally Posted by CJ Mastering View Post
You actually did, but you just didnt know you were doing it.
Gain staging happens pre and post.
Simply put, Gain staging happens at every path that can alter the gain in audio or MIDI. This goes for DAW;s and analog

Gain Staging (structure) refers to the signal level as it moves form its source to its final destination. Along this path, you can have points where signal level changes can be made. Monitoring the strength in each point is a must! You do not want to clip or over saturate a gain stage.


An example of a gain stage:
Voice to Microphone to Preamp to Compressor to Sound Card. With this normal vocal recording chain, you have 5 possible ways to change and alter the signal strength. The strength of your voice and the positioning of the mic to your mouth is a gain stage. The mic is also a gain stage, because it can have pad settings on it. All these components affect the signal strength.

You have to experiment with many approaches to find what works best for you and your set-up. Note, that not every approach will work for every situation. You need to trust your ears. If you're getting a great sound and your settings don't look right, that's OK, as there are no rules to gain staging. Over time you will build confidence when setting up gain structures. Most important is to use your ears and do not clip your audio signal.


Things To Watch Out For:

If your input level is too high, you track fader may have to be very low. This can make it too low to have control in the upper part. When your track father is low, its very difficult to adjust and fine tune the audio levels.
If your preamp level is too high, the signal can overdrive the sound card's input or the next gain stage in the signal path. Preamp settings are the upmost important. A bad preamp setting will result in failure
If your preamp level is too low, your track fader will have to be too high and you can get a bad signal to noise ratio.
Thanks for replying.

But what I still don't get is what do I actually have to do for proper gain staging (recording aspects aside. just starting with some loops inside my DAW).

For me the simple, easy way seems like selecting all tracksw with faders left at 0dB and reducing the gain to an arbitary number, so that all tracks are far from
clipping. Should take 20 seconds.

Am i oversimplifying the process or is that a legit way of gain staging?
Old 5th October 2018
  #4
Lives for gear
 
BT64's Avatar
There are three things to keep in mind with digital signal levels.

The moment you make it digital.
When recording you want to stay away from the 0dBFS, your A/D converter (audio interface) has a fixed bit depth so there is nothing above 0dBFS.
Not to soft ever because you want to keep your signal a fear distance from the noise floor.

In your DAW.
If your DAW is working with floating-point (32 or 64 bit) the signal will not clip, but..
Plug-ins (especially the ones that emulate old gear) have a sweet level, just like there analog counter partners.
If you hit them to soft or to hard they sound different and can distort, saturate or sound lame.

The moment you go to a fixed bit depth.
If you export to 24 or 16 bit 0dBFS is again the maximum limit.
Old 5th October 2018
  #5
Quote:
For me the simple, easy way seems like selecting all tracksw with faders left at 0dB and reducing the gain to an arbitary number, so that all tracks are far from
clipping. Should take 20 seconds.
For recording yes, for mixing no..

When you record, you control the gain with everything before the A/D conversion in your sound card. So you control the gain with everything before that. The track faders in your DAW have no control over the gain when recording.
Old 5th October 2018
  #6
Quote:
For me the simple, easy way seems like selecting all tracksw with faders left at 0dB and reducing the gain to an arbitary number, so that all tracks are far from
clipping. Should take 20 seconds.
For recording yes, for mixing no..

When you record, you control the gain with everything before the A/D conversion in your sound card. So you control the gain with everything before that. The track faders in your DAW have no control over the gain when recording.
Old 5th October 2018
  #7
Gear Maniac
 

Quote:
Originally Posted by CJ Mastering View Post
For recording yes, for mixing no..

When you record, you control the gain with everything before the A/D conversion in your sound card. So you control the gain with everything before that. The track faders in your DAW have no control over the gain when recording.
Yes, I understand that. That's why I said that the faders are all at 0dB and left untouched. I would reduce gain via the first plugin in the chain or with a dedicated gain plugin.

That is how it's supposed to be done, right?

As I said before I'm not referring to recording. Only mixing.


Also: Do you gain stage right at the start of a session or after arranging is finished?
Old 5th October 2018
  #8
Quote:
Yes, I understand that. That's why I said that the faders are all at 0dB and left untouched. I would reduce gain via the first plugin in the chain or with a dedicated gain plugin.

That is how it's supposed to be done, right?
No its not right. I said in recording you leave your DAW's faders to 0Db because trey do no effecy anything when you record..

In mixing you move your faders and what ever else needs to be adjusted, like buses, sends plugins effects. Everything in your chain needs to be gain staged.
Gain staging starts when you start recording and it ends when you finish your song

CJ
Old 5th October 2018
  #9
Gear Maniac
 

Quote:
Originally Posted by CJ Mastering View Post
No its not right. I said in recording you leave your DAW's faders to 0Db because trey do no effecy anything when you record..

In mixing you move your faders and what ever else needs to be adjusted, like buses, sends plugins effects. Everything in your chain needs to be gain staged.
Gain staging starts when you start recording and it ends when you finish your song

CJ
Okay, thank you. Bear with me. It might be obv. to you but I don't get it quite yet.

The channel faders are for leveling the output volume of each channel, right?

If I had a track that doesn't clip but does after I EQ and compress I would have to lower the input gain to prevent the signal from clipping.

Lowering the channel fader would just reduce the volume of the output where the signal is already clipped.

What am I missing here?
Old 5th October 2018
  #10
Gear Guru
 

Quote:
Originally Posted by Hip Hop Head View Post
Okay, thank you. Bear with me. It might be obv. to you but I don't get it quite yet.

The channel faders are for leveling the output volume of each channel, right?

If I had a track that doesn't clip but does after I EQ and compress I would have to lower the input gain to prevent the signal from clipping.

Lowering the channel fader would just reduce the volume of the output where the signal is already clipped.

What am I missing here?
You're right.

I think CJ is maybe thinking you're misunderstanding something.
Old 5th October 2018
  #11
Gear Guru
 

Quote:
Originally Posted by Hip Hop Head View Post
I would reduce gain via the first plugin in the chain or with a dedicated gain plugin.

That is how it's supposed to be done, right?
Well, if you need your levels that go into your insert-chain to be at a certain level then you could definitely do that. And some DAWs have "trim" at the beginning of a channel so that's an option, as is adjusting "clip gain" (or "event gain" depending on nomenclature). But you basically have the flow right.

One thing to note is that a lot of engineers prefer to adjust gain at the output of every plugin since that makes it easier to A/B using bypass. Otherwise if a plugin slightly boosts the level then one might like the actual level increase rather than whatever the effect applied is (i.e. it's not the EQ boost at X Hz that's nice, it's the level change... ). This of course doesn't negate what you wrote as a practice, just keep in mind that it's another place to do it for reasons of convenience.

Quote:
Originally Posted by Hip Hop Head View Post
Also: Do you gain stage right at the start of a session or after arranging is finished?
I don't do music as much as I used to, but yeah, in some cases it's convenient to just drop the level of everything right off the bat.
Old 6th October 2018
  #12
Lives for gear
 
BT64's Avatar
Quote:
Originally Posted by Hip Hop Head View Post
If I had a track that doesn't clip but does after I EQ and compress I would have to lower the input gain to prevent the signal from clipping.
Lowering the input will change how the plug-ins will work, think about a compressor with lower input what that does with the threshold.
Gainstage every step in the chain.
Level your mix with the faders.
Old 24th February 2020
  #13
Here for the gear
Sorry to revive such an old thread but I think my questions are relataed to the op one, in some way.

I recorded an acoustic guitar track played in fingerstyle (Paul Simon fashion) and after I download it to the Daw (ableton)
I can see that the average level is mostly between -31dB to -26dB and the peak is at -10dB but I also see that there are a few bass notes (Gs) that in some places are a 2/3 dBs louder than the average of the rest.

To do the correct gain staging I understand that first of all I should apply a gain plugin and set it so that the peak goes to -6dBFS,
but this would end up having the most part of the track between -27dB to -22dBs, 6dBs lower than what it should be (-18dB): correct?

I can automate a first gain plugin to reduce only the loudest notes by 3dBs and then use a secong gain plug in to take the peak to -6dBFS, after the automation,
but that could be time consuming so, would it be ok to have only one gain plugin set to 6dB to take the average to -18dB and then insert a dynamic EQ (FabFilter Pro - Q3) set with dynamic notch reduction on the frequency of bass G (98 and 196Hz) and then, put a new gain plugin and set it to have the peak at -6dB after the EQ?
If this approach is not correct, which one would?
Thanks.
Old 24th February 2020
  #14
Lives for gear
 
BT64's Avatar
There's no "should be".
An dynamic recording is dynamic.
If it's to dynamic for the goal you can reduce
(compress) it to your licking.
Old 24th February 2020
  #15
Gear Guru
 

Quote:
Originally Posted by fabiospark View Post
Sorry to revive such an old thread but I think my questions are relataed to the op one, in some way.

I recorded an acoustic guitar track played in fingerstyle (Paul Simon fashion) and after I download it to the Daw (ableton)
I can see that the average level is mostly between -31dB to -26dB and the peak is at -10dB but I also see that there are a few bass notes (Gs) that in some places are a 2/3 dBs louder than the average of the rest.
There is nothing inherently wrong with the above as long as it sounds good.

Quote:
Originally Posted by fabiospark View Post
To do the correct gain staging I understand that first of all I should apply a gain plugin and set it so that the peak goes to -6dBFS,
You don't have to do that. There's nothing more "correct" about that than -6dBFS.

Quote:
Originally Posted by fabiospark View Post
but this would end up having the most part of the track between -27dB to -22dBs, 6dBs lower than what it should be (-18dB): correct?
mm... it doesn't have to be -18dBFS average. It's just useful in some cases to aim for that as a ballpark value. You can make up gain in different places and typically people use dynamic processing in which they tend to boost levels anyway.

Quote:
Originally Posted by fabiospark View Post
I can automate a first gain plugin to reduce only the loudest notes by 3dBs and then use a secong gain plug in to take the peak to -6dBFS, after the automation,
but that could be time consuming so, would it be ok to have only one gain plugin set to 6dB to take the average to -18dB and then insert a dynamic EQ (FabFilter Pro - Q3) set with dynamic notch reduction on the frequency of bass G (98 and 196Hz) and then, put a new gain plugin and set it to have the peak at -6dB after the EQ?
If this approach is not correct, which one would?
Thanks.
Typically people would probably use a compressor instead, along with an EQ to shape the tonal balance. But certainly what you describe above isn't something you should do for the reasons you describe.
Old 26th February 2020
  #16
Company Rep
 
puremix.net's Avatar
 

Quote:
Originally Posted by Hip Hop Head View Post
I watched about 25 videos and read dozens of forum posts about gain staging but it seems like no one agrees with each other.

Maybe there's no definitive answer but I'm looking for the Keep It Simple answer.

FYI:
I'm talking about digital gain staging and gain staging recorded audio and midi clips.


Now, I never gain staged at all because I simply wasn't aware of what gain staging is and my DAW has no pre fader metering.


Researching the topic I've found conflicting answers.

Some say you should get ALL tracks to -18dBFS before you start your mixdown.

Others say you should gain stage before arranging (which imo is the most logical approach cause I want to EQ and tailor sounds from the start).

Then there's a group of people that say input signal clipping in DAWs doesn't matter at all. Only the master channel is important.


My theory:
Input gain doesn't matter when it's below 0 and has enough headroom for EQ adjustments and other effects that add gain.


Is my theory right or am I missing something?

Hey maybe you can find an answer on our blog! take it away https://www.puremix.net/blog/gain-st...r-session.html
Old 26th February 2020
  #17
Gear Guru
 

Quote:
Originally Posted by puremix.net View Post
Hey maybe you can find an answer on our blog! take it away https://www.puremix.net/blog/gain-st...r-session.html
Old 27th February 2020
  #18
Here for the gear
 
Patog's Avatar
Someone directed me to this post when I asked how hot should the output signal from my interface be when using an external hardware (a Warm WA76 in this case). If the input of the compressor is line level, should the signal be +4dBU to get the "Sweet spot"? So, should it be peaking at -18 dBFS in my DAW when it comes out?

Thanks!
Old 27th February 2020
  #19
Gear Guru
 

If your average level hovers around 0VU you're hitting nominal operating level, meaning the device is operating where it supposedly sounds the "best", which could be "cleanest" or whatever. As you drive it higher you obviously get more distortion eventually.

Most likely if you have peaks at -18dBFS in your DAW at the output then that's going to be far lower than the 0VU on your device. Even though the meter only goes to "+3VU" on your device you actually have a lot more headroom than that before ("bad") distortion, so if your signal has an average of -18dBFS instead of peaks, and the peaks now are -6dBFS for example, that's probably still well within the headroom of that compressor (assuming that the reference level of your converter is -18dBFS=0VU or +4dBu).
Old 27th February 2020
  #20
Here for the gear
Quote:
Originally Posted by mattiasnyc View Post
There is nothing inherently wrong with the above as long as it sounds good.
No, those outplayed notes don't sound good. But the rest of the performance can be ok so...
What I meant was that, being an amateur guitar player, there can be few notes that came out far too loud than what they should, so, if I don't lower them they will affect the peak readings of all the track and could lead to some distortion.
To avoid that, I automated the gain of a dedicated Utility audio effect (Ableton) put right at the beginning of the plugin chain to reduce the gain only locally where I need it.
I'd rather do that than split the clip and apply a gain reduction to the split clip so the original clip is left untouched.
Any advice on why that is a wrong approach (if it is)?
And what should I do instead to get the best starting point to mix?



Quote:
Originally Posted by mattiasnyc View Post
You don't have to do that. There's nothing more "correct" about that than -6dBFS.
Sorry, I don't undestand your reply (Bear with me but I'm not an English mother tongue, so maybe it's due to that).



Quote:
Originally Posted by mattiasnyc View Post
mm... it doesn't have to be -18dBFS average. It's just useful in some cases to aim for that as a ballpark value. You can make up gain in different places and typically people use dynamic processing in which they tend to boost levels anyway.
Reading around I understood that an average level of -18dBFS (0dBVU) often is the best level to get into a plugin/effect so I thought that, before I start to apply effects or plugins, after gain staging, I'd like to take the average level around -18dBFS.
Also, as I often record the same instrument setup, in this way, for each effect I can easily reuse the custom presets I saved previously already knowing that the dynamic thresholds are in the right ballpark.
Again, anything wrong here?



Quote:
Originally Posted by mattiasnyc View Post
Typically people would probably use a compressor instead, along with an EQ to shape the tonal balance. But certainly what you describe above isn't something you should do for the reasons you describe.
The EQ I'm using lets you make a band dynamic setting an independent level threshold for each band, so, I see each band as a compressor + EQ, as you wrote.
I'm recording at home in an untreated room and I found out that, in that setup, my recorced guitar presents a few overboosted frequency, so, with this tool I can tame even the narrowest frequency and only if it goes above a certain level, with a little or null effect on the rest of the spectrum.
That's why this is always the first plugin I put after gain staging.


Let me add a new question about the levels of recorded tracks.

I record with a digital recorder and than I download the tracks to Ableton via USB.
Once in Ableton, often I can see that, with the acoustic guitar track, the "distance" between the average level and the peak level is more tha 12dB so, if I take the track average level to -18dBFS (0dBVU), the Ableton peak meter can show -2dBFS or, sometimes, even a few points above zero. Can I safely leave it like that and don't pay attention to that meter?
(If I understood correctly, that meter measures the single samples and so it can "hear" single peaks that won't cause distortion).
If not, shall I have to reduce the guitar track dynamics at first?

Thanks.
Old 27th February 2020
  #21
Gear Guru
 

Quote:
Originally Posted by fabiospark View Post
I automated the gain of a dedicated Utility audio effect (Ableton) put right at the beginning of the plugin chain to reduce the gain only locally where I need it.
I'd rather do that than split the clip and apply a gain reduction to the split clip so the original clip is left untouched.
Any advice on why that is a wrong approach (if it is)?
And what should I do instead to get the best starting point to mix?
There's nothing inherently wrong with that, it's just that a fair amount of people would use compression instead, and the reason for many to use compression is that it's easier and faster, and that it can change the sound of the instrument for the better. So that's why a fair amount of people might not automate the gain.

But again, there's nothing wrong with that and if it works for you that's good.

Quote:
Originally Posted by fabiospark View Post
Sorry, I don't undestand your reply (Bear with me but I'm not an English mother tongue, so maybe it's due to that).
My mistake, I think I was editing the post and got my numbers mixed up so the sentence I wrote doesn't really make sense.

What I meant to say was that when signals get too loud they can clip in digital (fixed point), and that ends up being 0dBFS. So it doesn't matter as far as clipping goes if your peaks are at -6dBFS, -12dBFS or -2dBFS because none of them will clip the signal at the output. Your impression seemed to be that you had to get your peaks to -6dBFS and I was simply saying that you don't have to do that.

Quote:
Originally Posted by fabiospark View Post
Reading around I understood that an average level of -18dBFS (0dBVU) often is the best level to get into a plugin/effect so I thought that, before I start to apply effects or plugins, after gain staging, I'd like to take the average level around -18dBFS.
Also, as I often record the same instrument setup, in this way, for each effect I can easily reuse the custom presets I saved previously already knowing that the dynamic thresholds are in the right ballpark.
Again, anything wrong here?
Nothing wrong. My emphasis was on "don't have to". In other words if it's close enough that's fine. It's really mostly a practical matter. Something tells me most bread-and-butter plugins that people use aren't programmed to distort or act non-linearly and therefore don't have a 'sweet spot' at around 18dBFS. So it's a practically convenient place for your average signals, but it's not like you end up with problems automatically if you are off.

Quote:
Originally Posted by fabiospark View Post
Let me add a new question about the levels of recorded tracks.

I record with a digital recorder and than I download the tracks to Ableton via USB.
Once in Ableton, often I can see that, with the acoustic guitar track, the "distance" between the average level and the peak level is more tha 12dB so, if I take the track average level to -18dBFS (0dBVU), the Ableton peak meter can show -2dBFS or, sometimes, even a few points above zero. Can I safely leave it like that and don't pay attention to that meter?
(If I understood correctly, that meter measures the single samples and so it can "hear" single peaks that won't cause distortion).
If not, shall I have to reduce the guitar track dynamics at first?

Thanks.
I don't use Ableton so for specifics you'd have to check the manual or with other users.

Signals in modern DAWs are typically floating point, and that allows for signals to go 'above' 0dBFS. Sooner or later however we convert it to fixed point which is what CDs are and streaming etc, and then the limit is back to a solid 0dBFS. So again from a practical standpoint it's useful to keep your peaks throughout your chain below that point. If you don't you'll have to deal with them sooner or later anyway.

To put it differently: If you do what you said you did above and you have a +1dBFS reading for example then that's possibly not a problem anywhere in your signal chain until it finally gets converted to your interface output, which is most likely fixed point and limited to 0dBFS. Then you get some sort of clipping.

So yes, I'd say it's a good idea to deal with peaks that are 'too loud' sooner than later.
Old 27th February 2020
  #22
Here for the gear
Thanks for your time and for your replies.

f.
Topic:
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump