I understand gain staging well and have a very broad history in analog recording.
I was just reminded of when I recorded jazz. Actually, it was Billy Higgins on his first solo album.
All the instrument levels were set so that they would not peak above 0 VU. Once I had that set, I just let the musicians play. Their dynamics would go from really soft to loud. This was gain staging because I would not let the signal over load.
Then, I had to adjust the playback faders to allow a good live mix. This meant that for some instruments, they naturally needed more level than others. SInce my source was properly controlled, I had to make adjustments the way the musicians wanted to hear.

Then when mixing, I did whatever I needed to balance it all out.
So it all flowed very naturally. I never minded if my faders were at different positions on the board (called desk now). If I needed more gain, I used input trim.

When I was recording on automated Neve boards at Sound City in L.A., most fader levels were not identical. Why should they be?

As a studio drummer having played for years in many pro studios back in the day, the only really concern was not clipping. It was never an issue to try to get all faders at the same position. We used our ears most of time and VU's to double check ourselves.

So I see what Lawrence is saying and what Jeff has been talking about for years now. It's all good. But if you are new to this, don't think you have to master this idea before you actually know how to record at decent levels. Your faders don't have to all be the same. If they end up that way then great. It really is not important until you master the basics.
If you don't understand gain staging, don't worry about. It will come.

Edit: Gain staging is not about fader level but I used that to illustrate that music is dynamic in all its shapes and color. It is not about trying to make everything equal. It is about recording the best level possible for the source audio at hand.

For the mix down process I never mind the faders all being at different levels. With the track rms levels all consistent, the mix is reflected in the fader positions. It allows for good visual feedback.

I know there is some school of thought to get all the faders at unity and create the mix by altering levels elsewhere. I am not a fan so much of this because the faders don't tell you anything anymore. It is good for fader movement though.

But even if a loud track channel fader ends up down at -15 say, you can always insert some attention and bring the fader back up for better finer control up around unity. Tracks already recorded at the ref need no fiddling this way. When a track fader is at -15 though it is telling you this part is quiet. And when that snare fader is up around +3 it is telling you it is slammin! (probably too loud like most drummers! Many hit their snare too hard. It is such a loud drum before you even do anything to it!)

My response to trying to find the best level to record at for the given sound source is assuming the sound source is high quality and the signal to noise ratio is high and distortion low etc, the best level to record at is not at one fixed point like it was in the analog world. In digital it can vary. You can record at say K-12 or K-14 or even K-20. The result is they will all sound the same. What is different is the amount of headroom that exists above. Recording hot is certainly not the answer mainly from the point of view that with acoustic instruments it does not take much to change what you do in order too create a much louder transient and hence clip incoming audio.

Jemusic wroteAlmost every type of signal is good for a VU meter. Only drums are tricky because often they can be fast. Too fast to move a (real) VU meter at all very much. Use Peak metering around drums.

The standard meter on a channel will show you peak and rms. But the rms is a little line floating some distance down. A third party VU meter add on such as the Klanghelm VUMT meter gives you a needle that reaches 0 dB VU when the digital ref is down where it is. e.g. -18.

However the existing Presonus Level Meter can be inserted over input channels and set into K System mode. K System has three references. The closest one to -18 being -20. So put the meter now into K System K20. The meter will now move slower showing rms up around 0 dB VU now where it should be. Peaks will slip past and not show so much like in a real VU.

And yes while the average rms level is being set around 0 dB VU there is 20 dB of headroom above that to take care of peaks. (At -20 that is. At -14 there is only 14 dB of headroom but it is 6 db hotter than K-20 though.) I alternate between K-14 and K-20. You have to recalibrate your VU meters and the SPL level in your room also has to be adjusted when changing cal levels.

A good meter to monitor incoming transient material is the standard Level Meter you mentioned. Let the rms float down low but watch the main meters now. Put it back into TP mode.

thanks for all the input everybody. it took me awhile to read through this and catch up. i actually think i understand what you're all saying. oddly enough the only thing i'm not sure about are what each of the meters in S1 measures?!

in S1 mix console you have:

- an input channel with it's own strip meter
- playback / track channel (std mixer channel) with it's own strip meter

these are peak meters right (the bars indicate the loudest signal)?
however, i don't see any of the additional RMS "white" bars that some are suggesting should be superimposed on these strips even if i turn up the signal really loud?

then you have the presonus meter *plugins* "level meter" and "vu meter". if i set both of these to k12 and give them the same source

- the level meter (which looks like a peak meter to me;)) is measuring RMS right?

- but the vu meter with the same signal at k12 is jumping hot into the red so is that measuring peak or rms? why the difference between what's being shown if the two input signals and k12 setting are identical?

then you mentioned

"A good meter to monitor incoming transient material is the standard Level Meter you mentioned. Let the rms float down low but watch the main meters now. Put it back into TP mode.

i'm not sure what you mean by main meters or TP mode here .. do you mean the console main master strip level? ie are you telling me that i should let the rms float down on the input channel strip meter but look for peaks / clipping in the master fader? i think you're telling me that b/c it's transient you have to let the avg come down to accommodate the dynamic range but check for peaks so that it doesn't clip or get too hot. i'm just not sure what strip / meters i'm checking here for the peaks or what TP mode is?

sorry this is a bit new for me. i just want to really understand what my meters are actually doing if i'm going to use them

thanks!

Maybe this helps:

The loudness of a signal comes from the power of the signal, and can be measured with averaging/rms type meters (VU, K-system). These meters are way too slow for peaks. What is shown as 'peak' values is overshoot from power fluctuations.

The energy of a signal comes from the peaks of the signal, and can be measured with peak (hold) style meters. These meters respond to the highest and fastest top values in the signal. The meters tell you very little about the loudness as there isn't much power in fast high peaks. But it's the peaks that usually cause the clipping, if any.

So peak (hold) style meters tell you how close you are to clipping and VU/K-system style meters tell you how loud the signal is.

In Studio One there are various meters around. The meter on each channel on the mixer page are showing peaks only. Not rms anywhere.

Buss meters are also similar. No rms indication. Only the main mix meter on the stereo buss has the little line that floats down lower and that is the rms level. The top part of the bar graph is showing you peaks.

Sorry for me leaving out some info. The Level Meter I am referring is a Presonus plug-in. Under the Presonus banner as it is bundled with the program. You can insert it anywhere especially on the input signals as the signal is approaching the track. (Click to open the inputs panel when in the mixer page)

There are are no actual proper VU meters (needle movement, not bar graph display) in Studio One. One has to invest in them at this stage. Although there is a free Presonus VU meter but it has some issues. So check out Klanghelm here for example:

https://klanghelm.com/contents/products/VUMT/VUMT.php

There are others from PSP and various. There are free ones too but the ballistic of there meter is not great.

With this type of meter you decide what ref the rms component is going to 0 dB VU. If we choose K-14 for example then the the signal is down at -14 dB for any length of time then the VU will read 0 dB VU. There is now 14 dB of headroom above that reading. (there are other ref levels e.g. -12 and -20 this last one being for best dynamics etc and transients)

The Level Meter also shows rms with a bar graph showing 0 dB instead at the ref. level The Level meter can be put into K System mode so choose K14 and then this meter will show the same rms part as a VU meter does. It is quite effective at this. The Level meter can be put into Peak/rms mode and a display the same as the main outs meter pops up. Back to peak with the little while rms line floating under.

Read the manual on the Level as well. It explains all its functions.

Energy is not a good word to describe what peaks contain over the total length of a sound. It is OK to describe the energy in the actual peak though. I would prefer to think that the guts of the signal or meat or loudness energy is tied up in the rms component. On a basic level, the loudness of something is linked to its rms value.

Peaks contain the transient information in a signal. If a peak is high and fast it usually means that sound is going to hit you harder and faster than other sounds maybe e.g. a snare drum is a fine example. Peaks are gone real quick so they are not contributing the loudness of something so much. But they do contribute to loudness for the duration of the peak and even a small duration can be very loud.

So we do need to keep an eye on peaks everywhere actually and peak meters are always going to keep you informed on this. They are fast in their response and are allowed to track a peak right up to any level.

I just like adding in rms meters as well and keeping rms components consistent everywhere in the production e.g. signals on their way in , on tracks, buses and the main stereo buss. In and out of plug-ins and plug-in chains etc..The rms or VU meters are slow (on purpose) and ignore the peaks which is great. It means what you are looking at is primarily rms.

When you get all your rms levels more consistent you don't have to worry about peaks so much. They will usually take care of themselves. They will all vary above the rms components but usually wont break through into 0 dB FS. But it is still smart to keep a close eye on them with your peak meters.

Energy is exactly the right word. Maybe not in your vocab but it is electrophysically speaking. Just as power is regarding loudness and averaging/rms style meters.

Aside from setting gain levels at the recording stage, which has been covered from every angle here, and I believe good information from all aspects of it (Yes, it all works).

Just an additional note that I think is quite relevant. After your gain "goals" are in check, really take the time to consider what you're recording, and how to stay on top of your signals. Such as, occasionally extreme dynamics "at the time of recording" can put a huge damper on a reliable signal, and compressing afterwards isn't a cure all (though quite helpful). A dynamic recording could benefit from some compression at the inputs. Though some real care here needs to be weighed because you don't get to take it back (relatively). Hard compression would be largely a no-no, but some good use of scaling the hardest hits, and lowest drops "could" reward later on, and definately help a recording. What needs to be considered is the instrument, player's dynamics of the part, and if necessity dictates initial tools (mic, preamp, compression/limiting). It's not a time to add reverb or coloring, because it just not necessary at this stage.

One last thing, and its a bit of an additional insurance policy. Parralel track the recording with no processing. It's like (for you photography buffs, who will understand), shooting multiple exposures. It costs nothing more than a few seconds to add, and allows you the opportunity to compare, or even merge results.

When one is contemplating gain structure, one is also looking at the potential for options, benefits, and oops factors.

The reason I feel Jemusic's contribution to a lot of detail (or emphasis) in preparation of gain staging is important is that when one looks at the consistency of a project, with numerous songs, varied instruments, and takes, being on top of gain structure is only going to help the whole project.

Sure, any numb nuts can watch, and keep a basically descent and even forgiving recorded signal, but some look ahead project, and recording technique care is just going to help the possibly larger picture.

If ones project is large, with multiple songs, and varied dynamics lurking at the recording stage, then look to some help even at the recording input(s). That might include some dynamic processing.

Methinks, but mmv and other paraphrasing.

thx. i think i finally got it. instead of responding separately i'll just summarize for confirmation here.

when tracking in S1, *if* you want to record at a given reference level, put the presonus level meter plugin on the input channel and set your desired K level to monitor average / rms levels. in this type of situation i want the level meter as consistently as possible sitting on 0 (it's averaging and it's leaving me plenty of headroom more or less depending on a higher or lower K setting). i fully understand i'm monitoring the overall loudness or power here (regardless of what one's definition is). i've understood that from the get go here.

simultaneously, keep an eye on the input channel peak meter (or the TP mode of the presonus level meter plugin; or even main stereo master out channel when soloing input channel and not direct monitoring through headphones) to make sure your peaks are ok and not clipping. i understand that this is a given and simple check that should always be made for any and all signals. i also understand you usually have a lot of headroom ( leeway) when setting levels pegged to a k system reference level so in these cases, you should be more concerned about peaks mostly when the signal is transient in nature as focusing too much on the rms isn't a really good indicator (too slow) and can fool you into recording too hot if you're not careful.

that seems pretty rock solid clear to me. yes?

the only thing i'll add is regarding what was said above about the presonus vu meter plugin. although it is an rms meter with k system options and in theory could work in exactly the same fashion as the presonus level meter (rms) if i inserted it on the input channel, apparently it's not very accurate ("has some issues" and even i saw different values between what it was reporting and the presonus level meter were reporting when provided with the same signal and same k setting) compared to the aftermarket Klanghelm vu meter suggested, so it's not a good option. if i want to use a vu style meter don't rely on the presonus one. right?

in sum, the stock presonus level meter plugin provides the basic rms monitoring capability i'm looking. the stock presonus vu meter plugin isn't very good. get a Klanghelm *if* i decide i want to use VU metering.

The Presonus VU meter is free and a nice gift and can be used. I am only being very picky. It's OK in the -12 position when working at K-14. And the -18 position when working at K-20. What is great about it is the fact you can explode it and see a small thumbnail version of it in the mixer page. Very handy for actually seeing VU levels at a glance over your whole session. I have been using two wide screens with my computer lately and really liking it. (I don't think I can go back now either) It is very good having the mixer on its own screen. In the fully vertical mode. The Presonus VU works well in that mode. The workflow is better and faster when you can see the mixer like this.

I would like the Presonus meter to have all three K Systems levels as well as -18. The ballistic is how the needle dances to the music. I have real VU's. Nice quality movements. They look and move great. I am spoiled. It is the nicest movement of all. The ballistics can tell you other things about the signal too. I have devised a series of tests that test the ballistic of a VU meter. It is in this test they move very differently. They all show 0 dB at the ref but that is only part of it.

The Klanghem has got a very close ballistic to the real meter. By fine tuning the settings I can get it nearly perfect. My real VU's are just under my main monitor to the left of centre in a special meter box. (I can monitor any level from any point in my production even right from the input signal direct to my hardware meters) I can put the software versions down on the screen close to them so I can see both at once. (not easy to organise!) Presonus have modelled their ballistic to one of their own meters I have been told. In fact after watching it for a while I am liking it more. It does the job very effectively. There is a setting on the Presonus VU meter (Sensitivity) that will alter the ballistic enough to pull it into line for my liking that is. I just turn it a little anti-clockwise e.g. less sensitive to movement I think and it works well.

Because the K System ref levels are not there either, and the Presonus meter has settings that are only 2 dB off, by the time you set the sensitivity control you can get the needle to move nicely up to and around 0 dB VU as well. So overall it can be used.

The Klanghelm is not expensive though, has a slightly different ballistic and also has a few other functions which are cool. One is being able to add and subtract gain right there with a little screw adjustment. So you can make quick adjustments here and there. There is a Pro version of the Klanghelm meter which gets into filtering and all sorts of things. The improved update V2 of even the standard meter has got some extra ballistic options and scale readouts which are just excellent. e.g. actual rms levels as opposed to VU levels. There is new real slow liquid ballistic I just love now. Good for just seeing where overall rms energy in a signal actually is.

Presonus have added new features to their Level meter lately. It is even better now. You have Loudness readings now as an addition. Highly recommended to check. I also like the K system mode display in the Level meter. It is easy to see and has a nice ballistic for a bar graph display which can often be too fast at times. They have slowed it down to match what a VU would be more likely to do.

jemusic thx for the detailed info regarding the VU meters. in my previous post i summarized my final understanding of using various presonus meters while recording. i'm pretty sure it's spot on but i was also looking for some confirmation on my understanding from the crowd here if what i said summarizes well and makes good common sense. thx all

-12 db at the input stage and then, pray for good notes.

After that, use the output gain of plugins etc., to maintain that same -12 db level all the way to the mix stage. That is not always practical, but it's useful in guiding your process from start to finish so that when you turn off or remove plugins, the actual output level is not affected.

This is all I know. The rest I make up as I go along.

I love the Klanghelm meters, too. Very useful little plugins.
A few features that maybe aren't immediately apparent: they can be used as "trim" plugins, which can be automated, and the plugin can be used for VU or PPM in a single click, can be calibrated easily and yet easily reset to the defaults by a single click on the numbers (-18 on VU, -9 PPM), the PPM meter can also be calibrated to read extrememly fast: can read dirac spikes the length of a single sample, and the single pane version appears to be mono - but can be set to read stereo so that two "needles" appear in the single pane - which saves valuable screen real estate.
Great little plugin!

(I don't work for Klanghelm - but that little plugin really works for me!)

matthewgorman wroteGAIN STAGING GAIN STAGING GAIN STAGING GAIN STAGING GAIN STAGING (That was for Lawrence)

LOL

on most recordings it is easy to set levels for recording but vocals are tough... 9 times out of 10 the singer moves in and out of the mic causing levels to fluctuate. even if you set a decent level at one point in space it will not stay that way. actually imo gain staging has nothing to do with recording. it is simply a process of leveling audio for the next plugin in the chain.

p.s if there is a two word phrase i hate the most... it is a toss up... 'game changer' and 'paradigm shift' everyone and their brother uses it as if it makes them sound important. as for 'gain; staging' it has been worn out too but takes a back seat to the two previous phrases. cheers


jay

Jbrannam wrote: p.s if there is a two word phrase i hate the most..

For Jay......

"game changer", "paradigm shift", "gain staging" "game changer", "paradigm shift", "gain staging" "game changer", "paradigm shift", "gain staging" "game changer", "paradigm shift", "gain staging" "game changer", "paradigm shift", "gain staging" "game changer", "paradigm shift", "gain staging" "game changer", "paradigm shift", "gain staging"



Hah! 

Fact is, you're right Jay. Too much emphasis. That's why I added the points about what to consider upon initial recording (benefits of pocessing, and acquiring headroom = increased dynamic range). Not thinking you have some pre conceived notion of gain staging. That's just a fundamental element. You can't elaborate on it without adding more it's usefulness, and the bigger picture.

Far be it, anyone listens to logic. People now need instant gratification. Google it before you even considers its context. "Oh, that's what that is?" "Good thanks". "Well, now I know".

Put this post right up there with "How do you rename a file?".
"Manual? What's a manual?. I'm a (bedroom) producer ".

Lol.

Nothing wrong with the subject matter, but the over emphasis is a bit much.

you're right jim... about instant gratification. it is sad and thnx for the laugh man that was funny cheers

p.s. but if it is any help... like roland said.. -12 db is a great spot. i aim for it be it audio recording or even dumping/printing a waveform (audio) from a sample library using midi. you get the meat of the element and is workable from there either way (up or down)

Thx Jay. Glad your back up and running with your new PC. Nice ride!

Hey, if you or anyone need to find out more about gai , I mean the G word, the OP posted now...

Gain Level Post Mix. Then Mastering.


Does anyone have any info on Gain levels between final release submittal, and pre press? I believe there's some meter reading involved.

I'll start a post.

Sorry, couldn't resist.

If you can try to get a reasonably close accurate rms setting at tracking stage you are part of the way there. But as jay mentioned things may be more wild at the time of recording. I have dealt with dynamic singers too. If they go very loud at times, and I am still not sure, then a hardware limiter set to just clamp any very loud bits to a ref level anyway is a start. Light conditioning compression on the way in can be nice too. But take care here. Tracking with a careful eye on a VU meter also helps.

The digital system will handle an enormous range of levels during recording. As long as the loudest parts do not clip into 0 dB FS. Extremely loud parts could reach -6 dB FS. (a little more is better here)

By the way remember Studio One has 3 dB of extra headroom at the time of recording. Once a peak light comes on, in an editor you would expect something very close to 0 dB FS but it is in fact at -3 dB. That is good to know. It is often just enough to save a peak from not being clipped compared to being clipped.

I like to edit prior to mixing. Opening up the vocal track and checking it with an editor and a VU meter. I edit all over the place bring the loudest things down a little, very softer things up, overall rms level of the whole track up 2 or 3 dB. Even loud parts of words can be lowered and overall phrases lifted by a db here and there. The overall vocal track is now checked with a VU meter (set to the ref) and adjusted to the reference. Dynamics do not have to be killed here either. Louder sections can still change and increase in power. Just not as much as they did. Things can get softer here and there too.

You end up with a very nice vocal part prior to mix. Backed up with a new name of course. I add in extensions to the name e.g. ED 1..ED 2 etc.. That way I know at what stages the vocal track has been edited to. You may have 2 or 3 versions of the final edited vocal tracks to choose from.

This is what you start your mix vocal sound with. Good editing beats plugins you need on a track to maybe match the edit or not come quite as close. Less plugins are needed.