How Legendary Engineers Set Input Levels

Mix engineer Dave Stagl examines input and recording levels from legendary engineers including Bob Clearmountain, Joe Chiccarelli, and Al Schmitt.
◎ MIXING & MASTERING in Dolby Atmos & Stereo: www.staglproductions.com
◎ Training & Consulting: www.goingto11.com
◎ 5 Tips for a Great Dolby Atmos Release: tinyurl.com/y76pdt2u
Sonarworks: tinyurl.com/bdf4a7y8
SoundFlow: soundflow.org/invite?fpr=293i6
Waves Plug-Ins: waves.alzt.net/93xg4
FabFilter Plug-Ins: www.fabfilter.com/shop?code=W...
---
SOCIAL MEDIA
/ fohdave
/ fohdave1
---
00:00 - Start
00:52 - Bob Clearmountain Session
02:00 - Metering Considerations
03:06 - Bob's Levels
05:40 - Why Push Levels?
07:30 - Joe Chiccarelli Session
07:52 - Joe's Levels
12:32 - Al Schmitt Session
12:59 - Al's Levels
16:19 - The Big Level Takeaway
17:20 - Wrap-Up

[@-303-]

-18 dBFS (or -20 or whatever) is the audio equivalent of wearing your seatbelt. It protects against unintended and unwanted consequences of using non-linear processing that may occur with things like emulations of analog equipment. You can always increase the gain before the device using some sort of utility device if you want the extra saturation or something. A good habit saves your brain for more important decisions.

[@MarkusAudio]

First video of yours I watch, instant sub 🙂 After all these years I never knew the difference between VU and Digital meters readings. And getting to know Al Schmidt was the cherry on the cake 🙂

[@deniscadieux1676]

Great video! Al is king! He probably records stuff a little lower because he records mostly jazz and acoustic stuff, so he doesn’t want any of coloration and harmonic distortion that you would get by saturating the console. Also, he never uses compression on the way in, and most often not in mixing as well, so he plays it safe to not hit overs in the digital world. I believe He was recording 24 bits/96k most of his last recordings before he past away with Steve Genewick so it kind of explains his conservative levels. The resolution is there and he using all of it.

[@AdamGotheridge]

That was a really interesting angle, and I'm going with Bob and the overs.

[@DamonZenDrummer]

Really enjoyed this

[@maxdonosos]

Hey! This is the first video I've ever seen of your channel, and even though very informative, it left me with a lot of questions regarding metering! As someone that uses Ableton Live to produce, I was left with a lot of questions regarding "what kind of meter does my DAW, have". I wouldn't consider myself a formally trained engineer (or an engineer at all) and a lot of this was very eye opening. Thanks a lot!

[@BR_READDY225]

it has a full scale meter which is very accurate

[@SuperHonkyPodcast]

I've been producing music for 20 years now. Want to know something incredibly embarrassing? I literally just found out about keeping my signal at -18 dB for setting levels and just learned how to gain stage! 20 years folks! I seriously have instrumentals that could grace the billboard top 100 and yet I literally just found out about VU meters, how they work, what our levels should be at before mixing etc. I was always under the impression to "get it as close to 0 dB as you can but never clip!" That's what I was always told! I never had anyone take me under their wing over the years of me learning how to create music. Because of that, I've had to learn everything 100% myself! Only recently have I been relying on stuff like KZbin tutorials to help me out at a faster pace!

[@mrratskins]

I'm totally with you and I'm following the same path. I think it comes from being used to working with analog tape and as a result not being afraid to clip a little for that warm warm sound. But the digital world is different. Now trying to stay at -18. It's working well for me.

[@Tru-Luv-is-2-Giv]

Myself as well. Until, just right before COVID I saw a video on someone explaining on why 118dB is where the plugins sweet spot is. So that's where I been parked since. I believe it was Kenny Gioia?. Man does that guy talk fast, but love his stuff.

[@jefftum4177]

That jazz session appears to me the recorded levels were the ruff mix. I’ve seen guys do this. The play back on the console all the faders would be at 0 and the mix is already 90% there. The advantage to recording on a console is it can be mixed prior to hitting tape. Another advantage of a console is the ability to run hot signals then turn down the fader to lower the level going to tape.

[@themattprofessor]

Interesting perspective. This idea of using a VU for tracking into a digital recorder, is perhaps a red herring, it made sense with tape, and calibrating a tape machine to an analogue console. The simple fact is that digital recording needs headroom so the plugin processing chain does not “clip”, so out of all the recordings here Al Schmidt had the right idea. It doesn’t matter what you are recording jazz, classical or rock, you don’t want to clip any recording channel, in fact most of these meters do not show true peak, which is what you really need, particularly if you are getting anywhere near 0dBFS. Driving preamps is entirely a different thing, so in fact if you want to employ this type of recording method you need preamps with output gain as well as input. -8 to 10 dBFS is a good max figure to aim for, so approx what Al is doing.

[@christianmartinez1]

I enjoyed the video. There a few comments I'd like to make (you asked for it ;)). The first is that the thing about Al Schmitt is that he was pretty judicious about hitting the "red" of the VU meters on the Neve. He's used to analog VU so it makes sense that much of his stuff never got to far past 0 VU or -20 dBFS on a PT digital VU scale. Which brings me to my next comment that -18 dBFS and -20 dBFS are sorta the same thing. -18 is what they reference in the UK as 0 while here in the states the convention is -20. All this is reference from some technical papers from the ATSC. Doesn't really matter which you use but it does explain why certain brands choose either. The last thing is, as much as creatives and sound "artists" would like to make setting levels a subjective or creative thing it really is an objective goal of maximizing signal to noise and avoiding distortion. Not saying you can't drive pre's for saturation but from an engineering perspective it's about getting the highest fidelity transfer from capsule to speaker.

[@jwright8838]

Thanks for the video. It is interesting to see how the different engineers set the input level. Do you plan on going further into gain staging with external preamps and outboard gear?

[@goingto11]

Probably not outboard.

[@CarlCamp]

Great video!!! What would be a good average level for drums, when they’re so transient? And yes to your video on how you gain stage

[@goingto11]

Thanks! I don't worry about an average with drums because you don't really see an average from the quick transients. I like to record drums about 3-6 dB below clipping, but it ultimately depends on the drummer. When I'm working with a session player, I can usually go hotter because they are consistent. When I'm working with Animal, I'm more conservative. When it's a live/remote recording gig, I'm usually a little more conservative as well.

[@caffeinastudio832]

it be good to see the actual wave formes, to get an idea how much compression on there way in was used, because some of those peaks are high but never clipping

[@DaipayanKarmakar]

What works for me is setting a - 12 dB peak during recording... For mixing other projects which I didn't record... I gain stage to - 12 dB... I use the standard meter in Reaper master vol... Nothing fancy... Although I like to use a Vu meter just to get the visual cue of the movement

[@jas_bataille]

That's great but one thing I wanna point out : "digital VU" is just a meter in dBU or dB universal scale where anything above 0 means it's clipping (so there is no concept of "overhead"), but it's not actually a VU meter. That's useful for recording digitally because anything that's digitally clipping means it's going over the amount of data the file can contain. On analog gear, the meters are actually VU meters, which measures the loudness dynamic of a signal, not it's linear signal input. You can calibrate a VU to whatever target you want. I much rather like the traditional VUs with needles because it displays the overhear past the 0 with the black portion. That way you have a good idea how much you can push a signal for saturation's sake. My dad worked in the 70's onward and said that you could hear the needle go "tick-tick-tick" on the his drum bus's VU...! Harisson Mix is great if you want a real old school board emulation, VU, gain and all. Cakewalk ain't bad either, and there are tons of plug-in to check your loudness dynamic, too. It's much nicer to check it on a VU than a dBU meter where you're always squinting to try to see exactly where you at (I hate those things tbh). Cheers!

[@Studio22mix]

I saw there are VU plugins and thought who the heck needs that. I now have it because it was part of a bundle and I use it a lot now, who would have thought that 😃

[@stevegeorge7773]

Yea, seeing your take on gain staging would be cool, great.

[@chinmeysway]

hmm unless it sounds like clipping which sounds not good!

[@stevegeorge7773]

Well!@@chinmeysway that could be a good thing to address in doing a vid on gain staging. To mean what options, alternatives to address clipping if one has no band to redo the bit that clipped. What can be done??

[@citizenworld8094]

This is a great video and even if you were wrong, I think you're right. You are a true credit to music recording, you always give the personal touch and your sincerity in communicating your recording ideas is hard to come by these days no matter if it's like Pure mix or any of those fancy schmancy channels who never fail to try to look slick, but never really cut it imo.. Just thanks man. I've watched a load of your vids but I thought it was time I told you that you are just great. Greetings from Málaga, Spain.

[@goingto11]

Thanks for the feedback! I'm glad you're enjoying the videos.

[@TomCawoski]

I wasn't going to watch this video because there's so many videos out there already on the subject and I didn't necessarily want or need to watch yet another so-called self appointed KZbin expert. What I find interesting is the engineers (and their projects) you chose to give a better idea of what input levels they use. Certainly you can't go wrong with Bob Clearmountain and Joe Chiccarelliand then to add Al Schmitt was the icing on the cake. It's interesting that if you watch Al's DVD "The Art of Recording a Big Band" you'll hear Steve Genewick say that Al rides the big faders while watching the levels on the console when recording a group. The big faders were reversed with the small faders and used to adjust the input gain being recorded. That certainly says something about Al's way of doing things. When setting levels and gain staging I usually use the Klanghelm VUMT meter calibrated to -18 dbfs for most things and anything that is high in transients I will use Waves Dorrough meter. The Klanghelm meter is modeled after a real VU meter and should have the same ballistics. Just as drums are hard to meter through a VU meter is why I use the Waves Dorrough meter for transient stuff like drums or percussion. I too would be interested on your take on gain staging.

[@DarkTrapStudio]

It all boils down to the input, the input of the track is the instrument, the output of the track is the input of the bus, and the input of the master is the output of the bus !

[@chinmeysway]

so true. so obvious. so what

[@DarkTrapStudio]

@@chinmeysway So this is a great idea to do a video focused on Inputs level

[@jkasko7]

Neotek had PPM meters that you could switch to VU. Very cool.

[@jeffwitherell4272]

What I'd like to see is just preamp input levels, how do we know what other front-end gear is resulting in those input levels into protools?

[@Studio22mix]

Thanks for this, most people tend to record too loud something that is not necessary in the digital domain specially when recording in 24bit. I still am not sure about my master-bus mix levels, I read somewhere they should be "-6dB Peak loudness, with an average of -18 LUFS integrated". In my opinion that is a little bit to low, I tend to hit around -15 LUFS but have no idea I'm doing it right because higher levels means lower dynamic range. For mastering some engineers go for around -12 LUFS integrated, with a peak of -8dB. Anyone any suggestions because I'm still not convinced about these levels, they probably also differ by the genre such as pop, metal, acoustic duo's, etc

[@jeepo500]

I would expect LUFS to vary by genre because, as I understand it, LUFS is affected by the frequency spectrum too, so brighter genres with more distortion, lots of cymbals, or synths are genres that I would expect to have higher LUFS to peak ratios.

[@Studio22mix]

@@jeepo500 yeah you’re right. I heard a mastering engineer say that by boosting certain higher frequencies the perceived loudness will be higher without pushing a limiter too much. Especially in EDM and base heavy songs LUFS of -9 to -5 are no exception. -12 LUFS seems to be a good target for streaming services. I think I need to run some reference tracks through an analyzer and a lufs meter and see what’s going on.

[@stevedoesnt]

I wonder if some of those hotter channels have anything to do with how they Ike to hit the output stages of the console channels. My console is set to match digital Zero to negative 20 on the console, and still I have more headroom on the console before clipping.

[@Stu.6]

The red lights mean it’s working !

[@user-zl8fg3lu2o]

Ty

[@anonymousyoutube7259]

If you are recording at 24-bits, that means full scale is 146db and you have over 16 million steps between there and silence. That's an insanely huge amount of resolution. 16-bits gets you 98db and 65k steps - also more than plenty resolution. So a goal should be figuring out where on your meter gets you about 16 to 18-bits of resolution - which is probably a lot further down the meter than you think - each new bit represents a doubling of values.

[@gregorykusiak5424]

Al Schmitt knew to leave room for mixing and mastering.

[@noeqplease]

Remember boys and girls, dBFS = Decibels Full Scale, which is the Digital Scale (-inf to 0, zero being the Maximum, above which there is distortion.) Setting your levels between analog devices and a digital (DAW) converter, you want to see what the converter uses as the level reference (-14 to -20 levels are the most common) and set your ANALOG devices ZERO to that digital reference. So, if your analog console states that their Zero has a headroom above that of = 16 dB, then you set your converter to be at -16 dBFS at the converters input gain, when the analog signal equals its Zero. This matches your Analog devices headroom (the place ABOVE its Zero) to your converters 16 dB above that -16 dBFS. So when you get a red clip in your analog device, that will translate also to a digital clip. Cheerio and Merry Christmas.

[@Notinserviceij]

What if you want analog clipping but not digital?

[@TylerDarlington]

Do you have methods, or even bother, monitoring digital VU on live consoles and DAWs that don’t support digital VU natively on the meter bridge?

[@jmoemorris4133]

I don’t understand why so many are so harsh about someone bringing some observations to the table,if you can make such harsh comments that are results I think you would be famous but I have never heard of any of you!just experience the notice this guy is giving and you can take it or leave it but you definitely don't know it all!

[@mSarimaa]

Al’s stuff peaks are a lot hotter than the average, means a lot less compressed going in, which means it is more important to play it a bit safe and leave some headroom.

[@georgecain123]

Question: did they record those tracks to analog tape and then transfer to digital? Or were they recorded digital?

[@goingto11]

I'm not sure about the Clearmountain stuff, but I'd guess it was all digital since Bob has spoken before about his distaste for analog tape. I'm not sure about Joe's session. The Al Schmitt one was definitely digital.

[@taraszakordonskk5731]

What about when we recorded on to tape and wanted to hit that tape hard. We wanted that level almost at red or overload all the time. Not clipping or distorting, but rite there. A pinch of distortion or clipping actually helped a bit. Such as snare.

[@goingto11]

Sure, but most people aren’t tracking on tape these days.

[@johnphillips752]

"Gain Structure"... Are we talking Channel Gain or Channel Fader? There's usually a difference, but if not, then gain input is controlled by the output of the source. Gain is corrective, but fader levels are creative. Of course, "gain" is pre and "fader" is post in this description. If that's the scenario, then it is highly recommended to structure gain settings in a way that allows the faders to be equally referenced. If you've done this, then you've set up a level playing field to subjectively create. (This is all assuming that you have left enough headroom with no clipping in the first place). Disagree? Let me know.

[@tekis0]

I've been recording, professionally, for over 30 years. I tend to mix the work of others more and more these days. I find that 90% of what I get, which has been recorded by someone else, is far TOO HOT. In the "Clearmountain Session" notice where the faders are sitting; it's right around UNITY GAIN. On a console that's where your hand tends to sit. Also, the fader is most accurate around there. "Yellow is the new red."

[@Tru-Luv-is-2-Giv]

Newbie here. I like that Yellow is a new red. From what I gather from other videos is -18 is where plugins find it the sweet spot. So the yellow would I would assume would be playing it safe I would imagine. This was eye opening for me.

[@MrMrMrMrT]

32 bit float?

[@kempdaw6726]

G R E A T ! ! !

[@ComplexConfiguration]

Seems like there would be clipping on the master. Are people in general against turning up the gain on a limiter? just have it hit the master bus hard and clip and slap a limiter on it? and make sure youre not losing transients I guess? I find it a bit odd, was expecting something else, i dont see how this leave a lot of room for parallel compression, reverbs and delays, whatever else could be used.. Huh, would have loved more info about how the projects handled that. But was interesting anyways. THanks.

[@MarkKunoff]

You left out the WHY or the benefit of setting input levels this way. It’s for preserving headroom for eventual mix down, but that was way more crucial back in the analog tape era.

[@kilswitchengaged]

Took me a while to understand you were saying neg (number). I thought you were saying neck 20 or neck 10. Saying negative would have been more clear.

[@goingto11]

That's good to know for the future.

[@ernieg]

For minute i thought this was Bill Gates Talking I was thinking he was into engineering ? LOL Good Info That i did Not Know Thanks !!!

[@PaulPaul-vj2vx]

Only matters with analog gear, with 32 bit around the corner and already happening it doesnt matter at all …. Total flexibility in post

[@goingto11]

As far as I'm aware, AD converters are still fixed point, and thus clippable. Sound Devices is doing some interesting stuff with this, though...

[@PaulPaul-vj2vx]

not sure, but I think before the AD-converter will clip, the mics will do so long before... at least thats what they tell us in the ad.. .@@goingto11

[@BLKLTRSH]

Read "On the Record" and Thank Al after.

[@goingto11]

Great book.

[@argumentchannel]

PPL talt about dB. Arbitrary numbers when in digital wee should be talking floating point 1+ 1- and bit depth. Say my meters go up to 11

[@bassmandan12]

Why do people have such an issue with clipping? Minus Al, both those other engineers will clip. I’ve been there and seen it. Especially sources like drums. It adds edge and punch if done correctly. Not understanding its limitations and when it sounds awful is most people misconception.

[@notlayjeno6258]

-6db

[@Dubsteppah]

I'm a musician first, not an engineer, so I let my engineer worry about gain staging etc, I just focus on making good tunes, and then get the engineer help. Cheers!!

[@anotherchrisperry]

Nothing wrong with a little clipping on the way in these days. ❤

[@bluesky2145]

Yeah these were probably recorded on a reel to reel

[@TheRealCalijokes01]

I keep getting too much noise

[@uncleted3961]

“Neg”

[@tracyeaves4847]

Analog, push faders, Digital conservative

[@synthzizer3324]

There is a sweet spot for levels. Just dont clip. Id rather gain down than up. Always find loudest part of playing and ill even record in at -3 no problems because that could have been a sweet spot of the front end chain when it hit digital. Dont over think it. Worry more about crest factor and thickness of waveform is more important IMO. Rec Neg 20 to pesk is BS. That means gaining down your analog. So BS. I can understand if you want to gain down. Then what mr levels?? Bubble the whole mix up on the master bus from neg 20???.. thats alot of heavy work on the master. You need to understand the cumulative effect of levels when they all converge on the master. Do your own testing combine multiple waveforms say a dozen at -20 and what do you get level wise on the master??????

[@baseboardmatt]

You should probably just go ahead and not use an abbreviation for “negative”.

[@roberthahn5235]

ballistic wise those levels are 20 feet away how about a Studer at plus 9 no wonder music sounds like shit these days id plug that nonsense into my mic inputs not line returns

[@matt926uk1]

You remind me of Rivers Cuomo

[@JacobHeldt]

remember to set all faders to zero - otherwise your Recording levels wont show correctly !

[@kylegood9875]

This was useless

[@cassetteo]

I went to audio school and we never learned about “gain staging” it was something that was so obvious and so standard we never even learned about it. It’s pretty easy to understand, don’t overload the input…. This may be something that fruity loops morons don’t understand, but any common sense will dictate this.

[@whitex4652]

I help you out just quickly: DO NOT CLIP! Done. Everything else is discussing non existent BS. And who cares what celebrity does when,?

[@PippPriss]

I'd not go as far to say "don't clip" and be done. Julian Krause has nice graphs in the distortion levels based on the pots position. If you'd smack those inputs in with super high gain, then reducing within the DAW, you still get non static elements within recording. But as soon as you have your signals in the DAW, as long as you don't clip the output, you're golden, and on the single channels clipping does not matter. But there are plugins which are tailored to accept different signal levels and react to them as if they are saturating etc.

[@gener2842]

Well, sunshine, you should pay attention to what highly experienced professionals do because if you cut tracks too hot, when you insert channel strips or other plugins, you will have a lot of problems with gain staging. Or you can be smug and ignore this, resulting in a blob of crap as a record.

[@martdod]

Oh look, a know-it-all. 😂

[@-303-]

It’s an awfully big assumption that digital audio processors (VSTs) react linearly. It’s a completely wrong assumption for digital emulations of analog devices. Input level affects the sound. You don’t wear a seatbelt because you expect to be in a crash. You do it in case you are in a crash. Likewise, you don’t set input levels to -18 dBFS because you expect to use VSTs with non-linear responses. You do it in case you add them later.

[@mirkomarkovic3438]

​@@-303-analog emulations sound like crap anyways, why bother