Headroom is the "distance" between the loudest signal going through your device, and the loudest your device can go without distortion. More headroom is good, because it avoids distortion. EDIT: But it's a battle, because on the opposite end, you have the "noise floor" of the device, which you are always trying to be as much louder than as possible to minimize background hiss. So you can't just turn everything way way down unfortunately.

You can get it by proper gainstaging, which is the volume/gain settings on every device, you want to turn it up loud enough to get something out of the thing and be a lot louder than the background noise, but not so loud that it distorts itself, or distorts the next thing in the chain. It's not really something I can explain easily how to do, it involves trying to maintain unity gain (no change up or down) for most of the chain, but sometimes you have to deviate from that - it depends a lot on your source (the vocalist) too, so we can't just tell you settings to use.

I would google or search this site for "gain staging" tutorials or tips, and read everything you can, then practice with your gear making sure it doesn't distort. If it does distort, then you'll have the fun (and educational thanks tothere being multiple things in your chain) task of trying to figure out which device (or devices) is/are clipping, and whether they are clipping because of their own settings or the settings of the device previous to it/them in the chain.

Fun times! Read what you can, then ask whatever you cannot figure out.

Noise floor is discussed ad nauseum lots of places from GS to RO to HomeRec. Noise floor IS an important consideration but with modern digital gear and modern QC on common acceptably decent components, it is not really the defining factor in gain staging. 15 years ago this was a different story. Even today's newer live PA rigs have so much less inherent noise that gain staging is easier to deal with. In the chain listed by the OP for instance, there is nothing there to automatically "add" noise. The DBX could ruin things like any compressor, and the MBox2 has deplorable headroom due to it's analog wiring but nothing that can't be accommodated by reasonable levels.

You do need to go gentle into that good night regarding the MBox inputs. They don't have salty chocolate *****. To mix my Dylan Thomas, Southpark and Monty Python metaphors even further, they are Wafer Thin.

Basically to maximize the headroom you have to make sure that the output of the Presonus TubePre is not slamming the MBox. Your peaks in ProTools should show up as a range between -20dB to -12dB and definitely no louder than -6dB. This is a plenty hot signal to work with. Remember that make up gain on the DBX will exacerbate the problem. Leave the DBX output at 0 gain or less as you can always normalize the track hotter in ProTools. Better yet, until you understand what the MBox can handle well, remove the DBX from the chain.

Now, if you find that to make the mic sound good your need to push the preamp harder (unlikely to work on the lower end tube preamps but YMMV) the output of the preamp may be too hot for the MBox. In that case you will need to engage the 10dB pad on the TubePre or use a line level attenuator right as the signal comes out of the TubePre. Again there is no advantage in pushing the TubePre at all since it is a starved plate design and not full voltage.

There is more that could be said but this is my thoughts on your specific chain.

Here is a good example of what headroom is really about. Most old-school professional audio consoles typically could run a low impedance load of 600 ohms to +24 to +30 DBm output levels. While most low-cost consumer oriented recording gear CAN'T drive a low impedance 600 ohm load (which requires plenty of current drive). That in comparison to the "Prosumer" gear which can't drive any load lower than 10,000 ohms without clipping & output levels that barely make it to +18 DBu. Which of course is substantially less than +24 & +30. So not only does real pro equipment practically have the POWER to almost drive loudspeakers, they also have extra voltage capabilities that provide for +6 to +10 DB additional output level. This is the true separation between the real deal & the toys. And it's that headroom that provides for the extra snap & punch that denotes a professional sound. So is it not possible to get that additional headroom sound without spending a boatload more money on esoteric gear? You bet your ass. I've been forced (not exactly forced) on numerous occasions to have to record some rock bands on crappy old broken Peavey & TASCAM consoles with no decent headroom. Everybody marvels of the quality professional sound I get out of these crappy consoles these bands can never make a good recording with. This is because, of a deep understanding of gain staging, headroom & signal to noise. You see, I'll run these boards in a way to buy me 10 to 15 DB more headroom. I do this by trimming the microphone preamp gain staging down by 10 to 15 DB lower than normal. The output of the microphone preamp then passes to the consoles bus output amplifiers which I'll drive 10 to 15 DB higher than normal. This will improve my headroom by 10 to 15 DB while increasing noise by 10 to 15 DB. This really doesn't equate to any problem especially with pop orientated music recording. It would however be an obvious problem while trying to record a quiet oboe solo from 30 feet away. And whatever additional noise can actually be reduced artfully with downward expansion & carefully tweaked threshold levels or, with the software equivalents. So I tweaked the board completely wrong to make it sound good because I knew what was inside it and I knew what it could and couldn't do. So I always get a professional sound out of toy/nonprofessional equipment that still retains a great amount of transients and snap with plenty of punch & bounce. The basic problem is that most consumer-oriented equipment is anemic at best in its performance capabilities. Electronics in most budget oriented equipment can get quickly saturated resulting in a flat lifeless sound. The analogy here is if you are towing a 10,000 pound trailer getting up hills is far more effortless utilizing a Chevy pickup truck, with a 5.7 L V-8 engine, than towing the same 10,000 pound trailer with a 2 L, 4 cylinder, Toyota Corolla. That's headroom baby! Of course, if you're crafty, you could drop a V-8 engine into that Corolla. But that's not practical. It's easier changing levels.

I'm basically telling you how to run the console (and/or even standalone preamps with more than a single gain stage) incorrectly (I love to contradict myself) to get what you want from it. Never mind how the electrical engineers designed it to work. That's theory. I'm involved with practice. I can't say how many times I've done things wrong to make things right. It's crazy. It's even Bozo like. It's even worse when you know how to do it right but the equipment isn't up to the task. Heck, nobody needs more than +18 DBu worth of internal headroom. WRONG, PROFIT MONGERS! When I look at specifications for equipment, I always first look for its output capabilities. I don't care about response, that's flat. I don't care about distortion specifications, they're all lower than any loudspeaker. Although I am somewhat interested in its rated signal-to-noise ratio. Which is generally more than adequate in all modern-day equipment.

Of course there are instances when we contradict our desire for headroom. Those instances are with certain pieces of equipment that when run extremely hot in gain and shoved balls to the wall as with early API & Neve equipment will change its character of sound when running on its raggedy edge of saturation that creates a unique character of sound & coloration. But that scenario generally isn't applicable to IC chip styles of component design. They just get crunchy bad sounding, real fast. But remember, even when I'm gaining up too high, my output capabilities are still close to +30 DB even if I am over gaining up on the input side. Understanding what's going on here helps by also understanding that I'm over driving the input of a class A input amplifier and not over driving the output of the class AB output circuit of the input amplifier. This actually allows me to tailor sound on my analog console without even playing with any EQ.

Unfortunately, as John also indicated, most class A cooperation amplifiers have largely been associated with tube equipment. And as we all know when you push a tube, which is class A cooperation amplifier, it is always biased at 100% on. And when it starts to go nonlinear, it begins to distort the waveform asymmetrically. This asymmetrical distortion component produces musical sounding second harmonic/even order distortion. That occurs in real life and is musical sounding. But when things get symmetrically clipped, that's not something that happens in real life. It's an electronic phenomena and is harshly dissonant sounding. Transistor circuits can also be designed to be class a circuits. This doesn't mean that they soft clip like tubes but they certainly produce even order harmonic distortion components that make those circuit designs to be more musical sounding. Unfortunately there are tube circuits known as starved plate. This actually means that the tube is not part of the amplification circuit. It's not amplifying anything! It's merely their to be in constant overload distortion mode. Not only is the plate in the tube not running at full voltage, some have heater circuit voltages so low that the tube never even glows! Oh sure it looks like it's lit up but is just a yellow LED behind the tube. Nobody needs those kinds of bogus tube preamps. Because it's not a tube preamp unless the tube is part of the amplification circuit. And it's not going to be part of the amplification circuit if it's plate voltage isn't at least between 250 & 300 V DC.

I love lots N' lots of headroom
Mx. Remy Ann David

Eloquent as usual. Apparently we went to the same school of gain-staging.

Dave, I would certainly love to meet you face to face some day as I always love your responses. Having a great picture of you says it all so I love your avatar. The look of a competent professional having a great time which is why we do this. And we're not done doing it.

We have to stop meeting like this... Dave. (Sob sob sob)
Mx. (sniff) Remy Ann David

Hi Remy,

Being a highly visual person I am trying to noodle through this. If I understand this correctly with "Prosumer" gear you would want to have the mic preamp, if you had one in the chain, coming into the board 10 to 15 DB lower, as you said. On the channel strip or main output meters instead of the signal being at or near 0 it would be 10 to 20 DB less?? Then you would 0 that channel and trim above 0 by 10 to 20 db?? This giving the extra headroom in the overall out put stage as you wouldn't need to push that channel up to get the level you want thus lowering the noise ratio. I hope I got this. By the way the truck car example was great.

Thanks,

Karl

You have to keep the trim rotary on the top of the mixer stick lower. What you feed to it has to make sure that it does not max out the voltage of that input (generally lower on prosumer gear) therefore your external preamp's output will likely be lower than you think "optimal" too. You make up any missing game with the fader at the bottom of the mixer.

In the musical My Fair Lady I think the after she learned how to say the rain in Spain falls mainly on the plane, followed by her teachers proclaiming " I think she's got it". Yes. But here's the important part to understand: when using an audio console, all that is, generally, is a series of similar small "Operational Amplifiers", all strung together. But within this series chain of amplifiers that create gain, there are also losses in level from passing through various pieces of other circuitry. Because of the losses, the next amplifier stage has to then elevate the gain some more to be able to then pass it on to the next stage in the chain. Many console manufacturers would also include, in their manuals, a graphic, which was a representation of each stages amplification and each stages of loss from input to output. It was the example of how the designers felt best in the optimization of the console's properties. In fact you'll find that graphic gain staging example of a lot of Yamaha equipment. Now since somebody else designed it, it's generally their goal in this gain staging routine to optimize each operational amplifiers levels & signal-to-noise ratios. The analogy here might be when riding a multi-geared bicycle, you would go to the larger gears which would make you pump faster & easier by sacrificing speed for torque to go uphill, then switching to the smaller gears to sacrifice torque for speed. One can somewhat equate gain staging to the gears. So the whole concept is " geared" toward actually optimizing yourself as opposed to the bicycle. Each amplifier has an optimum specification which you have the ability to finagle to suit your needs, regardless of how the designers felt it best to be used. So, it can also be looked at as robbing Peter to pay Paul. However, if you are utilizing an à la carte method of creating a " channel strip", there may be extra variables one will have to consider. Such as a standalone microphone preamp. A standalone microphone preamp will not only have your internal operational amplifier " gain setting", it may also contain an output volume control. This additional output control should be generally set to full bore to the "0" marked optimum output setting. Then you might be plugging it in to say, a limiter. Those generally also have input or threshold sensitivity controls along with an output volume control to make up for the " dynamic gain reduction" that limiters do. So if it's automatically turning things down dynamically, you may need to make up that gain at the output. What's difficult to understand here is one must have a fairly solid understanding of what the optimized range is for each stage of amplification. With this knowledge in hand, and in brain, you can then (given that you know there will be transients on instruments like drums that you can't fully see even with a peak meter) that may exceed the microphone preamps optimum output capability. So, if you know it's going to start to get ratty sounding, you won't want to go there but instead, reduce the internal gain of the microphone preamp to make damn sure there is going to be enough ooooomph to get those transients out before that little Operational Amplifier starts to go nonlinear. Of course, since you don't have the preamp "balls to the wall" while it's sitting back relaxing instead of jumping around, the next guy may have to do some extra duty that it wouldn't normally have to do. You see with really good sounding prosumer gear don't quite have the guts but have the glory to deliver. You just have to decide how much guts to how much glory YOU want out of your fine purchases. This is particularly equipment that utilizes IC chips a lot. And IC chips while they are fabulous, do not quite necessarily function like that of a discrete transistor design of the higher-end and/or old-school style equipment. And even when talking about prosumer gear, it may in fact be almost identical to classic consoles such as the Auditronics, Neve's. My Neve is all transistor but within a few years of the release of mine, Neve went to utilizing many of the same little integrated circuit chips, identical to those you would find in a very cheap PA board! No joke, this Signetics 5534 chip. So why did the $85,000 Neve sound so much better with the same $3.50 chips in it? Because they just stuck a couple of extra output transistors onto their circuit boards so as to keep the chip from crapping out so quickly. And it was those output transistors that made the difference between an output rating on the PA board of capabilities to +18 DB in comparison to the Neve, with its extra output transistors which provided the ooooomph to provide the extra headroom to damn close to +30. So those output transistors were providing 12 DB of additional headroom that the 5534 chip couldn't deliver. Of course there is a lot of valid folklore that discrete transistor circuitry sounds different from IC chip circuitry. When Allen Sides, owner of Ocean Way and other million-dollar studio complexes was asked about one of his Neve consoles, he indicated that " it's the best sounding IC chip Neve ever made". But those, in spite of their lowly quality, can't necessarily be pushed or manipulated as much in its gain structure as the older transistorized Neves. And so even if you own a top-of-the-line piece, you still have to understand the way its internal character delivers. So the way you make your lesser expensive compromise is to create the headroom & sacrifice a little more noise for a more PRO sounding product. And noise today can be so much more effectively dealt with in our computer environments that you actually may not have to really sacrifice anything. So this is all good. And while I frequently tell people not to shop for equipment by looking at the specifications, it's understanding the specifications that is interpretable differently by folks in the know. And now, you are ready to bring to us some fabulous sounding stuff, that can rival the finest sounding Neve consoles since you now know how not to over abuse little IC chips.

Come here Chip! (BARK!) Gooboy. Now down! Gooboy, here's a biscuit. And that's gain staging.
Mx. Remy Ann David

BTW

You can tell that I failed in school at English grammar. Holy moly that's the biggest paragraph I think I've ever written?

At least this post is short and sweet like me.
Mx. Remy Ann David

Hi Remy and Jack,

Thank you Remy for the awesome detailed explanations. It clarified my understanding and added knowledge to the electronic side of things.

Thank you Jack for your input and your clarification. It helped as well.

Now I need to get my behind in gear and record/mix.

Karl