View source for Music - studio and stage notes - devices you'll use

{{#addbodyclass:tag_tech}}
{{#addbodyclass:tag_media}}
{{avnotes}}
=Mixer notes=

<!--

Remember that most mixer faders don't ''mute'' a channel, they just turn it down 60dB or so,
and that's close enough. 
(In fact, mute buttons may do the same)

But I have previously noticed that, with two decent mixers chained, you can amplify that back to be audible. 
It's about as noisy as you'ld expect, but audible.


-->
===Channel plugs===
<!--

Most mixers have two inputs:
* '''Mic in''' (XLR)
: pro mic level, goes to a preamp that is often gaining at least 10dB

* '''line in''' (TRS)
: typically balanced - but plugging in unbalanced makes it unbalanced (it shorts cold to ground)
: either
:: pro line level, doesn't need any gain
:: instrument level, which might needs a bit of gain (instrument here meaning guitar pickup)
: so it's often still sent to a preamp, you just might leave it at 0dB


A fancier mixer may add
* '''insert''' (TRS)
: unbalanced line level
: send+return -- that by default is normaled to not be in the path {{comment|(basically a short, see [[#Insert|details below]])}}

* '''direct out''', sometimes 'recording out'
: copy of the channel, meant to allow individual many-track recording
: might be provided via a breakout rather than another row of XLR
: Varies per mixer if it's between preamp and gain, gain and EQ, EQ and fader, or even after the fader.
: (note that inserts can be abused as a direct out)

-->

====Insert====

<!--
'''Insert''' is basically short for "inserting an effect in the middle of the channel's signal path"

It's useful to e.g. put in an external compressor on individual channels. Or perhaps reverbs or wilder effects.


Since you'ld be using line level effects, this plug sits 
: after the mic/line in - after any mic preamp so that it is useful on XLR ins as well as the TRS in
: before basically any other part of that channel  (and before the gain/trim section ''if'' that is separate from the preamp{{verify}}).


When introducing these it can help to explain them as effectively the same
* as an unbalanced TS send 
* unbalanced TS return
...shoved in a single (TRS) socket. 

In fact, if you want to put a guitar pedal into an insert, that is exactly the adapter you need to buy/make.

For reference: tip is send, ring is return, sleeve is ground



Notes:
* It's in a single TRS socket mostly for practical reasons.
: Partly that it takes less space (and components), for something you don't often use.

* putting it into a normaled TRS socket meant it needed no ''further'' components,
: because the way this is switched is purely mechanical {{comment|(without a plug in there, send and return are shorted.
: When a plug is inserted, that connection is broken and goes through your wire instead.
: This was a really simple and cheap ''and clean'' way to add this feature)}}

* It's usually fine that this is unbalanced, because the effect will usually be closeby {{comment|(though there is still some risk of ground loop)}}.


Because of this physical design, plugging in a TS plug does something completely different - which may be useful:
* inserting a TS plug fully gives you an output on the wire, and silences the mixer channel
: because consider that the insert's tip is send, ring is return, sleeve is ground
 T R S
 T S-S
: so it takes the send on the tip - and shorts return to ground

* inserting a TS plugs halfway (to the first click) gives you a mono unbalanced output without silencing it
: ...because that first click is the spring that ''would'' be pushed away from shorting the socket - but isn't. send and return are still connected and your tip is connected to that
: can be a bit loose, though, so this is a bit of a hack
-->

===Gain and trim===

<!--

Normalizes the channel levels to something reasonable compared to the others.

You could do without, but without you may e.g. find that a keyboard is too loud with the fader at 30% and a mic still to quiet at 100%. Gain and trim lets you use faders more sensibly (and helps avoid some distortion).


You may have single gain/trim knob, which can then typically gain a bunch (for mics) and trim a little (e.g. for line-level instruments).

Sometimes you may have a separate knob called trim or level, for subtler adjustment, like -10 to +10dB. It's sort of redundant, but its precision can be convenient. 

You often use gain mostly on mic inputs, as there's variation between mics.
Things that come in on line inputs are, in theory, all relatively similar. 


If you understood all that, you can figure out why and how other mixers might do things slightly differently.


Side note: on digital consoles, trim typically applies once it's digital{{verify}}, so it can make sense to set it lower to keep some headroom.
-->



===Aux===
<!--

'''Aux send''' let you send one or more channels to a speficic output called aux.

It taps off the channel before the fader, and into a specific aux channel -- fancier mixers will have multiple aux channels


They have a few uses
* send one or more channels into an external (pedal or rack) effects
:: if it's a more creative effect, you might feed that into a new channel
:: if it's subtler (e.g. reverb), you might put it into aux return, which mostly just goes to the main bus
:: (Note also that different effects work differently at differnt levels, so the amount of send (and gain/trim being roughly correct) can rather matter)


And, if your mixer does not have [[subgroups]]:
* use them as sends towards stage monitors 
:: to present the unchanged levels to your artists, while your channel faders control what your audience hears.
:: "Foldback" on a mixer is a similar idea - foldback seems to be an older term for what we now call monitors.

* if you have multiple aux channels, another send-only use would be to output to a multitrack recorder to figure out further mixing later
:: and not having to undo your fader tweaking in ''that'' process


{{zzz|differences with subgroups:
: aux is often a single mono bus; subgroups are one or more stereo buses
: aux is usually pre-fader, while mixers fancy enough to have subgroups often also let you switch ''those'' between 'pre' and 'post' - whether we send the channel just before or just after the channel's volume fader - makes it more flexible in what it is convenient for

Post means the channel fader affects what gets sent. 
Since your main out is also send out level, 
this is probably mainly useful for non-returning uses
where you want 
}}

-->

===Per-channel buttons===

====Solo, PFL, AFL, SIP====
<!--
{{stub}}

These are all flavours of solo - "I want to listen to just this, please".



"Is it the same as 'mute all others?'  Can you think of them as tapping off at different places within a channel's knobs/fader, but otherwise the same idea?"

...mostly.

The buttons mostly represent variations of
* where within a channel we tap off
* what else to do, especially when multiple are pressed

And mixers may individually vary a little on where exactly it goes
* often it means "I want just this channel on ''everything'' - the VU, monitors, and main mix".
* sometimes it means "I want to tap just this thing into my ''monitors/headphones'', and leave the main mix alone"
* exactly what happens with extra buses (if any) ''may'' be another footnote.


In order:

'''PFL''' ('''Pre Fader Listen''')
: taps after input gain, but ''before'' channel's dynamics/filter, fader, and pan
: seems intended to quickly set a reasonable input gain on all channels, also so that you can use most of the fader range without worry
:: (go through each channel, asking for people to play their instrument the loudest, and set a reasonable gain).
:: note that ''because'' it ignores the fader, you can listen to that channel even if the fader is all the way down (and the volume at which you listen better make sense -- which is also why in a soundcheck you maaaay want to ensure the gain is turned down before you press PFL)
: easier than pressing the mute button on all other channels, putting the fader on 0dB, disabling any bus routing, and centering the EQ.


'''AFL''' ('''After Fader Listen''') {{comment|(sometimes post fader listen, but that doesn't acronym very usefully)}}
: taps after fader, before panning
: e.g. if you think there's one instrument that is not sounding right in your mix, AFL lets you quickly swap between just it, and the whole mix, so that you can e.g. EQ it more appropriately.
: less necessary, great for more precise mixing, arguably more of a studio than stage thing


'''Solo In Place'''
: taps after everything (post-pan)
:: which also makes it a stereo signal (where PFL and AFL are functionally mono)



Notes:
* You never need all of these. Mixers tend 

* some of these have an analog-mixer background

* Solo In Place (SIP) is theoretically a clearer term than Solo
:: in practice people are sometimes fuzzy about Solo / SIP / AFL

* In analog mixers, AFL is mono. Digital mixers can be more flexible
:: {{comment|(in analog, a channel only becomes stereo after the pan pot, which while often placed with the rest of the knobs, is ''wired'' after the fader. As such, stereo AFL would require a stereo pot and some redesign)}}

* DAWs are digital mixers. 
:: They can make it more flexible, but tend not to (at least by default) to keep things understandable
:: they do add concepts like 'solo safe' - channels that ''won't'' be muted by use of Solo/SIP elsewhere

* in DAWs, it is easier to add effects post-fader
:: contrast this with physical consoles, where 
::: it is possible to have them before input
::: it is possible to use insert (which is post-gain but before anything else)
::: ...all of this is pre-fader
:: which makes AFL and PFL less useful


AFL and PFL are mainly for live/stage

-->

====Subgroups, or, 'buttons that say 1-2, 3-4, etc.'====

<!--

Indicates that this mixer has multiple extra '''subgroups''', which are extra buses -- extra channels, just inside the mixer, that you can use.


This is just routing, and this isn't even the only extra routing around.
Say, aux/return, mainly have the goal of selectively adding effects.

Subgroups have a different goal, and the knobs aim at different behaviour because of it.
Selecting a subgroup switch usually means '''instead of putting this channel on the main mix, put it into this specific subgroup bus instead''.

Each subgroup bus has its own fader into the main mix. You do not have to use it.

A common use is to, say, send all mics on a drumkit to one "drums overall" group-of-sorts.
If you need the drums louder, you no longer have to touch the relative adjustment of each of its mic (on each channel's faders),
and have one "drums overall" fader you can use instead.

Maybe you group all the vocal mics too. 
This means it's now a lot easier to adjust the difference between drums, and vocals, and the rest,
without messing up the tweaking you did for each channel.

Yes, there might be a near- equivalent in some uncomfortable hand-claw move, that leaves things ''more or less'' right.
But it's nice not to have to.



'''When mixers also have physical outputs for each subgroup''', there are some extra uses.
Consider:
* recording each group separately on a multi-track recorder
:: note that in practice, if you have multiple aux sends, they may be preferable to this so you don't record your fader fiddling (because aux is before the fader, and subgroups after)

* at a venue, sending fewer than all channels to the bar 'outside'  (to give an indication when to head back in)
: somewhat awkward, but doable

* send vocals to a vocal monitor, drums to a drum monitor, etc.
:: only really relevant on larger stages
:: and arguably ''not'' so useful if you want to inform artists of how they sound in the mix


Notes:
* Since these buses appear mostly on larger, more complex mixers, they're usually set up to allow stereo -- which is why they are set up in pairs, labeled something like 1-2, 3-4, etc.
:: {{comment|(If they were mono, you will need to use two of them at a time with some hard panning per channel, if you can. Perfectly doable when necessary, but nice that you often don't have to)}}




-->


=Stage mostly=


==DI==

{{stub}}
{{name|DI box}}, {{name|DI}}, {{name|DI unit}}, {{name|Direct box}}. {{comment|(some people argue over whether it stands for direct input, direct injection, direct induction, or direct interface. Most people don't care, because DI names a specific thing and function that we recognize, and isn't that the main point of a name?)}}.



'''Functionally'''

Takes a high-impedance, unbalanced signal, typically on 6.35mm TS.

Outputs a low-impedance pro-mic-level balanced signal on XLR
{{comment|([[pro mic level]], because that is roughly the only thing XLR carries in audio context)}}.

DIs were arguably ''made'' for guitars, but have other uses as well.



'''Passive DIs''' do so with a single audio transformer, so need no power,
which makes them a no-worry functional addition - plug in and forget.

...but their simplicity comes with assumptions that are only usually true,
and sometimes they lead to undue sound coloring, or higher noise levels.


In those cases, you want '''active DIs''' to fix that. 
They do a little more, but they do so using power (battery, adapter, and/or [[phantom power]])
so take a little more thought than just plugging in.


<!--
"What can the 6.35mm TS be?"

DIs were arguably ''made'' for guitars, but have other uses as well.


Passive DIs are little more than a roughly-12:1{{verify}} [[audio transformer]], 
which are aimed mostly at the voltage of electric guitar pickups.

Voltage aside, electric guitar pickups (passive, magnetic) are high-output-impedance,
and electrically ''cannot'' carry far without picking up a lot of noise, and/or a lot of coloring.

Also, if you want to put it on the main speakers,
then chances are you need to feed it into the stage snake, which will likely take only XLR.

There are roughly three ways to deal with that:
* feed it into a guitar pedal. It's right where you are ''anyway'', and 
:: many of them will ''need'' to do something that already happens deals with that 'high impedance output can't carry far'
:: doesn't fix the XLR part 

* feed it into a guitar cab, and put a mic on that. 
:: That mic is just a mic like any other, nothing special to you ''or'' the sound person.
:: fixes the XLR part too

* if neither of those, then you probably want to go more directly to the mixer
:: but have solved neither 'needs to become low impedance to carry further' or 'needs to become XLR'
:: a DI does both of those things


There are more flavours, like "guitar cab has pass-through TS that you can plug into DI to have two versions of that guitar" but 



When your instrument is a (low-output-impedance, higher voltage), 
then you don't need it for voltage or impedance reasons (a DI would still crank up, and you would need to turn it down a bunch),
and the main reason is because [[stage snakes]] going to the mixer tend to take XLR only.


-->
<!--

"If they are so simple, why are passive DI boxes sort of pricy? Can't I make one cheaper"

[[Audio transformers]] aren't that simple to make.

I'm not sure why economy of scale hasn't made them cheaper than they are



"Why not put a DI inside a guitar and give it an XLR output?"

This is in fact a ''great'' idea to avoid noise, because guitar pickups being high output impedance is basically just bad design.
That has existed, for a very long time.

There are a few reasons.
* this gives a very clean guitar output -- byt that is not something we care in a lot of musical styles.
:: you could give yourself the choice of either clean, or adding a lowpass and noise, but at this point, you're making your life overly complex

* there are a lot of practical ones, mostly results of conventions
:: say, no guitar amps take this, and we really like using guitar amps
::: Yes you could adapt it, but that's effectively two adapters for a few meters of cable


"Why not put a DI inside a guitar alongside the classical output?"

Cost.

And again, there are a few styles of playing that would like this,
but most would not care at all.



-->


<!--
The most basic DI is just an TS input, XLR output, that ought to pass signal as-is.

Passive DIs are little more than a roughly-12:1{{verify}} [[audio transformer]]. 


Optional features on a DI:
* a second TS socket, marked link or thru, wired directly to the first.
: lets you use the DI as a simple splitter on that end, e.g. so you can send both to the mixer, and maybe a cab, or stage monitor, or something else stage-side

* a pad switch, which means weakening overly-strong signals (e.g. active instruments, some keyboards) by 20dB or so

* basic frequency filtering (coloring mostly)

* [[ground lift]] switch
: (TODO: explain)

* RCA input (to connect laptops or whatnot, while avoiding ground issues)

* XLR input, in case you somehow have a ground-referenced XLR cable you need to avoid using the ground of. {{verify}}





-->

<!--

Bonus implications / other reasons DIs are used:
* XLR's lower impedance makes it easier to do a run a longer cable (or same distance with less noise) than you could with the cable type you typically connect on the DI's input side
:: only partially true for passive DIs because all of the power comes from the instrument

* DIs make it easy to avoid connecting the TS-side ground
: which can help avoid ground loops you ''might'' have when you e.g. plug unbalanced TS directly into a mixer.
-->






====Passive versus active DI====

<!--
'''Passive DIs''' are little more than a roughly-12:1{{verify}} [[audio transformer]] (...with a bunch of footnotes).

This implies an impedance ratio of 144:1, so if the target is a 1.2kOhm mic input, the sweet spot for the input is a ~170kOhm source.

They also step down the signal voltage, which is why passive DIs are fine to use on things with stronger outputs - mostly meaning powered instruments (e.g. keyboards), in which case you're using the DI largely to cover distance over the slightly-handier XLR cable.

Passive instruments, in particular unamplified pickups, often have weak signals and high impedance, so and adding a passive DI (remember, mostly just a transformer) to the pickup electronics it a more complex and hard-to-precisely-describe analog circuit.

It ''is'' going to color the sound, because the whole thing is one circuit with interesting interactions.

Sometimes in ways you like.
Sometimes in ways you don't like, and won't understand or be able to control -- in which case you want and active DI.


'''Active DIs''' add a buffer and amplifier to the design, 
and it is the power you supply to the DI, not the signal, 
that determines the strength of the signal it delivers.
(They don't strictly ''need'' transformers at all, but it's still a sensible design choice)


These are primarily targeted at guitars, and may add some guitar-specific features (like two 6.35mm plugs that are actually the same thing, so you can send the guitar to your cam ''and'' to the mixer, without needing extra hardware).



'''Roughly'' 
: you may need active DI for passive instruments (pickups), 
: passive DIs will work mainly for active instruments.
: if you want just one model, go for active DIs.




https://en.wikipedia.org/wiki/DI_unit



''''DIY DI'''''

Transformer-wise
* the impedance conversion means the impedance should be around 12:1 or 8:1

* you don't want it to saturate under typical use, so...

* the impedance itself matters
: higher impedance means better low frequency response
: higher impedance comes from more windings (implies more cost) and/or or a more permeable core material


* and if you want to be fancy, you care more specifically for ''flat'' response.
: (though note that one of the reasons transformers have a specific sound is that they distort nonlinearly, and favour bass {{verify}})


Studio grade transformers can be a few dozen bucks. 

Even a halfway decent one seems to be a dozen bucks.




https://www.gearslutz.com/board/geekslutz-forum/131616-transformerless-di.html

http://www.nfaudio.com/wp-content/uploads/DIY-DI.pdf


-->

==Monitors==
<!--

'''Studio monitors''' are for the mixer to get an uncolored 
: the uncolored is mostly so that you won't EQ for frequency bias that isn't present in the mix, just in those speakers
: also mostly [[near-field]], so that you won't hear more reverb than there actually is

Studio monitors aren't strictly necessary, in that headphones work too,
but it's nice to hear both how the mix works in isolated-ears headphones and in arrives-in-both stereo.

People working in DAWs may also like them while experimenting, 
in part because any stereo imaging going on will be more true to how other would people hear it.




'''Stage monitors''' are for the band, 
and seem primarily used to let each 

(This is sometimes a less-processed version than is going out on main)



* Consider monitor speakers
: stage monitors, and studio monitors, are intended to be near-field, i.e. close enough to give mainly unreflected sound
:: Studio monitors are set up this way, just to give a cleaner, less-room-reverbed version of what you're doing
:: these are also the things on stage close and angled up at the performers, largely so that they can hear each other in a sensibly-mixed way (rather than whatever the venue does to the sound)
: these are often good at lower frequencies, which makes them one tool that helps find mic rumble that you want to isolate.

-->

==Feedback==

<!--

Feedback is that loud squeal that happens when people use mics wrong on a sage.



That mic-and-input almost necessarily has amplification,
and there is a point where the audio 


It happens when audio output makes it back into the input,
and specifically when the overall system is set up so that it comes out of he speakers a little louder than the first time - because that will repeat and repeat, and soon be at the maximum level the system is currently prepared to output, which is usually pretty loud.


On stage, the path is usually from [[stage monitors]] to the stage mic.

In an audience, it might happen if you have a [[roving mic]] and the path is from audience speakers to roving mic.




there is a point where it arrives at the mic is strong enough that it will be put out on the speakers louder than the first time, and again, and again.
(it need not be ''originally'' caused by the source next to that mic, just that it comes back to the same mic)




When you want to avoid this, you might add a little mental modeling.

Say, you could say there is a 'loop gain' that is positive. 
It's often easy enough to figure out what that point is - just ask someone to futz with the mic. 
Mark the point the fader is at.
It doesn't matter much what you call it, just don't go above that.



Meaning its grows stronger, become annoyingly loud, and not go away without some help.


There are some forms of feedback that are not an issue, because they are more controlled.

There are devices to do intentional feedback on guitars for long sustains, which is modulated by the player.

Delay/echo effects are sometimes (temporarily) run with positive gain, e.g. for a wall of sound effect.


It's a single peak of a tone because while feedback could peaks at any pitch 
(if you don't believe me, play with the EQ and try again), one will be the strongest,
and feedback is a sort of resonating effect.

That also means that that pitch may change with the nature of the mic, speaker, acoustics, amp, 
with directionality of speaker/mic, with distance, with EQing, and with specific effects being used,
and with other explicit and implicit filtering in the system.

For example, public announcement systems at your local station tend to have speakers with relatively little bass,
and be used for voice, meaning its feedback is likelier to be a screechy few-kHz thing. 



How to avoid it:
* use cardioid mics (or other directional, but most others have side or back lobes).
: Especially on stage, especially if you use monitors.

* point the big speakers at the, possibly even shield them from you -- and have monitors for yourself
: yes, monitors are sort of the same thing, but monitors are a lot more controlled 
: and feedback issues can be fixed, individually, from the booth, without changing what the audience hears

* "ring out" a stage setup. This refers to taking finding the most feedback-sensitive mics, and frequencies, and EQing down them from the monitors 
: needs two persons (or a ''lot'' of running around, or remote control of the booth) to do, and a decent EQ between mixer and monitor, and needs to be done for each set of equipment, and preferably with some example playing
: (similar things can apply to a PA system, which is part of why they can sound tinny)

* there are feedback detectors. 



https://en.wikipedia.org/wiki/Audio_feedback

https://en.wikipedia.org/wiki/Ringing_out
-->

==Micing cabs==

<!--
"Micing a cab" means pointing a microphone at a a guitar amplifier's speakers.


This has roughly three uses:
# you ran out of adapters or sockets somewhere and this saved you
# you are avoiding a ground loop
# you like the EQing this adds
# you like the specific distortion


The main reason people started doing this is the first two, which are practical bits of stage wiring.

It's not the best way to avoid that ground loop at all - but a simple to understand one that works.


Guitar people make a really big fuss about #3, but it's also the most nonsense,
because if it's primarily EQ, you can do much the same on a mixer,
without picking up noise.


The fourth is a better reason, but more rarely is the main argument from guitar people, {{verify}}
because distortions often come from pedals, because that's easier to control.
-->


=Studio mostly=

==Audio interface notes==

Audio interfaces are often a combination of 
* sound card and
* some specific-purpose inputs and outputs
:: both in terms of the plugs
:: and the amplification that ''some'' things need

<!--

You may want one if
* the thing you want to connect uses a plug you don't have on your existing sound card
* the thing you want to connect uses a voltage you can't use as-is on your existing sound card
* the noise specs are better


More practically: 
: if you have a microphone with an XLR plug, guitar, or other stage instrument.


And yes, you can make those work on regular sound cards as well, but not ''quite'' as well.
The difference can definitely be overstated, though.



How much of each spec do you need?  How much can I expect to spend?

You can get entry-level for EUR40, a more useful one for EUR100, and a good one for around EUR200.  More than that is largely for features.


That depends a little. 

Broadly speaking, a lot of halfway decent interfaces can be used very well, 
especially in fixed installations.


'''Take noise floor / EIN'''

Audio interfaces are the first component in processing, and in some sense most crucial:
if you do something wrong here, nothing later can save you, e.g. in terms of noise.

A 20dB lower noise floor can translate to being able to make a 20dB mistake in gaining
without much consequence. But you don't ''need'' it if you are paying attention.

If that 20dB lower noise floor means 20dB larger dynamic range (which it should), 
that .



'''Take 24-bit, 192kHz'''

While irrelevant for most playback, recording is actually one of the few places it ''can'' matter, even if it usually still doesn't.

A fixed, one-purpose installation may need neither, in which case this is just label bling,


-->

=On gain, and levels throughout=
{{stub}}

===In theory===
<!--

For context, gain is amplification.

'''gain staging''', then, can be seen as setting reasonable audio levels, at every point of interconnect and gain/attenuation, for two reasons:
* Avoiding settings that mess with quality
* Setting Similar levels in a way that is useful to you



'''Avoiding settings that mess with quality'''
: at every step, you can set gain and/or volume (if they are different, they often are not) too low (SNR issues due to the noise floor) or too high (distortion issues) 
:: once you've messed that up, you can't get it back
:: so when you have issues, you probabably start looking at the earliest part first 
::: {{comment|(this is also part of why niver mixers have things like [[PFL]] - for a few more bucks it it makes this debugging faster)}}



'''Setting Similar levels in a way that is useful to you'''
: this is the relatively boring part, though important if your everyday life is sound checks and such


Once you've messed it up in terms of 

you can't fix it later, 
so it's always useful to see the earliest point at which something might have gone wrong.
The ''input'' gain is most important, but if you want to, you can 



In other ways ways, it's also important to not lose later.



Sure, the most important point of gain is arguably at input because you can't get back what you never had,
yet the point of the ''stages'' is to ''also'' not lose it later.


Why?
* analog [[audio voltages]] imply a practical minimum and maximum, and about noise levels,
* but even in digital systems that would perfecly preserve everything (and that's not quite how that works) you would still do it, to consider effects. The concept would be easier to explain if they trigger at specific levels, but most have a smoother response. If it emulates something analog, this probably matters.


While 
: in the digital world, you could theoretically carry through a signals of wildly different amplitudes with minimal effect on fidelity,
: in analog every device has some low-level noise you can let signals fall into, so you have to think about the amplitude. 






'''A practical maximum''' because, well, there's a defined maximum, 
so above that circuits do not have to make any guarantees (though circuits designs often have some leeway).

So yes, every individual volume knob could be too high and cause distortion
: 'make things loud in a room' is up to the very last step, the amplifier


'''A practical minimum''' because descrease the signal amplitude enough, and one of the other parts of the system is going to make more noise than that.
: Low quality designs aside, you will hit thermal noise soonish after, and it's a lot more practical to stick around reasonable levels than to, say, liquid nitrogen cool your audio gear.
: And, when that is fed to digital sampling, its quantization mostly presents as a noise floor.
: Noise floor used to be more important - early days, things were recorded louder, primarily because old gear had a higher noise floor than even most cheap hardware has today
: today we have a bunch more leeway (but it'll always be there)

-->
===Gain staging===
<!--
'''Gain staging''' mostly just means
* Get it on a reasonable level as early as possible, often using [[#Gain_and_trim]]

* Keep it there throughout
: because each part of the chain has such a minimum and maximum, and this is the best way to stay way from both

Gain staging won't make your audio better quality, it avoids it geting words for avoidable-now, hard-to-fix later reason,
so it's a good habit.



A little more practically
* when turning it up to be loud enough, consider that the source's peak may be perhaps 20dB or so above its typical levels
:: keeping the typical levels below the maximum is called giving it '''headroom'''

* keeping it loud enough is called keeping it above the '''noise floor'''
:: also it avoids having to do extreme amounts of amplification later, but in most hardware that is a secondary concern, or none at all


Arguably you're best off with every part at moderate levels, and attenuating things down as late as you can.

But to avoid that noise floor being higher than necessary, the useful advice for most recording is "aim for the typical levels to be 20dB or so under the maximum".



That annoyance is why this is useful even in all-digital mixing, where the 'too low' might theoretically matter very little (because in theory you could actually sum up all the gains and attenuation, and do that once, at the end. ...if there was no other processing going on, which there is).


There are much more practical reasons for gain staging, like that it's often useful to reduce the amount of channels using mixing groups like 'all drums' and 'all guitar mics', and that only makes sense to do if they're at comparable levels.

Also it may be handier than starting off with some sliders near the bottom because the adjustments (on a dB slider) are more granular there. (not about quality, just pragmatism)

And various audio processing devices, pedal or not, assume things about levels. The best examples of this would be compressors, limiters, or other dynamics processing (because their thresholds put some absolute meaning to levels that most other things care less about).



And to a larger point, once a signal is digital,
it will incur no more noise, at least not in the way an all-analog setup would.

...and yet, there is still a habit of people to record very hot. 
Out of habit.

Sometimes that's merely unnecessary.

Sometimes it's actually a bad idea, because you lose the headroom to deal with peaks, but also, a lot of gear is less comfortable once you push it to their maximum,
from effects to preamps. (you might call it their ~0dBFS but you don't actually know that point)


Generally, 
: you want everything you can meter to hang around -20dBFS usually,
: you want peaks to rarely stray beyond -10 .. -5dB range.
for a good deal also because that means peaks that ''will'' happen will also still be in its comfort zone.

(Some styles of PPM/VU meters put focus on specific figures, like -18dB and -9dB.  These figures have useful historical reasons, but aren't magic.)


(...while recording or mixing anyway, because in both cases you simply will not know what your peak is.
In mastering you can know precisely, and can master for very specific goals)


https://www.musicianonamission.com/gain-staging/



'''Mixers muddy the concept a little'''

Mixers make some useful choices for you:
* the first knob may do both gain and attenuation. This depends on context:
:: mic inputs often range from a little to a lot of gain
:: line inputs often range from a little attenuation to a bunch of gain

* the second is often the fader (slider, can be a knob on smaller mixers)
:: this is often also a range of 'bunch of attenuation to unity to some gain'




-->

===On mic preamps===

<!--

Pre-amps is mostly about [[gain staging]], putting things at comparable levels in a mixer. 

...but with some extra constraints from the typical specs of mics.



"So do you need to buy a preamp?"

There are, broadly, two arguments:
* amplification of unusually very quiet mics
* noise
* its sound


If gain is your own reason, you generally ''don't'' need one, because all XLR inputs have their built-in preamp. 
They have to, to be compatible with the variation of how much signal each mic puts out.



'''The sound'''

One reason to buy a separate pre-amp is probably to '''color the sound'''. 
Whether it's via some intentional filtering, some extra features possibly even including some basic dynamics processing, 
or by the preamp's design nature (e.g. tube preamp).

[[Transparency]] doesn't really apply here - we care about it sounding nice.

And yes, some of that is just EQ, and can be done equivalently later, e.g. at the mixer, or in software.
As can the dynamics processing.
Even if you have characterized exactly what it does (which is nontrivial) and can replicate it closely (also not necessarily simple), it can still be ''handy'' that you have hardware that is fixed to the thing you like.

But if there's any non-linear part (e.g. from tube amps), that coloring, while often subtle, is harder to imitate,
so if a specific piece of kit does coloring you like, then you like it.


https://songwritingandrecordingtips.wordpress.com/2012/02/14/recording-the-great-mic-preamp-debate-are-they-really-necessary/

https://www.neumann.com/homestudio/en/will-a-better-preamp-give-you-lower-noise



'''The noise argument'''

There is an argument that your mixer/interface's preamp is probably crap, 
and you need an external one to have anything decent.

This is bullshit in ''most'', but not all cases.


That "probably crap" argument was maybe true for cheap hardware fourty years ago - you didn't know what you were going to get.
Get overly cheap stuff ''today'' and you might still manage to find crap.

But today's component quality, and long history we have now,
means that even amateurs can design something that gains 30 or 40dB comfortably - fairly cheap and simple even.

And 40dB is a ''lot''. 

You'll never need that much for condenser mics, because they by definition have their own preamp built in (also why they need phantom power)
meaning at the mixer you rarely have to gain more than 10dB.


Dynamic mics, on the other hand, work with ''only'' the mechanical energy,
That, and often being intended to record louder things, means their output may easily be 20 to 40dB quieter.


That's still mostly within the comfort of many halfway-decent preamps.

This  is where  the 'known quantity' argument comes in:
it's more practical to use a known-good box inbetween so that 
you could plug it into crappy cheap mixers with crappy cheap pre-amps, and still be good.


There are however some dynamic that output a signal that may be very clean, very low-distortion, but also very quiet. 

This is the part where it gets interesting.

For one, amplifying ''any'' electric signal 60dB is not as easy to do cleanly in one go, and most things don't try.

It's not about the amount of amplification - that's fairly simple.

It's not accidentaly introducing noise in the proces.

That noise has to be well ''below'' the input signal strength, because if introduced early, then even a perfectly clean amplification would just... amplify it right along.


Put another way, it's mostly the ''first'' amplification that needs to be low noise. After that it's more of a regular gain staging thing.


In fact, the cloudlifter ("the thing you must have to use an SM7B properly!!!1" only gains up to 25dB.
...but does so with negligible noise (which is why it costs a bit).
which means a regular preamp now only needs to do a very comfortable 30dB or so.




'''So do you ever need +60dB?'''

Rarely. 

One reason is that environmental noise is often less than 60dB below what you want to record.
Your room? Probably around that. 

Put a sensitive mic there, gain it +60dB, and suddenly you will hear things out the window perfectly, you computer's fans, your typing, maybe footsteps from the next room.  Unless you wanted to hear all that, you are doing something wrong (sitting too far away, usually) that gaining cannot fix.
: You might notice that with voices in a quiet room, which is why voice booths are really useful for such vocal work. 
: On a stage, though? Most pop or rock vocal stage mics are dynamics because it's fine (and also recording louder vocals anyway)
: and why e.g. streamers can sanely choose to ignore this completely


Most mics's inherent [[self-noise]] is also around -60dB of typical use (ballpark) For fancier mics that's lower.

One example is the SM7B, a dynamic mic aimed at vocals, known for being a quiet mic,
and one of enough quality that when you amplify it up about 60dB, you still don't really hear its self-noise.

That makes it useful to deal with both medium-loud vocals, and still be able to pick up really subtle things.

You generally wouldn't use it that way, but you ''could'', and that flexibility makes it great for a few specific uses - and also mostly an exception.

Point is, there are cases.



In practice, (these are estimates; citation very much needed)
Most prosumer can comfortably gain 40 to 50dB without noise, some up to 60. 
Doing 60dB is basically edge where the circuit needs more care than usually spent.
Studio grade equipment always spends that care and has no problems with 60dB.

Cheap prosumer XLR gear might be an unknown. Basically, if it doesn't dare list its noise specs, they're probably not great.


When you know there's a limit to how much you can cleanly do, devices have a choice
: allow the gain to be dialed up higher, because you asked for gain, not low noise, ha-ha

: more expensive prosumer is more likely to limit the dial to what they can do without noise
:: how much that is depends a bit on whether it was a priority in the design specs
:: There ''are'' varied prosumer devices which can do 60dB well.

: professional devices will just have the specs to deal with a little more than 60



Above 50dB or so is where the "known quantity" argument starts to make sense.

If you ''know'' your particular box does a clean pre-amp, 
then you know that preamp in your mixer/interface (that you can't change) will never have to be pushed to its limits,
and you won't end up with stuations that sound a little worse or quiet because the PA system is fourty years old.
I mean, it probably still won't sound great, but it's the ''principle''.





Note that a preamp does nothing to ''improve'' the mic's signal quality. 

It will not lower noise - unless your setup was unnecessarily adding noise.

It can't save you from a long/worn mic cable's noise. 
It can't save you from a mic's internal noise.

Any noise that the mic itself introduces alongside its signal (tends to be more in cheaper designs) are inseparable - you amplify both, always.



Note also that this doesn't apply to condenser mics. 
Obvious maybe, but people ''still'' fall for "I've heard buying a preamp will make my mic better" .
-->

<!--
"So why don't they put two amplifiers in a row?"

That's what a bunch of people with more complex setups end up doing, actually.
It's not the lowest-possible-noise setup, ''but'' it may avoid cases where not enough gain means falling into a noise floor somewhere.

And yes, that's a relatively valid solution where the limitation is the op amp's noise. However, it's usually more the device around it.
-->