Proper Gain Structure & dbfs?

[jupiter8]

Just for funsies i once lowered the volume of a sample with 120 dB,rendered it as 32 bit float. Increased the gain by 120 dB and it canceled out with the original completely IE they were exactly the same.

I bet you're a lot of fun at parties!

Yeah that was a complete laugh riot!

[MyKill]

- this YouTube video:

The link says the video doesn't exist. Got another link to it?

P.S. Thanks to everyone taking time to weigh in

[Compyfox]

Cause I'm stupid to post the correct link:

Some compressors and ampsims like to see a certain dbfs due to inflexible input gains, but it's easy enough to fudge things by adjusting levels before and after with volume filters by ear.

I actually do that with Slate VCC (bus plugin), if I want to overdrive the bus on purpose. So the chain would be:

Gain (boost) -> Console plugin -> Gain (attenuate the same ammount as with boost)

I got a more saturated signal, but I'm still at a suitable worklevel.



I also compensate the ouput of compressors and to a certain excend even EQ's. So if I hit "global bypass" for a channel, I can easily hear what's going on and whether or not it's doing the signal any good without having any huge volume/loudness bumps.

[MyKill]

... I use Satson CM for that. I love that plugin so much. Except the stupid reflections on its "screen"

To check it out, I got the digital version of the issue that the Satson plug came in (~$5 online opposed to $12 print edition). It's the only way you can get it, believe me, I looked. If anyone's interested:

http://www.musicradar.com/us/computermu ... ded-560053

I like it, especially for $5, so thanks for that! The reflection doesn't bother me, but it does look like a reflection of a 4 pane window or something, not the typical 1 directional light look so popular on GUIs these days. I'm a graphic designer by trade in addition to a musician/producer, so I pick apart every GUI from every plug in I have. Waves stuff could use an update- for $3-500, could I get something that doesn't look like Windows 95 please?!? Nit-pickey I know, but when you do graphics for a living, you get irritated when something looks awful..

ANYWAY..

A question... not sure if it's Cubase or it's just how it works (the manual didn't say much, and perhaps it's my inexperience with analog here), but when I put it on a buss channel, it only monitors the input. Meaning, if I put it on my Drum Bus, since the Cubase routing is Pre-Fader, moving that bus fader doesn't affect the plug in's VU reading; only if I drop a fader level from one of the tracks feeding it does the VU level change. Now when I crank some of the plug ins gain (on the Drum bus chanel) to hover around 0VU, obviously my channel dbfs peaks raise. So if I drop a Satson on just a Kick channel, I'm only seeing the actual audio file's RMS since dropping the Kick channel fader doesn't affect the reading.

Obviously if I wanted to monitor the channel with fader movement, I could just slap it in Insert Slot 7-8 (Post-Fader), but the manual states: "Load one instance of Satson Buss on each bus channel .Finally, insert Satson Buss on your DAW's master bus channel. In all three contexts (channel, bus, and master bus), we recommend using Satson plugins as the "first inserts" in your effects chain.

So, just to be clear, when you say

... just keep the volume of all the tracks in the mix around, but not beyond, -18dB average RMS, and use the VU meter for that. ... I keep my tracks at around -18dB average RMS and usually finish up peaking at about -14dB average RMS on the master. Usually I get to around -3dB dBFS peak on the master channel this way.

are you saying keep all your individual "TRACKS" (kick, snare, guitar) @ -18 average RMS, or were you refering to the Bus "tracks" that should be that level, then leveled by ear and keeping an eye on peak levels?

[Compyfox]

Obviously if I wanted to monitor the channel with fader movement, I could just slap it in Insert Slot 7-8 (Post-Fader), but the manual states: "Load one instance of Satson Buss on each bus channel .Finally, insert Satson Buss on your DAW's master bus channel. In all three contexts (channel, bus, and master bus), we recommend using Satson plugins as the "first inserts" in your effects chain.

Such manuals are a bit misleading.
What is actually meant, is that SATSON does act as a channel preamp if it's on the first insert.

Signal flow:

Signal (mic, VSTi, send aux, group, etc) -> channel -> Insert 1 (Satson) -> Pan/Fader -> Summing Bus

Satson on the summing bus would then act as a "summing" device, adding crosstalk and the likes, and also show the strength of your signals.

So, Satson as Insert 1 uses pre-fader measurement (PFL gain boost/attenuation)
Satson on the Summing Bus is the actual console PPM/VU for the whole sum.


Pretty much all console emulation type plugins work this way. Just like a regular mixing console would as well.

are you saying keep all your individual "TRACKS" (kick, snare, guitar) @ -18 average RMS, or were you refering to the Bus "tracks" that should be that level, then leveled by ear and keeping an eye on peak levels?

I think he meant channel individual.
Though keep in mind, a Kick is more bass intensive than a transient heavy Snare. So at 0VU/-18dB RMS, a kick would give the right readout, but a snare (due to it's lack of lowend for the short transient it's measured) it might only show a fraction of the signal strength.


This is why I said:
Use both a VU/RMS meter and a digital meter.

[DuX]

What Compyfox said.

I should have said: keep the *loudest signals* from exceeding -18dB RMS and set other channels as it fits the mix. Usually the loudest signals are bass, drums, vocals.

As you probably noticed, Satson CM's VU meter is calibrated for -18dB which is great, isn't it?

[Tricky-Loops]

Recently I made a thread and asked if there's a negative sonic impact (i. e. distortion) if my fader is in the red light, and some people said, no, there isn't as long as it's not TOO much.

Now I read -18 db headroom which is A LOT. I think you mean the group bus faders, not the master bus fader?

I had my master fader always at -3 db, is it a crime?

[MyKill]

keep the volume of all the tracks in the mix around, but not beyond, -18dB average RMS, and use the VU meter for that. It's that simple.
I keep my tracks at around -18dB average RMS and usually finish up peaking at about -14dB average RMS on the master. Usually I get to around -3dB dBFS peak on the master channel this way.

are you saying keep all your individual "TRACKS" (kick, snare, guitar) @ -18 average RMS, or were you refering to the Bus "tracks" that should be that level, then leveled by ear and keeping an eye on peak levels?

I think he meant channel individual.
Though keep in mind, a Kick is more bass intensive than a transient heavy Snare. So at 0VU/-18dB RMS, a kick would give the right readout, but a snare (due to it's lack of lowend for the short transient it's measured) it might only show a fraction of the signal strength.


This is why I said:
Use both a VU/RMS meter and a digital meter.

So, Satson as Insert 1 uses pre-fader measurement (PFL gain boost/attenuation)
Satson on the Summing Bus is the actual console PPM/VU for the whole sum.

Bear with me guys, we're almost there. The VU stuff in new territory. Which Satson am I monitoring this on then? If I'm trying to keep my -18dB average RMS for EACH channel, are you saying that is the reading for each soloed channel when viewed on the VU meter inserted on the Master Stereo Bus?

To get a 0VU reading on the master bus instance for any one channel, like the kick with dbfs peaks of ~ -5 and much lower RMS reading, I would have to slam it with compression or the Gain from a channel level instance of Satson, which itself shows a Pre Fader level, so if I bypass it, my signal level drops significantly, which its not supposed to...

When I put an instance on the channel and crank the gain till hist 0VU average, it sounds incredible (or just loud), but better sounding in the mix either way.



I know this is something so basic, I'm gonna feel stupid when it clicks, but ... I don't quite get where it monitor point is or something

[Compyfox]

Man, this will be a long post - so please sit back, grab some popcorn, and prepare to be hit by a lot of theoretical stuff!

Bear with me guys, we're almost there. The VU stuff in new territory. Which Satson am I monitoring this on then? If I'm trying to keep my -18dB average RMS for EACH channel, are you saying that is the reading for each soloed channel when viewed on the VU meter inserted on the Master Stereo Bus?

Ideally, yes. Average signal strength at -18dB / 0VU, peak (on a digital meter) can exceed up to 12dB. The faders need to be at unity for this.


Which is why you have VU's on the individual channels:
Bass intensive material should hover around 0VU/-18dB RMS
Transient intensive material should go up to -9dBFS peak maximum

Again, channel fader at unity.


The sum of you signals should ideally also hover between +/- 1VU (around the -18dB RMS mark), or up to -3dBFS peak as absolute maximum. Don't touch the master fader (leave at 0dB = unity), use your channel faders instead to create a mix (sum) that doesn't exceed these values.

When I put an instance on the channel and crank the gain till hist 0VU average, it sounds incredible (or just loud), but better sounding in the mix either way.

That is the way to do so. It's calling "leveling in your signal".

Once more: bass intensive signal = 0VU max, transient intensive signal = -9dBFS max. Then you can mix to your hearts content.

I had my master fader always at -3 db, is it a crime?

Actually, my faders are at unity (0dB) until I start mixing. I do not touch the master fader at all anymore.


We're talking about -18dB RMS, and -9dBFS digital peak at maximum on a per channel basis - showing on the meter(s). The faders aren't touched at this point. And with the approach of using a reference level, there is even no need to do so, and you'll never exceed the 0dB ceiling limit of DAC's (you won't get any clipping).



The reason for a reference level is to actually work at an optimum level both within the DAW, and maybe even integrated hardware.

Unfortunately, there is no relation between dBFS and dBu. It is something that someone decided after calibrating gear back in the days. So there are common suggestions(!) what voltage the DAC should have on output, if you use +4dBu as relation, which would mean 1,228V (a 1kHz signal measured with a voltage meter). And since a lot of hardware providers talk about "headrooms of up to +24dBu", we have a certain headroom to work with right from the start.

All this can be ported over in "theory". Actually also with measurements, thinking about it and throwing in some educated guesses. Which nowadays clocks in at about -20dBFS or -18dBFS. (also depending on impedances, prefilters, yadda yadda)



AGAIN - AND THIS IS IMPORTANT:

There is no decibel to dBFS converter

Notice - Comparing dBu and dBFS: There is really no fixed
world standard like e.g. -20 dBFS = +4 dBu = 0dB VU.
The digital peak scale is not equivalent to the analog RMS scale.

You can never match dBFS and dBu

Reference:
http://www.sengpielaudio.com/calculator-db-volt.htm


So don't take given reference levels for granted(!). Always measure with test tones, apropriate meters (including voltage meters!). But these are known values/reference levels nowadays. You just need to calibrate your gear accordingly.



Now, let's go back to the theoretical stuff:
If -18dBFS would mean +4dBu (in theory!), then we're at optimum level for most of the available equipment. So we have up to 18dB (peak) and roughly 6dB (RMS) as headroom. Remember the +24dBu that some hardware providers talk about? Yeah, that'd mean 20dB headroom in the hardware realm until stuff really starts to go downhill. Though exceeding more than 10-12dB is not recommended, and you'll start to notice that real quick.


Actually, ever wondered why mixing consoles have bargraphs that have positive values? Especially cheap ones like the Behringer for example?

Here we're talking about an optimum level as well. Again, +4dBu would (once more, theory and stuff) resemble 0dB on the meter. And with the bargraph going up to 12 (first red LED), we have a headroom of over 10dB for peaks (I'll come back to that in a minute).

BUT... now it's getting tricky.

This bargraph is a PPM. Meaning, it responds at/around 5ms. It's not sample acourate like a host. Which brings us to the PPMs I mentioned earlier (IEC 60268-10). These are setup to the optimum level (like +0dBu = 0,775V) and then declare what the maximum level should be (+9 dBu = 2.18V - measured in RMS!). I mention "measured in RMS" or better said "VU", since actually the voltage(!) is measured. And voltage meters are slow in response (see next paragraph). So using an optimum worklevel of 0dBu to +4dBu, we also have a RMS range of +4dB (which I'll get to in a minute as well), or even +6dB depending on the available scale for the RMS meter.

Since this particular measurement tool (voltage meter) is also slow (rise/fall up to 300ms), peaks can exceed that value up to 12dB (peak!) and still won't be registered on the meter. So we need a savety zone. And in order to measure what's really going on with your signals, you need both an RMS meter and a peak meter (though if it's a quasi PPM, then you need a suitable headroom, and that is the -9dBFS we're talking about).



Confused yet?
Trust me, I sure as hell was myself - until it made click. And that took a couple of days (not to mention ton of tea and hot chocolate).

*cough* Anyway...



Now let's port it over to the digital realm (even more theories!):
Assuming that a test signal at -18dB FS or RMS (1kHz sine) on a +4dBu output resembles 1,228V, then we have our optimum worklevel. And our reference points.

A VU (or Voltage Unit measurement device) could then show 0VU at 1,228V, and we have the reference level for the loudness measurement (300ms rise/fall - unweighted meter). But since the reference level was -18dB(FS), we know that the theoretical headroom is 18dB until we reach the digital ceiling (0dB).

Though it is advised, even with modern ADC/DAC, not to go near the ceiling. Some DAC's start to barf at -1dB, others already at -3dB. So maybe even pull it down another 3dB to reach -6dBFS and you'll never enounter any problems.


Effectively you then have (from -18dB to -6dB) a peak headroom of 12dB (remember the 12dB I mentioned earlier?), and RMS of 4-6dB (remember the voltage meter scale and the range between 0dBu and +4dBu I mention earlier?). Which is a lot in both cases. And what do VU meters have as maximum scale mostly? Yes, +3dB or even +4dB.

Bam! Filled the gaps.




Even though this concept is a lot of guesstimate(!!!), it's understandable why VST plugins have the reference levels we know (-20/-18dB). And why it is important to have measurement tools that we can setup accordingly, why hardware needed to be calibrated.

Someone simply decided, that this and that voltage needs to resemble this and that -xx dBFS with a test tone of 1kHz. Though ensuing from the fact, that both the ADC/DAC and the measurement tool use the same voltage (+4dBu). And we have since adapted it.


But in the pure digital realm, nobody cared about reference/optimum work levels. Until people like Skip Burrows and Steven Slate came along. One set things right, and the other one gave us a very good educational tool. And some crazy German sat down for a couple of days to make the connections.

Which I just presented to you. Hopefully in way that was easy to understand.





IMPORTANT DISCLAIMER (TL;DR):
This is a highly theoretical topic. Do not take my word for granted. Up until this day this whole "reference level" thing is highly debated in several audio engineering communities. I direct you towards the Sengpiel Audio infopool again in this case.

Someone decided that xx dBFS = +4dBu = 0VU, which was due to the fact that a meters were calibrated to a certain test signal (1kHz at x Volt). And that happens to be now the known "reference level" wel all talk about and some of us use on a daily basis.

So please take the upper post as explanation how this all might work "ideally" - but it's not a correct and scientific explanation. It's an allround accepted "guesstimate".





It's late for me, more tomorrow (if needed).

[MyKill]

if I want to overdrive the bus on purpose. So the chain would be:

Gain (boost) -> Console plugin -> Gain (attenuate the same ammount as with boost)

I got a more saturated signal, but I'm still at a suitable worklevel.

Bear with me guys, we're almost there. The VU stuff in new territory. Which Satson am I monitoring this on then? If I'm trying to keep my -18dB average RMS for EACH channel, are you saying that is the reading for each soloed channel when viewed on the VU meter inserted on the Master Stereo Bus?

Ideally, yes. Average signal strength at -18dB / 0VU, peak (on a digital meter) can exceed up to 12dB. The faders need to be at unity for this.
Which is why you have VU's on the individual channels:
Bass intensive material should hover around 0VU/-18dB RMS
Transient intensive material should go up to -9dBFS peak maximum

Again, channel fader at unity.

The sum of you signals should ideally also hover between +/- 1VU (around the -18dB RMS mark), or up to -3dBFS peak as absolute maximum. Don't touch the master fader (leave at 0dB = unity), use your channel faders instead to create a mix (sum) that doesn't exceed these values.

When I put an instance on the channel and crank the gain till hist 0VU average, it sounds incredible (or just loud), but better sounding in the mix either way.

That is the way to do so. It's calling "leveling in your signal".

Once more: bass intensive signal = 0VU max, transient intensive signal = -9dBFS max. Then you can mix to your hearts content.
...
We're talking about -18dB RMS, and -9dBFS digital peak at maximum on a per channel basis - showing on the meter(s). The faders aren't touched at this point.

I use VUMT set to -18db in all channel to set up the gain according to the rules described by Compyfox, with the Cubase channel faders set to default. VUMT is great for that, but there are other utilities one could use.
Then I use the faders to balance the tracks.

So I thought I was finally getting it...until I tried to put it into practice.



From what I'm reading, it seems that I need to use gain staging inside Cubase to get the correct levels. I need to have an instance of Satson (the VU Meter/Gain control) on each individual source channel and one on the master fader. I need to set my individual channels so that they are themselves sitting around 0 VU/ -18dB RMS,

So I set up a new project wit no plugins, imported just my kick (mono), bass (mono) and Snare (stereo, bounced from drum software. Put an instance of Satson on all three channels and one on the Stereo Out.
I ran a 1 kHz sine wave from the TestGenerator plug in on a channel, followed by the Satson VU, and yes -18dB from the Generator is 0 VU on the channel meter (actually -17.9), while the Stereo Bus VU is -17.7....???????
Ok, maybe in-precise summing from Cubase?

I started with the snare(stereo channel). At unity on the channel fader, Its channel had a peak of -2.93 dBFS, a VU reading of ~ -5 to 0, and the Stereo Out had a VU of ~ -5 to 0 as well. But I need a consistent RMS of 0VU and a peak of -9 dBFS on the channel, right?

I used the channel Gain knob (at the top of the channel, before the inserts, to get my dBFS to -9, which brought down both VU readings to -7. I then cranked the channel VU plugins Gain to reach 0VU, which brought my dBFS peaks back up to near 0. As Cubase has no Gain/Trim plugin that I can find, I started over and did the 2 previous steps as offline processes, got the same results, then tried to compensate for the additional gain by offline removing -9 dBFS to get the signal back to -9, and now the VU meters are back to -7 in the channel and the Stereo Out.
It seems I going in circles to accomplish nothing.

Apparently, I'm not getting how to "level the signal" properly. It's getting really frustrating because I thought I understood what you were saying and how to do it, but I'm more confused than when I started, and I don't get what I not getting
Fese, you are using Cubase for the same purpose.... what am I doing wrong?

Which is why you have VU's on the individual channels:
Bass intensive material should hover around 0VU/-18dB RMS
Transient intensive material should go up to -9dBFS peak maximum

Again, channel fader at unity.

If bass intensive material should hover around 0VU, what dBFS am I shooting for?
If transient intensive material should go up to -9dBFS peak maximum, is it still 0VU that I'm aiming for?

[Compyfox]

Just a short answer...

From what I'm reading, it seems that I need to use gain staging inside Cubase to get the correct levels. I need to have an instance of Satson (the VU Meter/Gain control) on each individual source channel and one on the master fader.

This would simulate a normal analog console, yes.

I need to set my individual channels so that they are themselves sitting around 0 VU/ -18dB RMS,

Or -9dBFS peak maximum. Depending on he signal source.

I ran a 1 kHz sine wave from the TestGenerator plug in on a channel, followed by the Satson VU, and yes -18dB from the Generator is 0 VU on the channel meter (actually -17.9), while the Stereo Bus VU is -17.7....???????
Ok, maybe in-precise summing from Cubase?

Totally correct behavior, since SATSON introduces saturation and a frequency fingerprint. That is adding up.

I started with the snare(stereo channel). At unity on the channel fader, Its channel had a peak of -2.93 dBFS, a VU reading of ~ -5 to 0, and the Stereo Out had a VU of ~ -5 to 0 as well. But I need a consistent RMS of 0VU and a peak of -9 dBFS on the channel, right?

Either or!
Either 0VU (bass intensive material) or -9dBFS maximum (transient intensive material).

It's normal that the VU doesn't show 0VU with the latter, as it's responding too slow.

I used the channel Gain knob (at the top of the channel, before the inserts, to get my dBFS to -9, which brought down both VU readings to -7. I then cranked the channel VU plugins Gain to reach 0VU, which brought my dBFS peaks back up to near 0. As Cubase has no Gain/Trim plugin that I can find, I started over and did the 2 previous steps as offline processes, got the same results, then tried to compensate for the additional gain by offline removing -9 dBFS to get the signal back to -9, and now the VU meters are back to -7 in the channel and the Stereo Out.
It seems I going in circles to accomplish nothing.

Since you compensate something that doesn't need compensation.

If the signal is leveled in properly, you don't need to use the fader to turn up the signal level again. The signal already is at optimum level. You just mix it all together and the summing bus should not exceed (ideally) 0VU or -6dBFS peak maximum.


BTW:
The Cubase "gain knob" in Cubase 6 is at the top of the channel strip. In Cubase 7 it's in the prefilter section of the MixConsole racks.

If bass intensive material should hover around 0VU, what dBFS am I shooting for?

Bass intensive material at 0VU would usually never exceed -9dBFS peak.

If transient intensive material should go up to -9dBFS peak maximum, is it still 0VU that I'm aiming for?

Depending on the signal. Sometimes it does, most of the time it won't.



Really:
Focus while leveling in on the specific channels ONLY(!), do not compensate the output of them with the channel fader. Unless you're mixing.

[MyKill]

Either or!
Either 0VU (bass intensive material) or -9dBFS maximum (transient intensive material).

It's normal that the VU doesn't show 0VU with the latter, as it's responding too slow.

...Since you compensate something that doesn't need compensation.

BTW:
The Cubase "gain knob" in Cubase 6 is at the top of the channel strip. In Cubase 7 it's in the prefilter section of the MixConsole racks.

I think that just cleared up a lot. I guess I wasn't getting the OR part. I thought it was a combination of the two on each source track. And, yeah, the gain knob is what I was refering to

I used the channel Gain knob (at the top of the channel, before the inserts, to get my dBFS to -9, which brought down both VU readings to -7

Each channel has a gain thats pre-inserts, they just don't have a simple Gain default plugin. One's only available as an off line process.

So I think I got it (it's too late to fire the DAW back up though):

My snare that is coming in peaking at -2.93 dBFS when the channel fader is at unity needs its gain (as a knob move or offline process) so that it's max peak is -9 dBFS at the same unity gain fader position. Then, treat the insert gain staging just as you would with a compressors make up gain compensation knob.

With the less transient, bass-heavier channels, I need to try to not go above -9 dBFS peak and shoot for a 0 VU RMS while the channel fader is at unity. Which, I could achieve using the channel gain, the Satson plug in's Gain knob, etc.

Now, all channels are at optimum level at unity and can be mixed to taste.

In the ballpark yet?

Now, to be clear, this 0VU in the bass tracks is a reading on the Stereo Bus Satson VU, while the Satson VU plug in inserted on the bass channel itself is more for gain staging with its Gain knob, but do I need to use it to get the signal to 0VU inside the channel plug in before compression, eq, etc so I am then able to master bypass inserts on the channel and get roughly equivalent volume, or am I using all those things to achieve the 0VU reading overall for that channel in the Stereo Bus VU?

[Compyfox]

I think that just cleared up a lot. I guess I wasn't getting the OR part. I thought it was a combination of the two on each source track.

In certain ways, it is a combinated use of both meters. You either shoot for -9dBFS peak or 0 VU / -18dB RMS maximum. Depending on the signal source.

With the less transient, bass-heavier channels, I need to try to not go above -9 dBFS peak and shoot for a 0 VU RMS while the channel fader is at unity. Which, I could achieve using the channel gain, the Satson plug in's Gain knob, etc.

You "rather shoot for" 0 VU / -18dB RMS than -9dBFS peak. See above.

Now, all channels are at optimum level at unity and can be mixed to taste.

In the ballpark yet?

Yes, you are.

Now, to be clear, this 0VU in the bass tracks is a reading on the Stereo Bus Satson VU, while the Satson VU plug in inserted on the bass channel itself is more for gain staging with its Gain knob, but do I need to use it to get the signal to 0VU inside the channel plug in before compression, eq, etc so I am then able to master bypass inserts on the channel and get roughly equivalent volume, or am I using all those things to achieve the 0VU reading overall for that channel in the Stereo Bus VU?

You can use whatever tool to get to the desired level. Satson has a built in gain boost/trim knob. I tend to go the following route in Cubase:

Signal -> Cubase's own Gain knob (Prefilter) -> VU -> whatever plugins I need -> Pan/Fader -> Summing Bus

The Summing bus itself either uses Cubase's internal metering tools (Control Room in Cubase 7), or I use another VU just in case as first insert. However I feel like.

So...
I level in the signal with the VU's on each individual channel first.
I then mix all signals to taste.
I try to shoot for 0 VU on the summing bus


Since I also compensate EQ's and Compressors to have a consistent felt loudness, I can then easily A/B the signal while using the global channel bypass.

Once you get the hang of it all, it's a fairly simple process.

[DuX]

Great posts, Compyfox. You have a patience of a teacher.

One tip about Satson CM: when you boost the gain on it, it adds a bit of bass at 40Hz, and of course more saturation at 2nd and especially 3rd harmonic that becomes stronger with higher gain and overtakes the 2nd harmoonic. The more gain, the more bass boost and saturation. When you lower the gain the frequency response stays rather linear and there's less saturation.

[Funkybot's Evil Twin]

Bear in mind that any analog modelled plug-ins you own will respond differently when fed different input levels and so may benefit from analog style gain staging. The same is true for some pure digital effects where things can be far less subtle, bitcrush being the obvious example. Also, some plug-ins (Satin being a recent example from what I can gather) may have a 'noise floor' of their own that you might want to avoid boosting later on. Other than that, the only reason to gain-stage in a DAW is for workflow reasons. You can whack up the master at the end to compensate or render to 32 bit float then normalise/boost. I prefer keeping things low so I never have to worry about headroom during mixing.

Agreed.

The problem some of the "it doesn't matter" posters are ignoring is that insert plugins are pre-fader, and you risk hitting one of these plugins too hard on the input if your tracks are too hot. If you were to put the plugin on a buss, or post-fader, then gain staging wouldn't matter. In practical use however, it's important due to the plugins we might use on a track.