Tracking nice levels

[NashBackslash]

Hello, new user here. Nice to meet everyone here.

Anyway, I'd like some advice from all you professionals here on recording nice levels that wouldn't give me much trouble during the mixing stage.

First of all, here is my setup:

Behringer 2442 (using the mixer's built-in pre amps - can't afford external pre amps currently :/) -> direct outs -> MOTU 828 MK2 -> Cakewalk Sonar 3.

I don't monitor from the mixer, that means all the channels on the mixer itself are not routed to any bus (not even main). I monitor from the MOTU, because I want to hear what is actually going into my MOTU.

I feed MOTU's main outputs into channel 16 of the Behringer mixer. This way I can use the fader on the mixer to adjust my volume. Channel 16 is the only channel that is routed to the main bus of the mixer. So in addition to being able to adjust the MOTU volume using channel 16's fader, I also have the main fader on the mixer.

I connect my monitor speakers to the "main out" behind the mixer. I'm using a pair of Alesis M1 active.

Before I start to track, I do the following. What I'd like to know if what I'm doing is correct or not.

I adjust the trim/gain on each of the channels on the mixer to be as loud as possible without clipping the mixer's pre amps. I then adjust the fader. I would monitor the level going into the MOTU (using the MOTU's front display meters) to make sure they don't go above -6 to -4dB. Obviously, this would mean that I am hearing low levels as I monitor the tracks. I end up having to crank up my speakers a bit. Then later during the mixing process, I would turn the speakers down so that I am hearing a "normal" listening level.

Is it normal to do that? Constantly adjusting the speaker levels?

Secondly, the MOTU has options to have the inputs at -4dB or +10dB, as well as a separate +6dB boost on every input. Should I use -4 or +10? Should I use that +6dB boost or stay away from it?

I will appreciate all suggestions and input. Thanks!

- Nash

 

[SouthSIDE Glen]

Hello, new user here. Nice to meet everyone here.

Welcome to the nightma...er...board, Nash\. Good to have new blood who knows how to make good posts, as this one is.

I don't monitor from the mixer, that means all the channels on the mixer itself are not routed to any bus (not even main). I monitor from the MOTU, because I want to hear what is actually going into my MOTU.
...
I connect my monitor speakers to the "main out" behind the mixer. I'm using a pair of Alesis M1 active.

If I understand you correctly, you are not actually monitoring from the MOTU directly. You may be watching the levels on the MOTU, which is fine, but your monitors are coming out of the mixer. This means one of two things, depending on how the monitor routing is set up on the mixer: either you're actually monitoring a live mix of what's conimg into the mixer or your monitoring channel 16 which is the return from the MOTU (after the signal has passed through a whole bunch of extra path.)

The 828 does have L/R Main Outs. If you're using the FireWire to connect the MOTU to your PC (recommended), then the Main Outs should be free to use as your monitor sends. This will not only tell you just what is actually being heard in the MOTU - in both directions - and not what is at the mixer, but it also gives you the front panel volume control for the mains so you don't have to be adjusting it at the speakers themselves all the time. No, adjusting at the speakers is not common, usually there is a "control room" or "main" volume for providing volume control to the monitors so the monitors can be set and left alone. This would be an ideal function for the main volume control on the 828.

Secondly, the MOTU has options to have the inputs at -4dB or +10dB, as well as a separate +6dB boost on every input. Should I use -4 or +10? Should I use that +6dB boost or stay away from it?

You should match the input settings on the 828 to match the specifications of the direct outs on the mixer. The 6dB boost shouldn't be needed.

HTH,

G.

 

[Cloneboy Studio]

I adjust the trim/gain on each of the channels on the mixer to be as loud as possible without clipping the mixer's pre amps. I then adjust the fader.

You should leave the fader at 0db and adjust the preamp until the RMS gain on the channel is around -12db, with peaks no more than -6db. By jacking the preamp and lowering the fader you are adding noise to your signal, and potentially clipping the preamp--which will add distortion.

I would monitor the level going into the MOTU (using the MOTU's front display meters) to make sure they don't go above -6 to -4dB. Obviously, this would mean that I am hearing low levels as I monitor the tracks. I end up having to crank up my speakers a bit. Then later during the mixing process, I would turn the speakers down so that I am hearing a "normal" listening level.

I recommend mixing at a variety of volume levels from soft, medium to loud.

 

[DonF]

I would monitor the level going into the MOTU (using the MOTU's front display meters) to make sure they don't go above -6 to -4dB. Obviously, this would mean that I am hearing low levels as I monitor the tracks.

No, that's not obvious at all. It sounds like you're confusing "VU" levels, which your mixer uses, with digital peak levels, which the MOTU uses. You should align the CH16 input so a -14dBFS signal on the MOTU gives you 0VU on the Behringer. (The -14 is arbitrary -- it could go a few dB either way depending on how much headroom you want.)

Don

 

[NashBackslash]

either you're actually monitoring a live mix of what's conimg into the mixer or your monitoring channel 16 which is the return from the MOTU

I'm not monitoring anything from the mixer. All the channels on the mixer are not routed to any bus that's available on the mixer. There are these little switches that I can press that tell each channel which bus they go to. As I said earlier, I don't route them to the mixer's main bus. That means it's totally silent. The only way out is through the direct outs behind the mixer.

And, yes, I realize that the 828 has Main L/R outs. But instead of hooking my active speakers directly to these Main L/R outs, I plug them into the inputs of channel 15/16 (stereo) on the mixer. I set the line gain to 0dB, leave the fader at 0dB, and route these two channels to the main bus on the mixer. If I wanted to control the master (speaker) volume, I would just play around with the main bus fader on the mixer.

The 2 main reasons why I do this:

A) I prefer using the main fader to control the speaker volume.

B) I have a Creative SB Live soundcard that I use. Why, you may ask? For talkback and MIDI playback (click tracks, etc). I've used up pretty much every input on my mixer. I even don't have enough for vocals - I have a SEPARATE 8 channel Yamaha mixer which I utilize for vocals and anything else, and I send a mono mix of that into one of the channels on my Behringer. So where does Behringer the mixer come in? The Behringer mixer has a tape in which is just RCA inputs. There's a button on the mixer that can route that signal into the main bus. So I just hook the output of the SB Live into the mixer's tape input. The SB Live has a mic input, so I use that for talk back. So, the performs would be talking to me using that 8 channel Yamaha mixer I mentioned earlier, and I get to talk to them using a mic that I plug into the SB Live.

And yes, I am using the FireWire connection so that it sends 10 separate signals into my main DAW of choice, Sonar 3.

ou should leave the fader at 0db and adjust the preamp until the RMS gain on the channel is around -12db,

Ummm. Pardon my sillyness, but what's RMS?

No, that's not obvious at all. It sounds like you're confusing "VU" levels, which your mixer uses, with digital peak levels, which the MOTU uses. You should align the CH16 input so a -14dBFS signal on the MOTU gives you 0VU on the Behringer. (The -14 is arbitrary -- it could go a few dB either way depending on how much headroom you want.)

Sorry, again. I don't quite understand this. What is a -14dBFS signal? Some of these terms are a bit new to me...

Thanks for the replies. Much appreciated!

 

[DonF]

What is a -14dBFS signal?

"dB" all by itself means very little. It's a way of comparing two quantities -- in this case, audio signal levels. Put something after the "dB," though, and you have an absolute measurement rather than a relative measurement.

The thing that comes after "dB" is called the "reference." There are several references that you'll run across in audio: "u," "m," and "V" are common, and they all refer to different voltage or power levels.

"FS" means "full scale," and it means the highest signal level your digital gear is capable of handling. Zero in this case is a digital signal that uses all the bits your system has. If you go over 0dBFS, you get nasty digital clipping. End of story. Re-track, or at least punch-in, required. So, you need to establish a reference point lower than 0dBFS; how much lower defines how much headroom your system has.

So, send a 1kHz tone out from your computer through the MOTU. I assume there's some kind of meter, either on the box or in a software control panel? Adjust the signal you're sending out from the MOTU until it's 14dB lower than the clipping level. That's -14dBFS. Then adjust the gain on the Behringer's input strip, with the SOLO button pressed, until you get a "0" reading on the Behri's meter. (That's how it's done on a Mackie, anyway. Is Behringer the same?)

I hope that's clear enough.

Don

 

[SouthSIDE Glen]

I'm not monitoring anything from the mixer. And, yes, I realize that the 828 has Main L/R outs. But instead of hooking my active speakers directly to these Main L/R outs, I plug them into the inputs of channel 15/16 (stereo) on the mixer. I set the line gain to 0dB, leave the fader at 0dB, and route these two channels to the main bus on the mixer. If I wanted to control the master (speaker) volume, I would just play around with the main bus fader on the mixer.

The 2 main reasons why I do this:

A) I prefer using the main fader to control the speaker volume.

B) I have a Creative SB Live soundcard that I use. Why, you may ask? For talkback and MIDI playback (click tracks, etc). I've used up pretty much every input on my mixer. I even don't have enough for vocals - I have a SEPARATE 8 channel Yamaha mixer which I utilize for vocals and anything else, and I send a mono mix of that into one of the channels on my Behringer. So where does Behringer the mixer come in? The Behringer mixer has a tape in which is just RCA inputs. There's a button on the mixer that can route that signal into the main bus. So I just hook the output of the SB Live into the mixer's tape input. The SB Live has a mic input, so I use that for talk back. So, the performs would be talking to me using that 8 channel Yamaha mixer I mentioned earlier, and I get to talk to them using a mic that I plug into the SB Live.

And yes, I am using the FireWire connection so that it sends 10 separate signals into my main DAW of choice, Sonar 3.



Ummm. Pardon my sillyness, but what's RMS?



Sorry, again. I don't quite understand this. What is a -14dBFS signal? Some of these terms are a bit new to me...

Thanks for the replies. Much appreciated!

If I'm understanding you correctly, your saying that your running the main outs from the MOTU to the stereo ins (15/16) on the mixer, patching those ins to the main bus out on the mixer, and then running the main outs to the monitors. If that's indeed the right understanding then you ARE monitoring from the mixer. You are monitoring channels 15 and 16 like I said originally (OK I left out the 15, my nistake )

My point is that's a lot of extra signal chain just to use a fader instead of a knob (the Main volume knob on the front of the 828.) Especially for a volume function that should only need to be used sparingly. Plus you still have the volume knob on the 828 in-line anyway with that setup, it's not as if you're bypassing anything. What I don't understand with that setup is why, with two different volume controls in line for the monitors (more than that if you add the channel strip trim and level for 15/16), why you still need to go back to the monitirs and adjust there. There should be zero need to do that.

I can understand the talkback capability, but are you saying that your using the same monitors for engineering as you are for talkback? If so that means that you and the artists are in the same room, and I'd have to ask what the need for talkback is?

I'm just trying to say that it sounds like you have things setup with your monitors to take three 90-degree turns to the left when one 90 degree turn to the right will get the job done cleaner and easier.

As far as the different measurements, "RMS" stands for "Root x Mean x Square". To keep it simple, that more or less means "average". Technically an RMS value and an average value (technically "X = Y / 2") are not exactly the same thing mathematically, but in layman's terms it's close enough to think of RMS as average. When one says dBRMS that's basically saying the average volume of an audio passage (v.s the peak, or maximum volume). One cannot tell the RMS volume of a signal unless one has RMS-reading metering. Most project-level peak meters do not have that. I'm not sure if the 828 has that option in its metering or not.

dBFS, on the other hand, refers to what you are reading on digital meters (peak meters reading a digital signal.) The "FS" stands for "Full Strength" or "Full Signal". 0dbFS in the digital domain means full digital saturation or maximum possible digital volume. this is the reading that the peak meters in Sonar display; Zero(0), at the top, above which there is only clipping, is referred to as 0dBFS. a -14 reading on those same meters would be referred to a -14dBFS.

Again, unless the Sonar meters have an RMS option where you could switch between peak and RMS values, you'd have to have a plug that calcualted or otherwise showed RMS value. But if you could read RMS average values and you saw thatf the *average* RMS volume of a song in the digital realm is, say -14dBFS (14 dB below digital maximum), then you could equally say that your signal level is -14dBRMS.

This gets confusing until you get a handle on it all. The above is where it;s easy. But when you switch over to ananlog metering and have to factor in measurements like VU, dBv, dBu, A-weighting, B-weighting, etc. it all begins to look like it has Excedrin written all over it .

All Cloneboy was trying to say there was to try and adjust your trim/faders/etc. so that a 0VU reading on the mixer yields a -14 (a.k.a. -14dBFS) reading on the MOTU. He also said that number is flexible. I personally might want to bump it up to around -9dBFS myself. But you''l find that different engineers are like different chefs., One will like more or less salt than another one will.

The the IMPORTANT point I think Clone was trying to make was that 0VU on your mixer is not the same thing as 0dB on the MOTU or in Sonar. They are actually using two different measurement scales and techniques and there is not necessarily a 1-1 correspondence in values between the two meters. By setting the digital dBFS reading below the analog VU reading, this allows you to drive the analog harder (louder) without having to worry about clipping the signal once it goes digital.

Hope I haven't confued you even more, but it's a very confusing subject to those new to it. Feel free to ask any followup questions and to searc this forum for other threads discussing "dB" metering and measurements in general.



G.

 

[NashBackslash]

So basically, the recorded signal should read -14dBFS?

Sorry, I'm a bit slow here

There are several places to set levels, in order from top to bottom:

1) Channel's pre-amp gain on mixer
2) Channel fader

Each channel is sent to their respective direct outs

3) 828 channel input gain (which I normally leave at 0dB, using -10dB recording level)
4) 828 Main out L/R (again is left at 0dB)

Main out L/R is sent back to channel 15 and 16

5) Mixer's ch15/16 gain (left at 0dB)
6) Mixer's ch15/16 fader (0dB)
7) Mixer's main bus fader. This fader is moved constantly to change listening levels.

So does that mean I set the loudest level possible at the pre-amp until the mixer's VU reads 0dB, and then at stage (3), I would lower the gain until it reads -14dbFS?



Southside Glen:

I can understand the talkback capability, but are you saying that your using the same monitors for engineering as you are for talkback? If so that means that you and the artists are in the same room, and I'd have to ask what the need for talkback is?

Um, actually the tracking room and the control room is very far away from each other (practically the other side of the building - 20 meters, I'd say). So talkback is definitely a must. I'm using very high-grade cables though so signal-loss isn't a problem. Oh and there's a two-way CCTV too so me and the musician(s) can actually see each other.

You are correct that I am taking an unecessary long route to accomplish a simple task, but the talkback capability and MIDI playback are key reasons why I have to do what I am doing now. I simply don't have enough channels on my mixer, and I do a lot of live tracking. Not every band can record single instruments separately...

nb. I hope to eliminate this problem soon with the FireWire mixers Alesis and Phonic are coming up with ;D

Regarding RMS: While there's no way use RMS metering on the 828, Sonar 3 allows me to change the metering formats, and it's possible for me to view RMS metering on the recorded signal, in real-time, from within Sonar. So no problem there.

Oh and before I go, a quick unrelated question: I'm trying to test the bass in my control room to make sure it's at a level that's pleasing enough. I don't have money to invest on expensive bass traps and acoustic foam. Currently I'm using the egg container-things to reduce mid and high reflections. Not sure what to do with the bass, though! Anyway what Hz test tone do you use to test bass?

Also, I'm using Alesis M1 active nearfield monitors, arranged horizontally, and the tweeters are in the inside (subwoofers outside). The bass is too loud when I'm sitting close, but taking steps backward (away from the mixing desk) makes the bass disappear slowly. Is this supposed to happen?

Thanks in advance! I appreciate everyone's help here!

 

[SouthSIDE Glen]

So basically, the recorded signal should read -14dBFS?

Sorry, I'm a bit slow here

There are several places to set levels, in order from top to bottom:

1) Channel's pre-amp gain on mixer
2) Channel fader

Each channel is sent to their respective direct outs

3) 828 channel input gain (which I normally leave at 0dB, using -10dB recording level)
4) 828 Main out L/R (again is left at 0dB)

Main out L/R is sent back to channel 15 and 16

5) Mixer's ch15/16 gain (left at 0dB)
6) Mixer's ch15/16 fader (0dB)
7) Mixer's main bus fader. This fader is moved constantly to change listening levels.

So does that mean I set the loudest level possible at the pre-amp until the mixer's VU reads 0dB, and then at stage (3), I would lower the gain until it reads -14dbFS?

Yeah that's basically correct. A little more detail:

In general you want your channel strips set with the fader at 0 gain. Then set your input trim/gain about as high as it will go before the signal overloads If you have an over/clip indicator, you can us ethat. Otherwise solo to the VU meter and use that meter. Either way, set the trim until the peaks just stat lighting the "over" or, alternately, your meter maxes at 0VU. Then back off the trim just enough to keep the "over"light dark and/or the meter just below 0VU.

Keeping the MOTU at unity gain in and out is probably a good idea, but keep those controls in your pocket in case you have to pull them out later to tweak.

The idea behind adjusting the input volume in Sonar to a negative dBFS reading on the Sonar meters (let's talk the "peak" reading setting for now) is to give youself some "headroom" on the digital side. To explain:

In analog it is possible - on some few specific occasions, desireable - to push a few dB past zero. Plus many analog VU meters are "slow" and don't show transient "spikes" that are actually blowing past zero into the red because the meters simply aren't fast enough.

In digital, however, there is literally *nothing* above 0dB. Shooting past 0dBFS in digital is like accellerating past the speed of light; it physically can't be done. At 0dBFS, all 24 bits (or 16, or 32, whatever you want) in the digital stream have a value of "1", the maximum value has been reached. There is no way to represent a higher value.

So you have an analog signal that may have transients that shoot above 0VU and survive, and that's OK as lng as it's not abused. But anything that tries to go above 0dBFS gets "clipped off" at zero and doesn't go any higher. This is what is called digital clipping distortion, or just "clipping", for short. Clipping is a bad thing .

This is why you want to set up your "gain staging" (i.e., your signal levels down the signal chain) so that 0VU on your analog mixer actually peaks somewhere below 0dBFS in your digital software. Setting it up that way gives you some digital "headroom" in Sonar to fit without clipping any signals that happen to shoot over 0VU on the analog side. This also allows you to get the maximum gain/volume out of your analog signal which, if done correctly at the right places (like specified) will increase the distance between you signal and the ambient noise in the analog circuitry; it will keep the electronic noise level down.

Now as far as the -14dB figure that Cloneboy shot you, there's nothing wrong with that figure, but it's not a number that's chiseled in stone, either. Part of it has to do with individual engineer preference, part has to do with how hot you want to push your analog readings, part has to do with the amount of actual inherant noise in your system, and part has to do with the number of total tracks you plan on mixing together once you're done. On very clean signal chains with just a handful of final mix tracks where you're not pushing things too hot onthe analog side (or have a lot of compression going in), that number can go as high as -6dBFS. On projects where those variables are flipped around in a major way, it can go to -14dBFS or lower. Go haead and start with Clone's figures, they should work fine for you and at least not cuase you any major problems off the bat. As you get more involved and experiences, you might find yourself playing a little with that number is all I'm saying.

Um, actually the tracking room and the control room is very far away from each other (practically the other side of the building - 20 meters, I'd say). So talkback is definitely a must. I'm using very high-grade cables though so signal-loss isn't a problem. Oh and there's a two-way CCTV too so me and the musician(s) can actually see each other.

You are correct that I am taking an unecessary long route to accomplish a simple task, but the talkback capability and MIDI playback are key reasons why I have to do what I am doing now. I simply don't have enough channels on my mixer, and I do a lot of live tracking. Not every band can record single instruments separately...

I have to apologize for still missing something in the equation and perhaps barking up the wrong tree because of it. I thought you were initially asking about the control room/mixing monitors and they were plugged into mixer 15/16. But it now sounds like you're talking about the live room playback monitors instead. In which case I have no idea where the C/R monitors are in the grand scheme of things. So I'll leave that one alone for the time being instead of confusing you with the erroneous stuff. Sorry 'bout that.

Oh and before I go, a quick unrelated question: I'm trying to test the bass in my control room to make sure it's at a level that's pleasing enough. I don't have money to invest on expensive bass traps and acoustic foam. Currently I'm using the egg container-things to reduce mid and high reflections. Not sure what to do with the bass, though! Anyway what Hz test tone do you use to test bass?

Also, I'm using Alesis M1 active nearfield monitors, arranged horizontally, and the tweeters are in the inside (subwoofers outside). The bass is too loud when I'm sitting close, but taking steps backward (away from the mixing desk) makes the bass disappear slowly. Is this supposed to happen?

There really is no one single frequency of test tone for testing bass, the room responds differently to all frequencies. What you describe about the bass changing volume is fairly typical in untreated rooms. Such rooms are laden with "bass peaks" and "bass nulls" where the volume of the bass changes with the listener's position. It sounds like you have a textbook diagnosis for that and are probably in need of some bass traps, indeed. (BTW, those egg cartons are probably having close to zero effect on your overall acoustics.)

Check in with the "Studio building" forum on this BBS for a tour-bus-full of excellent tips on building bass traps and other acoustical treatment tricks. Look especially for our resident expert, Ethan, who eats, breathes and sweats that stuff.

G.

 

[NashBackslash]

Thanks for the clear answers. I really appreciate it.

One quick question: What is "unity gain"?

 

[SouthSIDE Glen]

What is "unity gain"?

That's the black-tie term for "no boost, no cut" . Specifically it means, for example, on a mixer fader the setting about 3/4 of the way up that is labeled "0", which is the point at which you are neither boosting or cutting the volume.

G.

 

[iwantmypie]

Ok now I'm having problems with this whole thing. On my analog mixer (Behringer UB802, yeah i know) I have the channel set to "0" and the master set to "0". Then I take the trim pot and adjusted it to just before it starts going crazy red. Problem here is that for me to read around -14 to -9 on my programs digital mixer (n-track) I have to practically turn my soundcards input all the way down (SB Live!, yeah i know AGAIN). My biggest area of confusion is the setting of the soundcard inputs. I've read this article here that says to set you soundcards level to the max in windows control panel.

http://www.ethanwiner.com/mixer2daw.html

Can anyone help me cus I feel like I'm missing something. Hopefully this changes when I upgrage my card shortly.

 

[DonF]

I take the trim pot and adjusted it to just before it starts going crazy red.

I think I'd make a gain cut right there at the trim pot. You do have the mixer connected to the LINE input of the soundcard, right? The Behri's output is at 0dBu when the meters read zero; that's probably hotter than the soundcard's input expects. You might have to pull the main fader down a bit to compensate.

 

[iwantmypie]

yes I am going into the soundcard's LINE IN. Is that whole thing about keeping the soundcards input at the max a good idea?

 

[NashBackslash]

iwantmypie, it could be that your UB802's output are of a different level. Check your Behringer manual.

Does it say +4dB, or -10dB? Ideally, the outputs from your Behringer that go into the Line In of your SB Live soundcard should be -10dB.

"Is that whole thing about keeping the soundcards input at the max a good idea?"

With your current setup, I think what you'd want to do to have the cleanest possible sound is to set your soundcard's line in input to max, set the channel strip on your Behringer to 0, adjust the trim pot so that it doesn't clip, then pull down the "master" fader on your Behringer until you get the cleanest reading possible.

Try it!

 

[xstatic]

Most companies reccomend using somewhere between -15 and -20 dbfs as a "unity" scale measurement. The problem here is that a lot for your cheaper mixers do not have any sort of calibration options. For a quick down and dirty level match, you could try this.....

1. Send a 1 khz signal through your analog channel. Turn the signal up so that the channel meter (you may have to PFL that channels to see a meter depending on what mixer you are using) reads 0. This must be done with the channel out put (or whatever controls the direct output) at unity, or "0".

2. Take a look at your converters. What is that channel reading on them? Assuming you have no way besides the analog channel to change your gain structure at the converters, whatever your converter is reading is your "unity" gain setting.

3. Now that you know where the "unity" setting is in your system, when you are setting levels, set each track so that the average signal hovers around that level.

If you follow this approach you should generally have no real problems with clipping either you converters or your analog channels. At the same time you will have a decent input signal. For some tracks (that lack in extreme dynamics like maybe a heavily distorted guitar) you cna get away with setting an even higher level. Be aware though that if you do this be raising the gain or trim that you run the risk of adding some more distortion from the analog channels preamp. On a heavy guitar track this often goes unnoticed until the track has already been laid and it is too late. Depending on your preamp, this artifact may be pretty disappointing, or it may even be pleasing. I do this on purpose occasionally with my Chandler preamps, and sometimes with a DBX 165a compressor specifically for that effect.

If you do happen to have a nicer console that is calibrated, than you would run the 1khz sine wave through the channel and meter the direct output ( or group output if you have something like a 24 bus console) and you would set your meter to 0 when the ouput equals 1.23 volts. Now you have a more properly matched system.

 

[iwantmypie]

iwantmypie, it could be that your UB802's output are of a different level. Check your Behringer manual.

Does it say +4dB, or -10dB? Ideally, the outputs from your Behringer that go into the Line In of your SB Live soundcard should be -10dB.

"Is that whole thing about keeping the soundcards input at the max a good idea?"

With your current setup, I think what you'd want to do to have the cleanest possible sound is to set your soundcard's line in input to max, set the channel strip on your Behringer to 0, adjust the trim pot so that it doesn't clip, then pull down the "master" fader on your Behringer until you get the cleanest reading possible.

Try it!

BOOYA! That worked for me. I run the signal out of the mixers TAPE OUT which is at -10dB. I believe the main outs go at +4dB. Thanks a lot man!