Is there a signal to noise ratio problem in amp sims ?

[Logicno8]

I just want to check with you guys am I right on this one.

So....signal to noise ratio. As far as I know if we record external sound source in a room (acoustic guitar, bass amp, electric guitar, singer, trumpet...whatever) we need to achieve "healthy", firm signal (not hot of course) so we can later manipulate it in our DAW. I won't talk about levels now since there is so much about that already talked on the forum. What I want to discuss about is signal to noise ratio in a "real world" versus DI recordings.

What I mean by that is....if we record external sound source with a mic (SM57 on a Marshall amp, just an example) and we record it at a low level, than later on in the mix we decide to boost that signal which is too low to cut through the mix (so we add gain, normalize, EQ, eventually bring up the fader, whatever works for us) and we can probably notice that we brought up something beside our sound and that is the noise, hum, hiss....who knows what. Maybe some air conditioner room next to us, maybe some audible rumble (which wasn't present so clearly if at all, when we were recording), maybe some mic stand movement....so, in conclusion we have recorded this with our highest peak at, say -20dB'sFS or -17dB'sFS and now we boosted it up to -5dB'sFS and we experience problems with signal to noise ratio. My question is : IS this possible when recording DI ?

Example : I plugged my guitar into my audio interface, I put Guitar Rig as an insert on my audio channel, I have everything set and now for headroom sake I leave level of -18dB'sFS or -15dB'FS to be my highest peak. And I know I can do whatever I want since "the room" is virtual as my "amp" and "mic" inside that "room".

Can I feel free to leave more headroom and carry less about signal to noise ratio in a amp simulators such as Guitar Rig and Amplitube ?

 

Thanks.

[Matt Mayfield]

Recording with Mic:

 

Air conditioner at 50 dB SPL.

Guitar amp, at quiet volume, at 70 dB SPL.

 

Your recorded guitar track has a 20 dB signal to noise ratio. Doesn't matter if you turn it up or not.

 

Keep the track low in the mix - guitars at -20dB FS, air conditioner at -40dB FS.

Turn the track up - guitars at 0dB FS, air conditioner at -20dB FS.

 

Same signal to noise ratio: 20 dB.

 

Only way to improve this is to turn off the A.C. or turn up the amp. Say your neighbors don't mind noise. Then:

 

Air conditioner at 50 dB SPL

Guitar amp at 100 dB SPL.

 

Record this, and then your signal to noise ratio is 50 dB:

Keep the track low in the mix - guitars at -20dB FS, air conditioner at -70dB FS.

Turn the track up - guitars at 0dB FS, air conditioner at -50dB FS.

 

Think about what this implies for the amp simulator and you will have your answer.

[fuzzfilth]

The question is like 'is there a distortion problem with guitar amps ?'

 

Of course there is. If you crank either up to get a juicy sound it will a) distort and b) hiss

 

However, an amp sim does not necessarily hiss by itself once it's turned up. If you mute the input signal (turn down the volume pot of the guitar, use a volume pedal upfront, mute the input of the interface, use a noise gate on the input of the amp sim, it usually goes dead quiet, which is not at all what a cranked up guitar amp does.

 

But while playing, hiss can be a problem. During notes it is usually masked by the tone itself, but in the later decay phase of a note, it fades pretty quickly into a loud noise floor. Using the noise gate or expander as mentioned can help on this.

 

Christian

[Logicno8]

Matt Mayfield...so basically when my recorded guitar (or anything) has 20dB's signal to noise ratio, it means that my recording source is 20dB's louder than any background noise, hum or hiss ?

I understood you mic examples (real amp in a real room recording) but I'm not sure how to translate this to an amp sim world where conditions are meant to be perfect. Which means that I don't think I can bring up noise level as much by bringing gain up. More a pure tone itself.

 

fuzzfilth...No, you didn't understand. Maybe I explained it bad. I KNOW there is all kind of mumbling sounds coming out of a big amp with a lot of gain but

I wish to understand if boosting up GuitarRig amp sim guitar signal, (adding gain) which isn't present in the mix, will make me benefit with less signal to noise ratio than real guitar amp recorded in a real room.

Because I am pretty much in control over any amp simulator (just as with using virtual instruments) opposite to real life where we're not.

[ski]

IMO, Logic Guy, you worry about this stuff waaaaaaaay too much. You can't make music by the numbers. And there aren't that many recording techniques where you can do actual math to predict a particular outcome.

 

Aside from mastering engineers who need to see fairly precise levels, most recording engineers used their ears to get a good sound. I never saw anyone, EVER, take out the calculator and say, "hmmm... the ambient noise is at -50 dB on the meter, so if I get the guitar player to lower the gain and raise the master by exactly (tap tap tap on the calculator) 3.7 dB I'll get a little bit better signal to noise ratio..."

 

No. In this case, the engineer might ask the guitarist to remember their settings but turn down the gain and raise the master. They'll try that. If it doesn't work, they put the amp back to where it was and try other solutions.

 

You make music by starting with the sound that you like and composing the part. Then, when it's time to record you put on your engineer's hat and listen critically to the sound. Heck, do a test recording! And if you find that the sound is too noisy you start to make adjustments. You want to change the sound as little as possible but reduce the noise. Of course, save your settings first. You can TRY and adjust your amp sim to get exactly the same sound with less noise. If you can then that's great. But if you can't achieve that, your next step is to put a noise gate on the signal and set the controls -- by ear -- until you have a sound where the guitar comes through nicely and isn't chopped to hell by the noise gate.

 

Unless you have the crappiest signal in the world with hiss, hum, and buzz, chances are that you'll be able to get a perfectly acceptable sound without hearing noise in the background (when you're not playing, or during rests). That's what noise gates are for.

 

But if you have a really crappy signal then you have to start fixing the problem at the source. Is there some kind of crap coming from the guitar that's being amplified by the amp sim? If so, fix the problem at the source. Ground lift, new cable, reposition yourself to cancel out induced hum, whatever. Then put your amp sim back in line. Does it sound better? Great! Still a little noisy? I gave you advice above on what you can try to fix the problem.

 

Notice that nowhere above did I mention anything about levels or math. That's because it's just not necessary. Forget about levels, forget about math. Use your ears and your judgement and move on. That's my sincere advice.

[Logicno8]

IMO, Logic Guy, you worry about this stuff waaaaaaaay too much. You can't make music by the numbers. And there aren't that many recording techniques where you can do actual math to predict a particular outcome.

 

Aside from mastering engineers who need to see fairly precise levels, most recording engineers used their ears to get a good sound. I never saw anyone, EVER, take out the calculator and say, "hmmm... the ambient noise is at -50 dB on the meter, so if I get the guitar player to lower the gain and raise the master by exactly (tap tap tap on the calculator) 3.7 dB I'll get a little bit better signal to noise ratio..."

 

No. In this case, the engineer might ask the guitarist to remember their settings but turn down the gain and raise the master. They'll try that. If it doesn't work, they put the amp back to where it was and try other solutions.

 

You make music by starting with the sound that you like and composing the part. Then, when it's time to record you put on your engineer's hat and listen critically to the sound. Heck, do a test recording! And if you find that the sound is too noisy you start to make adjustments. You want to change the sound as little as possible but reduce the noise. Of course, save your settings first. You can TRY and adjust your amp sim to get exactly the same sound with less noise. If you can then that's great. But if you can't achieve that, your next step is to put a noise gate on the signal and set the controls -- by ear -- until you have a sound where the guitar comes through nicely and isn't chopped to hell by the noise gate.

 

Unless you have the crappiest signal in the world with hiss, hum, and buzz, chances are that you'll be able to get a perfectly acceptable sound without hearing noise in the background (when you're not playing, or during rests). That's what noise gates are for.

 

But if you have a really crappy signal then you have to start fixing the problem at the source. Is there some kind of crap coming from the guitar that's being amplified by the amp sim? If so, fix the problem at the source. Ground lift, new cable, reposition yourself to cancel out induced hum, whatever. Then put your amp sim back in line. Does it sound better? Great! Still a little noisy? I gave you advice above on what you can try to fix the problem.

 

Notice that nowhere above did I mention anything about levels or math. That's because it's just not necessary. Forget about levels, forget about math. Use your ears and your judgement and move on. That's my sincere advice.

 

 

Wise words. I became obsessed with getting best the possible sound I can with my "on a budget studio" that I really DO care too much about some unnecessary things. Thanks for advising me wisely SKI.

[Rev. Juda Sleaze]

And don't forget the "strip silence" feature in Logic. It is the best way I've found to get rid of hum and buzz between notes, and doesn't have the risk of making it sound weird like gating can.

[Matt Mayfield]

Matt Mayfield...so basically when my recorded guitar (or anything) has 20dB's signal to noise ratio, it means that my recording source is 20dB's louder than any background noise, hum or hiss ?

I understood you mic examples (real amp in a real room recording) but I'm not sure how to translate this to an amp sim world where conditions are meant to be perfect. Which means that I don't think I can bring up noise level as much by bringing gain up. More a pure tone itself.

 

Well, I didn't mean to draw emphasis to the numbers; I was just using them to illustrate the point that wherever you have noise, it accumulates, and you have to get a cleaner sound earlier in the chain.

 

With an amp sim, all the noise is going to come through your guitar's connection to the interface... and be amplified by the amp simulator. So get the cleanest sound you reasonably can out of your guitar with as little buzz as possible, and there shouldn't be any problem with noise in the mix.