Amp sim noise

[jthzp9]

Hi all,
So I consider myself a newbie at recording and was looking for some feedback regarding guitar amps and amp sims.

My set up:
Guitar > warm audio active DI > focusrite Scarlett 18i8 (2nd gen) > Mac mini 2018

My question:
When I’m playing my guitar into my guitar amp and and have a fair amount of gain dialed in I like to roll back the volume on my guitar so I can have a clean tone. The amp cleans up, volume gets a little quieter and the noise floor also decreases. If I were to turn the volume all the way down, put the guitar in a hum cancelling pup position or unplug the guitar, the noise floor is barely noticeable (even with the amp gain/vol settings loud enough that would be almost uncomfortably loud!)

However... when I use the same guitar, plugged into the recording set up described above, There is always noise when I use amp sims! With the guitar unplugged or volume all the way down and plugged in there is always an unusable noise floor present. This is even present at conservative levels of gain and volume dialed in on the amp sim.

I continue to read about proponents of amp sims being superior in that they impart less noise than a real amp. However, in my experience thus far, this is just simply not true. Whether it is due to the interface, computer or usb protocol I am not sure. What I do know is, it is unacceptably noisy in my opinion.

I know that the focusrite Scarlett is not exactly the top of the line audio interface, but would upgrading to something like an RME really make that big of a difference? Obviously another take is just to use a real amp, but I do enjoy the convenience of amp sims for quiet recording and would like to be able to use amp sims when necessary.

Any suggestions would be much appreciated!!

[CS70]

Well, they are two completely different signal paths, so it's not really about amplifiers and sims in general, but about your specific setup.

Have you tried to plug the guitar directly in the interface? In most interfaces at least one preamp doubles as an instrument in.

I guess you're plugging the DI into a mic preamp? If you do, you're taking your perfectly fine instrument input, bring it down to mic level and then re-amplifying it with the interface preamp to get it to a reasonable line level. What's the reasoning?

[forumuser931182]

It does seem that the warm audio di is unnecessary and could actually add a small amount of noise.
The noise you mention however is probably from the amp sim you are using. If you record the guitar directly into your computer ( try without the warm audio ) without feeding through an amp sim is the noise level ok? If so try playing with the amp sim controls when it is a track effect and see if you can get the sound you're after without too much noise. Maybe try other amp sim effects as well and maybe noise gates.

[jthzp9]

Plugging directly in to the hi-z input of the Scarlett does not yield any better results. This ultimately makes me think that it is the noise floor of the AI that I’m hearing because it is apparent even when the guitar is unplugged or plugged in and volume down.

I originally put the DI in place because it gave me the ability to know that I had a reliable 1M ohm impedance for my DI signals. I’ve also always read that passive pickup instruments sound better with active DIs. I owned an active guitar and agreed that it sounded better than my passive pickups when DI’ed so assumed this must be true.

Lastly, I’ve also read that if you are at mic level that the signal is transferred over longer lengths with less signal degradation due to the impedance matching on the front end of the DI and then the signal being dropped with a lower impedance needed at the preamp. This was slightly experimental on my behalf to see if it helped with lowering the noise floor of my AI due to needing to turn up the preamp less. I’ve noticed little to no difference in the noise floor but do think it sounds better than direct to the interface itself.

I will post some clips of the noise I’m referring to tonight.

[CS70]

Plugging directly in to the hi-z input of the Scarlett does not yield any better results. This ultimately makes me think that it is the noise floor of the AI that I’m hearing because it is apparent even when the guitar is unplugged or plugged in and volume down.

Could be for sure - something in your cabling or in the interface itself picking up noise, the computer itself, grounding issues etc. The electronic experts in the forum will hopefully come in but in the meantime you could try to borrow another interface and check. Or - aren't these Scarletts capable to work in standalone mode? Bit late now but it could be possible to devise some experiment that eliminates the computer and would allow you to detect a noisy singal out of the interface...

Could it also be the specific amp sims? Do you get the noise even with a "clean" low gain amp emulation?

[Sam Spoons]

It sounds like it may be a gain structure issue to me, the Scarlett is capable of very good noise figures so I'd be surprised if that was the problem on it's own. Some audio files would definitely help identifying where the noise is originating, it's not always possible to tell but many noises have an identifiable character.

You seem to have eliminated the possibility as you say "This ultimately makes me think that it is the noise floor of the AI that I’m hearing because it is apparent even when the guitar is unplugged or plugged in and volume down." but electric guitars, especially single coil equipped electric guitars, are notoriously noisy beasts and any gain you apply, wherever it comes from will increase that noise

[Drew Stephenson]

Any difference in the noise when you touch or release the strings? I'm just conscious you've not got a proper ground anywhere in your recording chain but you probably do when using your amp.

[jthzp9]

Ok here is a couple clips of the signal and noise I'm getting.

https://soundcloud.com/user-500797446/d ... m-audio-di

There is a two bar click in between the clips, please keep in mind the last two are much louder than the first four!

Clip 1: Scarlett (no amp sim)
Clip 2: Warm audio (no amp sim)
Clip 3: Scarlett (amp sim no playing)
Clip 4: Warm audio (amp sim no playing)

*** PLEASE BE MINDFUL OF YOUR LEVELS FOR THE NEXT CLIPS, AMP SIM AT RECORDING LEVEL!!!***

Clip 5: Scarlett (amp sim, playing)
Clip 6: Warm audio (amp sim, playing)

The levels were set for both so that the highest peaks are coming in at -6dbfs on the way in, before the amp sim.

Amp sim is Scuffham 2 S-Gear.

The guitar is a HSS strat. My frustration is not the noise that you can hear as I flick through the different pickup configurations, but the white noise that is prevalent throughout. That does not change is ultimately the reason I'm not happy with the amp sims. If I had normal amp noise, it wouldn't bother me as much. The never ending "digital ocean" sound is maddening to say the least.

I can record clips of the same guitar through my amp and that noise is not there when I turn down the guitar, it is completely quiet save a little warm hum of the amp in the room that is barely even perceivable.

[jthzp9]

To give a little more info on my system.
I use a panamax M5300-PM as a power conditioner. All of my gear is hooked to it and it is running to the wall outlet that I have checked with an outlet tester and says that it is a grounded outlet.

Computer, monitor, audio interface are all going to the power conditioner.

[forumuser931182]

Regarding warm audio 1 megohm input, the focusrite should also exhibit this when the input is set to "instrument" in the software control panel. Also a balanced cable/input will be less prone to hum/noise however as you are using a short lead from guitar to audio interface you shouldn't have to many issues with an unbalanced lead.

I tried the Schuffman amp sim as a send effect in my daw and didn't get any noise when I brought the input signal down. I tried this with various presets. Do you get the noise with presets or just patches that you have set up yourself.
If you record the guitar straight and then apply the Schuffman sim as a post effect do you get the same result?

[forumuser931182]

Actually tried it live and I noticed the noise. Is it just because the amps are all set with quite high gain. First preset American Clean has noise but Clean Duke which is a different amp has less noise.
The built-in noise gate seems to be post-amp? Maybe try a separate noise gate plugin that is situated before the Schuffman plugin.
I won't post anymore, I'm not even a guitarist haha.

[CS70]

Ok here is a couple clips of the signal and noise I'm getting.

To my ears, that could be combo of a really pushed gain stage mixed with some interference from the computer equipment. There's a dodgy connection somewhere (you can hear the crackling), possibly the cabling.

These are the classic things that are easy to debug if you are in the room and hard to do remotely, as there's a lot of guess-try-and-see involved. I'd ditch the DI for the moment to eliminate one factor, and start debugging with just guitar and interface.

Some possibilities:

1) Sam's idea on the gain structure is definitely an option. Are you positive the jack sockets you're using are in "instrument" mode? If they were set to line, you would have to push the gain a lot in software to hear anything and that would lead to exactly that kind of "noise even if you aren't playing anything" - because you would amplify a lot whatever tiny amount of noise there is.

2) Have u tried using a different cable? I've experienced that the mechanical tolerances can be so that a cable's jack plug works perfectly in a socket and is noisy in another. If the amp's socket is just a little bit smaller than the interface's that could lead to a solid contact in the first and not the second.

3) Also something that's different between playing the amp and playing the sim is that in the latter case you are probably nearer to the computer - especially the computer screen. However, it does not sound like typical RFI to me (well, a little bit mixed in but the bulk of the noise sound like very high gain).. but it could be worth moving around.

[jthzp9]

This noise occurs without the guitar even being plugged in or with the guitar volume all the way down. While you can obviously hear some RFI in some of the clips I posted, that is even lower noise than the “digital woosh” noise That is ever present.

Even at fairly conservative levels when I setup the amp sims (s gear is not the only one that does this) you can hear the noise.

Absolutely positive that the input is set to hi-z instrument.

I use a mogami neutrik connection TS 18” guitar cable. I’ve used others and this is by far the best cable I own. No noise on other guitar amps, and really no noise with the audio interface DI. Again, just my two cents but it does seem to be a noise floor issue.

If this is true, what are my options?

[merlyn]

While you can obviously hear some RFI in some of the clips I posted, that is even lower noise than the “digital woosh” noise That is ever present.

You seem to have a bit of a misunderstanding here. Fundamentally digital audio is noiseless -- if you put in a stream of zeroes (silence) you get a stream of zeroes out.

The "whoosh" is thermal noise, and that comes from analogue electronics. Every resistor adds a bit of noise -- that is completely unavoidable. The higher the resistance the more noise so I think you will find using high Z inputs adds more noise.

I've got an amp sim -- Guitarix and it does have noise. The more gain is used the more noise. But this is not coming from inside the plugin -- it's the analogue noise on the input that's being amplified. If I disconnect Guitarix from the hardware inputs there is no noise. Although amp sims are digital models of analogue gear it would be a pretty nutty developer who included noise in the emulation to get that real analogue feel.

[jthzp9]

Ok, so this is not my DI track but was made for folks trying to make kemper profiles. It sounds like it is an active guitar with humbuckers into some bit of outboard gear. I don't know this to be true as there is no info to go along with the track but just my assumption.

https://soundcloud.com/user-500797446/di-guitar-example

Clip 1: clean DI
Clip 2: SGear 2 (everything at noon except the input gain which was clippping the plugin turned down slightly, signal is not clipping the output meter) > NAD IR impulse response

Even as great as the DI track sounds, the same noise is still there. So not sure what to think.

[jthzp9]

https://soundcloud.com/user-500797446/a ... e-no-input

This is also what makes me believe this is the noise of my interface.
There is no cable connected to 2nd preamp of my interface. Hi-z is selected.

Clip 1: interface hi-z input
Clip 2: interface hi-z input with sgear applied as plugin.

There is nothing connected, so this is just the noise being generated by the interface itself. This is the sound that I'm trying to get rid of.

This is also the same exact noise I'm hearing when I have my guitar plugged in and volume off. It is present throughout the guitar track when recording which I don't like.

Again, comparing to a real guitar amp, when the guitar volume is off, it's quiet, even with enough gain on tap to be in crunch/rock territory.

[Drew Stephenson]

Apologies if you've covered this already, but if you take your current set up and record so that your peak signals are at a healthy -10dB and the average is burbling along around -18, what level is this background noise showing on the meters?

[James Perrett]

Recording your Soundcloud link into Reaper gives me an RMS noise level of around -93dBFS with no effects and -83dBFS with the amp sim enabled. I wouldn't call these levels excessive.

If I'm wrong it might help if you could provide a downloadable link so that we can see the real numbers.

[merlyn]

Again, comparing to a real guitar amp, when the guitar volume is off, it's quiet, even with enough gain on tap to be in crunch/rock territory.

Something to bear in mind is that the speaker in a guitar amp doesn't reproduce much above 5k. To do a like for like test you could take the output of your interface into the guitar amp with a clean setting or straight into the power amp if that's possible and you know how.

It's the high frequency noise that is particularly noticeable/annoying so you could look at a cabinet emulator with a 5k rolloff or tweak the HF and presence settings on the virtual power amp.

[jthzp9]

So I had a chance to do some additional tests tonight and just wanted a bit more feedback from folks.

I have come to the conclusion that the noise I'm hearing is either the preamp noise or the converter noise of my audio interface. Perhaps it is just me, but I find this noise extremely annoying.

I did a quick test between my Hi-z input on my scarlett interface vs using the fx send from my tube amp into the line input of the interface and the noise ratio is far more acceptable to me. Furthermore, the noise is acceptably dropped by turning down my guitar volume. Where as when using the DI, the audio interface noise is at an unacceptable level.

I'm just curious if there is a reason for this? I guess I'm just not meant to jive with amp sims which is too bad because I like the idea of them, but I just can't handle listening to the underlying noise floor.

Here is a short example of what I'm talking about. In the first portion I'm using my amp into the audio interface. When the guitar volume is down, no noise is barely perceivable. Then you can hear some 60 cycle hum from the single coil pickups when not in hum cancelling positions. To me this is just a natural sound of a guitar and it doesn't bother me. The 2nd clip is the same guitar, through my audio interface and the noise floor is just unacceptably loud to me. In the beginning the guitar volume is all the way down and yet the noise is annoying and loud in my opinion. And when I go to turn down the guitar volume to clean up the signal like I do with the real amp, the tone is gone and the noise is again too loud to be useable.

https://soundcloud.com/user-500797446/amp-vs-amp-sim

[ef37a]

I felt that the noise to riff ratio in the first clip very acceptable. The hum would bother me more.

I have looked up the noise spec of a 5watt valve guitar amplifier*.

With ALL controls at maximum and a shorting jack in the input the noise at the 8 Ohms speaker jack for a clean setting is -46dBV or better. For the overdrive setting it gets much worse at -15dBV (n.b. guitar amp tech s do not use an outdated measuring system based of over a hundred years old telephone technology!)

It is likely therefore that any amp sim' is simply going to boost system noise quite a very lot.

*And that is one of THE quietest guitar amps you will find.

Dave.

[Wonks]

On almost all guitar amps, the input is taken to ground with no jack inserted. On an audio interface or DI this is not the case, so with nothing inserted, the wiring is able to pick up electrical noise that is then amplified through the system and any software that adds gain.

If you want to find out what your true audio interface background noise level is (and it should be very low if it's not faulty), they you really need to insert as jack plug into the AI (in Hi-Z mode) with the tip and sleeve connected together with a 470k resistor (which will simulate the DC resistance of an 8k humbucker and 500k volume pot in parallel). You could also just short to ground (which is what a guitar volume pot set to 0 would do) but the resistor is a bit more real world 'ideal guitar plugged-in but not playing' scenario. Any noise you then hear on that channel will be the result of the noise picked up or generated within the audio interface itself with no contributions from the

Turn down the gain on any unused inputs and make sure that the AIs' and DAW's own audio routing setups aren't mixing in any signals from unconnected inputs to that channel.

Set the input channel gain to a value you'd use with a guitar plugged in. Record the noise onto a plain audio track with no plugs and no gain applied to the channel input in software. That will be the audio interface noise level you are recording and which is then getting amplified.

Then simply remove the shorting plug, plug in the guitar with the volume knob up full, position yourself for minimum hum with your hands on the strings so you are grounded and record the noise level again. You should notice a substantial increase in level. You'll need to tell us how much, it increases by, but I'd suggest that it will be a considerable increase. And it's that noise which is then being amplified by your amp sim plugs.

Which doesn't really help you in a practical sense, but it might help you understand where the noise is coming from, which is very unlikely to be the audio interface itself.

My HF hearing is completely shot, and I simply can't hear any real noise in the clips, though I can see the waveform on the screen, but it does also indicate that the noise is almost all at frequencies that a normal guitar speaker would filter out. So are the cab sims working properly?

An amp sim isn't a valve amp, it hasn't got the same input circuitry, so it won't react like a valve amp when you turn your guitar's volume down. It's got a high input impedance so that it won't load a passive the guitar pickup and roll off a lot of the treble, but that's as far as it goes. The rest is down to the amp software and how well it simulates the amps it's modelling. And that model will be based on getting an input signal that's equivalent to the signal level going into a real amp. Presumably the amp sim manual tells you at what level that should be, so try and get that level feeding into the amp sim. You may need to add a simple gain plug in to set that level if your DAW doesn't have an input gain control on the channel input. If it's supposed to be at a maximum of say -20dBFS, then try and get the signal peaking at that. If the input signal is too low or too high, then you'll get a different feel to a real amp. I've messed about with amp sims on previously recorded guitar tracks, but never tried to use them live. By which time I've normally gone and edited the guitar tracks to cut out noise between sections.

But I do have quite a few amps (Two Rock Studio Pro22, Fender 68 Custom Deluxe Reverb, handwired Fender 5E5A tweed Super style combo, Sound City 50 Plus head with the Hiwatt passive EQ mods, Laney Ironheart IRT30 combo, Orange Tiny Terror, Fender Vibro Champ and a Blackstar HT5) and none of them are particularly quiet, especially with the gain/volume turned up. In my real world, I get a lot more hum and hiss from them than you are getting with the Scuffham software. But if recording, the moment they get slotted into a track, the noise magically gets hidden by the bass and drums and vocals. Or at least it sounds like that to me.

[CS70]

But if recording, the moment they get slotted into a track, the noise magically gets hidden by the bass and drums and vocals. Or at least it sounds like that to me.

+1

[jthzp9]

So I finally was able to come up with a better way hopefully to share what I was experiencing with my audio interface. I'm still having the same noise problem and I hope this helps give a better understanding of what I'm dealing with.

https://www.dropbox.com/sh/qnjd81ohvzqm ... 9CrBa?dl=0

In the drop box link I have 3 files. One of my Soldano SLO 30 running it's power amp out into my Focusrite scarlet 18i8 2nd gen. The second is my guitar (Fender Strat HSS) directly into the Hi-Z input of the audio interface (instrument input selected). The third is the DI take clean without the amp sim processing.

My problem with what I'm experiencing is not so much the tone, but the noise that is super noticeable before the guitar kicks in and after. This is barely noticeable at all with the Soldano SLO 30 and there is really no noise floor to speak of when I turn the guitar volume off.

I'm curious if a better interface would help diminish this noise floor that is being magnified by the amp sim. I just can't stand that "shhhhhhhhh" sound being there in all of my recordings. It kills the ability to let notes ring out and I find it super unmusicial.

Hopefully someone might have some suggestions.

A few notes about the recordings:
The amp sim and real amp are both going through the same impulse response which is a Marshall 1960A by Emissary.
No silencing or noise suppressors are used on any of the tracks.
It is the same guitar, with the same pickup selection.
Same guitar cable. An additional cable is run from the power amp out of the Soldano to the interface line in (without the preamps).

[jthzp9]

Anybody have Any thoughts regarding the sample clips?