Levels hitting the amp (in dBu) when reamping and/or capturing

[Richard_G]

I'm curious what levels (in dBu) you guys are aiming for when reamping and/or capturing; what's your typical level of the signal going into the amp input?

Intuitively I think it would make sense to aim for dBU levels in the same ballpark that your guitars put out. On the other hand, I've seen people argue for having hotter signals than that (very much so) - at least when capturing. One argument seem to be to capture how the amp reacts when boosted etc. If you are in this camp; what levels (in dBu) are you hitting your amp with and are you worried about putting a too hot signal into the amp?

 

[LP59]

For reamping, the same level as your guitar is the right answer. Otherwise it won't sound the same as if you were playing your guitar through that rig. For capturing, it's complicated .

 

[Richard_G]

Intuitively, I totally agree.

A complicating factor may be the difference between the output level in dBu you get when doing this:

)

.. and the dBu level the actual DI and/or test file used during capturing (provided by NAM, Tonex etc.) will end up as in the same setup?

 

[Richard_G]

Let's take these nice NAM captures of a PRS Archon Classic from @2dor as an example: https://www.tone3000.com/tones/prs-archon-classic-50-6ca7-30381

In the description he writes:

"Reamp chain details
• RME Fireface UCX II -> Lehle P-SPLIT III -> PRS Archon CLASSIC 50 (6CA7) -> St.Rock React:IR II -> RME Fireface UCX II
• SEND level: 18.995 dBu
• RETURN level: 13.195 dBu"

Perhaps I'm misunderstanding something fundamental here. But isn't this a very hot signal to put into an amp?

 

[Chocol8]

That's hotter than most active pickups are going to get!!!!

 

[Richard_G]

In my own setup I'm going from the line out of a Motu M4 into a Lehle P-Split III, giving a dBU level of 14,8 dBu when following the simple @MirrorProfiles procedure in the linked video. I'm a bit worried this is too hot for running into pedals and or amps for reamping and/or capturing? But then again; at what level is the actual DI and/or training file ending up at...

 

[Richard_G]

An output of 14.8 dBu is like over 4 volts right? I'm pretty sure my PAF-like humbuckers don't deliver that. So if the actual DI and/or training file is close to 0 dBFS in the digital domain, there is a reason for me worrying about hitting the amp/pedals with a too hot signal in the analog domain?

 

[tekbow]

This is why I like the Two Notes Torpedo Reload so much.

The match function was amazing and they should have released a standalone version because its gone from the Reload 2

 

[MirrorProfiles]

In my own setup I'm going from the line out of a Motu M4 into a Lehle P-Split III, giving a dBU level of 14,8 dBu when following the simple @MirrorProfiles procedure in the linked video. I'm a bit worried this is too hot for running into pedals and or amps for reamping and/or capturing? But then again; at what level is the actual DI and/or training file ending up at...

Absolutely fine to do that, and honestly, I’d recommend using a level like that (slightly louder than what your guitar outputs) so that the training has examples of the amp being pushed hard. This’ll make it more accurate in more situations (like using different guitars or boosting with a pedal). FWIW, a TC integrated preamp (or peppers dirty tree) adds over 30dB of level. So I wouldn’t worry about hitting an amp 5-10dB higher than a pickup might output.

Much better to train with a loud signal, rather than too weak. The reamp file will have a wide range of dynamics already - the louder you can get it will give the model more information to train from.

 

[2dor]

I usually try to hit the amp as hard as possible while preserving a linear frequency response. Given the NAM training sequence has a lot of dynamic range, the profiles end up behaving much better when I then run a boost pedal / profile in front.
That said, I don't think there's any "correct" value so long as you fill in the metadata in. I see most interfaces hover around the 12 dBu mark so that would be a good enough choice; going higher will help as long as you don't add clipping / saturation by overloading the converters or the reamp box's transformer.

 

[Progressive_disclosure]

I usually try to hit the amp as hard as possible while preserving a linear frequency response. Given the NAM training sequence has a lot of dynamic range, the profiles end up behaving much better when I then run a boost pedal / profile in front.
That said, I don't think there's any "correct" value so long as you fill in the metadata in. I see most interfaces hover around the 12 dBu mark so that would be a good enough choice; going higher will help as long as you don't add clipping / saturation by overloading the converters or the reamp box's transformer.

I’ve been meaning to do a couple of new captures, both with NAM and Tonex. I’ll try blasting the amp a bit harder this time!

Haven’t fully understood what to put into the calibration data yet for NAM though. I’m gonna use my HX Stomp again for it which is I think 12 dBu on the input. How do I get the number for the output? Any hints are welcome! Read through the documentation guide but still don’t get it.

Update: added +6dB on the HX Stomp send to the amp and the captures just finished training. I think it sounds better than my last batch of captures!

 

[2dor]

I’ve been meaning to do a couple of new captures, both with NAM and Tonex. I’ll try blasting the amp a bit harder this time!

Haven’t fully understood what to put into the calibration data yet for NAM though. I’m gonna use my HX Stomp again for it which is I think 12 dBu on the input. How do I get the number for the output? Any hints are welcome! Read through the documentation guide but still don’t get it.

Update: added +6dB on the HX Stomp send to the amp and the captures just finished training. I think it sounds better than my last batch of captures!

These things ought to have all the info you need for shooting calibrated profiles & using them:

Post in thread 'NAM: Neural Amp Modeler' https://thegearforum.com/threads/nam-neural-amp-modeler.1698/post-406047

 

[paisleywookiee]

Someone needs to do a fucking USB box for reamping, where they have all the necessary IO in one (like the ToneX capture), BUT then have the reamp signal selectable for "This is single coils, this is passive humbucker, this is active humbucker".

ALL of this shit is trial and error, and frustrating as fuck. I'm still enjoying messing with NAM and Tonex, but Jesus Harold Christ, it can be frustrating.

 

[MirrorProfiles]

BUT then have the reamp signal selectable for "This is single coils, this is passive humbucker, this is active humbucker".

This part isn’t necessary with calibrated models, it’s basically aiming to make it akin to just plugging your guitar into any amp at its given settings.

If you want less gain you back the models input level down or you choose a model captured at settings that match your guitar and preferences. It removes the wonky guesswork of “is this supposed to be a really high gain capture or an edge of breakup capture”.

 

[bikescene]

I went with 11.5dBu for my rig. My initial motivation was to use my I/O as a hardware insert for pedals into Helix Native. Then I just went with it for NAM.

My line outputs are rated at 18dBu. I used my reamp box’s output to get it to 11.5dBu by following the info in the “Calibrating Input Levels for Plugins” thread. I set my input levels by adjusting my instrument input gain until I got unity gain with the output.

I have run into instances where this level is too low. I tried capturing my BBE Screamer with the level cranked and hit clipping with the NAM training file. I also had trouble with my Naga Viper clone cranked. I also clipped this level when my Duncan Custom pickup was set a bit too high.

 

[Ed DeGenaro]

This all would be simpler if folks would talk in mV instead of dBu.

Put a multi meter on be done.

 

[MirrorProfiles]

This all would be simpler if folks would talk in mV instead of dBu.

Put a multi meter on be done.

Why do you think that would be any easier? It needs to be in reference to 0dBFS so having it relative to decibels makes a lot of sense. It’s like saying Celsius is harder than Fahrenheit.

Regardless, it’s simple to convert between Volts and dBu:

https://www.analog.com/en/resources/interactive-design-tools/dbconvert.html

 

[Ed DeGenaro]

Why do you think that would be any easier? It needs to be in reference to 0dBFS so having it relative to decibels makes a lot of sense. It’s like saying Celsius is harder than Fahrenheit.

Regardless, it’s simple to convert between Volts and dBu:

https://www.analog.com/en/resources/interactive-design-tools/dbconvert.html

Simpler I said if we wanna be pedantic.

If I take a multimeter and read the output of the guitar, and then read what comes out of the reamp to the amp I get an exact readout.
More so than the meters on some of the audio interfaces folks use.

Personally I’d be too lazy for either and ear ball it.

 

[MirrorProfiles]

If I take a multimeter and read the output of the guitar, and then read what comes out of the reamp to the amp I get an exact readout.

Once you have the measurement in Voltage, you're done. Doesn't matter whether it's V, dBV, dBu, dBm etc. dBu is handy for Voltages when they're audio signals but there's no reason it has to be one or the other, they all measure the same thing. I like dBu because Voltage can get sloppy if people aren't clear on peak, RMS or peak to peak etc and also because its easier to think in terms of "I need to add 2dB here, and cut 2dB here". We adjust levels in decibels rather than turning things down by voltage.

More so than the meters on some of the audio interfaces folks use.

The Voltage measured is relative to the dBFS level being sent out, so it's important to check both. A lot of people calibrate their reamp signal by running their reamp out directly to their DI in and they'll adjust levels until the interface meters read the same value. If they know the headroom of the DI, and the reamp matches, then you know it's the same value and you don't need to use a multimeter at all. Same same result.

Personally I’d be too lazy for either and ear ball it.

Probably fine for your own use, but a bit shite if you were to create captures for others to use.

 

[Ed DeGenaro]

Once you have the measurement in Voltage, you're done. Doesn't matter whether it's V, dBV, dBu, dBm etc. dBu is handy for Voltages when they're audio signals but there's no reason it has to be one or the other, they all measure the same thing. I like dBu because Voltage can get sloppy if people aren't clear on peak, RMS or peak to peak etc and also because its easier to think in terms of "I need to add 2dB here, and cut 2dB here". We adjust levels in decibels rather than turning things down by voltage.



The Voltage measured is relative to the dBFS level being sent out, so it's important to check both. A lot of people calibrate their reamp signal by running their reamp out directly to their DI in and they'll adjust levels until the interface meters read the same value. If they know the headroom of the DI, and the reamp matches, then you know it's the same value and you don't need to use a multimeter at all. Same same result.


Probably fine for your own use, but a bit shite if you were to create captures for others to use.

Since I’m not sharing profiles that last part applies.