Input Gain Advice/Question in AT5

[jjguitar]

The best advice I can give is to ask whoever will listen at IK for more information on this topic. It is possible for them to determine and share this, and everyone benefits. No misinformation, no confusion, no guesswork, better tones.

I absolutly agree
IK, please answer this question

[Boogieshoes]

I very much doubt that IK will help here even though this subject probably doesn’t qualify as a ‘tech support’ query. Having a few of the real amps here and wanting to do some ‘Quiet recording’ I have battled with AT5 to try and replicate sounds that I have with the real amps and nothing comes close to giving the dynamics you can readily pull out of the real amp, by using guitar volume/tone controls to get clean up of a fuzz or an amp set at higher gain edge of breakup.

The sounds become flat and lifeless and the pedals, if needed I have to approximate in AT5 because even Max doesn’t have some of them eg. Retrovibe. You cannot simply put the real pedals in front of the audio interface, as has already been discussed trying to get the balance of audio interface settings right such the pedals work in an authentic way or overdrive the selected amps preamp stages is neigh on impossible…. Well it hasn’t worked for me.

At the best, the package is emulating a rig in the recording environment, so I have to accept that the emulated environments will not match the room that I have an amp in. Close mic’ing an amp is still impacted by the room dynamics. Possibly there are better things that could be done by experimenting with the rooms, mic choices and placement and more. At this point the package simply becomes overload and drives ‘analysis paralysis, loss of creativity’ for me anyway.

I hope you find some solutions, and some have said already, there is great deal more that could be done to document the user guide to help users get the best of the product, I am not holding my breath though as I doubt it’ll ever come.

[Matthias Mueller]

I think machinated explained it even further and better than i can.
Thanks for that.

So there is not much to add from my side.

Just want to add this link to his own thread for anyone who is interested to read.
https://www.thegearpage.net/board/index ... d.2482452/
(hope thats okay, that i link it here?!)


I think there are some support tickets open now about this whole topic. Also more and more youtube videos appear about this.

I cant put it in better words than machinated did, so again...

The best advice I can give is to ask whoever will listen at IK for more information on this topic. It is possible for them to determine and share this, and everyone benefits. No misinformation, no confusion, no guesswork, better tones.

Because it really seems that every amp model and pedal is "calibrated" different in Amplitube, this whole 1VAC=0dbFS calibration does only work to a certain degree (if at all).
But maybe you can use it as a starting point till we get information from IK Multimedia. I think its better than nothing...
If you havent already found out, here is another simple way to get calibrated to 1VAC=0dbFS:
- Set your inteface gain to 0
- Look up in the spec list or manual of your Interface how much Headroom your Hi-Z/Instrument Input have (eg. +12dbu (0dbFS) )
- boost this number + 0,79 (in this example 12,79db...) on the input slider in AT or with a gain Plugin before AT.
(- IF, you have problems with Noise you can split it and boost eg. 6db on the Interface/Preamp and the other 6.79 in Amplitube)
- Use this setting regardles witch guitar you plug in.
- From there you can boost more with eg. the slash pedal or the parametric eq in front of the amp if you need to. I dont think there are cases where you want to reduce the input from this startingpoint but...i mean... you can...

[machinated]

I have a strong feeling that 1V peak=0dBFS is used sometimes in Amplitube, but certainly not all the time. I would be surprised if several models didn't use this reference point, but there are significant number which don't. Its hard to recommend values when it could be that, but it could just as feasibly be -3dBu, -6dBu, 7dBu. 10dBu, 12dBu etc. Only IK can say for sure. Sometimes its easier to guess that its closer to say -3dBu than it is to 12dBu but its basically impossible to get anything certain without more information.

Despite what many of us may like the sound of or enjoy using, they may not be accurate at all.

Good tones and how you dial things in is down to personal taste and interpretation, but the internal operating levels of the models in amplitube is not. There is a correct input level that will correspond to the most accurate experience (compared to using the real reference amp).

[Boogieshoes]

Totally agree that Good Tone is a matter of personal taste and that is individual to each and every user, even though many are probably seeking to emulate another inspirational player that they have heard.

The biggest problem for me I think is in fact the connectivity, firstly how to set the interface input gain (I use a Roland Octa-Capture) with Hi Z input for guitar. That may seem easy but of course if you want, as many players do, to play with dynamics by riding the volume control on the guitar vs what should be an amp on the edge of breakup (in my case) and for the pedals in front of the amp to deliver the same as real ones in front of a real amp, that gets really hard to the point I haven't got it to sound right.

Playing the guitar straight into the simulated Amp (no pedals) can get quite good subject to 'Room/Mic settings' which can take time to sort. When pedals get involved it is a lot harder.

I have tried setting interface input gains with pedals in front of that at max impact without clipping of the interface input, but that doesn't translate to what you get with the real amp. Based on your discussion point this is similar but not the same as trying to cope with different guitars with different amps.

How much this is to do with whatever Model simulation calibrations were used on the inputs is frankly an anathema to me. I do expect different guitars to exhibit different characteristics, that is the whole point of switching guitars for double tracking etc. and amps would react differently as result of the change in guitar. I have wasted a lot of time trying things and flipping so many screens that is becomes overwhelming to the point that eventually I feel like I have all but forgotten what I was aiming for when I started.

[dmitch]

There is a correct input level that will correspond to the most accurate experience (compared to using the real reference amp).

I've been thinking about this issue a bit more lately since I posted last week. One thing I keep on coming back to is: what exactly is being asked of IK in terms of documenting the "proper" input level? Suppose you're working with a '65 Twin Reverb sim and you want the "proper" input level for a stock Stratocaster. OK - which Strat? Mine? Yours? My buddy Frank's? They all have slightly different output levels. Or do you want the proper level for an ES335? Again - which one? Or do you want the proper level for a PRS Custom 24?

My point is, all of those present a different level when plugged in to a real Twin. But in every case it will sound like a Twin (a Twin with a Strat, or a Twin with humbuckers, etc.) because it IS a Twin. Different sounds for different levels, sure, but it's a Twin. Would we not expect the same of a totally accurate and faithful simulation of a Twin?

I'm interested in knowing exactly what spec would be desired per this thread, and how that might apply to recording each guitar in my collection (presumably none of which has the same output level as the single guitar spec'd by IK).

I'm not trying argue or belittle anyone's point of view, I'm genuinely trying to understand what is being asked of IK and how that would be useful to me. Thanks.

[machinated]

There is a correct input level that will correspond to the most accurate experience (compared to using the real reference amp).

I've been thinking about this issue a bit more lately since I posted last week. One thing I keep on coming back to is: what exactly is being asked of IK in terms of documenting the "proper" input level? Suppose you're working with a '65 Twin Reverb sim and you want the "proper" input level for a stock Stratocaster. OK - which Strat? Mine? Yours? My buddy Frank's? They all have slightly different output levels. Or do you want the proper level for an ES335? Again - which one? Or do you want the proper level for a PRS Custom 24?

My point is, all of those present a different level when plugged in to a real Twin. But in every case it will sound like a Twin (a Twin with a Strat, or a Twin with humbuckers, etc.) because it IS a Twin. Different sounds for different levels, sure, but it's a Twin. Would we not expect the same of a totally accurate and faithful simulation of a Twin?

I'm interested in knowing exactly what spec would be desired per this thread, and how that might apply to recording each guitar in my collection (presumably none of which has the same output level as the single guitar spec'd by IK).

I'm not trying argue or belittle anyone's point of view, I'm genuinely trying to understand what is being asked of IK and how that would be useful to me. Thanks.

its a perfectly valid question, and its actually EXACTLY the reason I would like to know these specs.

Rather than basing the accuracy of a model from the peak signal from the guitar, its better to look at it from the other angle - from the input of the amplifier.

The amplifier doesn't know what signal is going to hit its input, a quiet signal will get less gain, a louder signal will get more.

This is EXACTLY what we want to get from the plugin. We want it so that your 335 responds to the plugin just as it would if you plugged that very guitar into the real amp. and with the same input settings, it would also give my Telecaster EXACTLY the same sound as plugging my Tele directly into the front of the amp.

If the input gain is set incorrectly, its just like doing the above, but with a volume pedal shoved into the circuit. The goal of this is to remove the volume pedal, so its an accurate response for any guitar, bass, vocal, drum machine etc as plugging those into the real amp.

The difference in volume we are trying to account for isn't because guitars are different, we want to preserve that. Its about removing the variable of everyone (including IK) using totally different converters, with their own definitions of what digital signal level will be produced from the same analog source.

Hope that makes sense, if it doesn't I can try and explain it another way perhaps.

[dmitch]

Hope that makes sense, if it doesn't I can try and explain it another way perhaps.

Excellent - now let's keep going. At multiple places in this thread I've read the request for IK to publish or document A NUMBER that will solve this problem. What is that number? I don't mean what numerical value, I mean what will that number represent?

Bear in mind that every audio interface I've plugged a guitar into has gain or trim control for the guitar input. Those gain controls are generally abstract, they don't have any calibration or units on them. So I'm presumably going to have to set that gain so my Strat sends the right level to the AT5 Twin sim. How am I going to do that?

Maybe everyone who cares enough about this issue will have to obtain a signal generator that can emit a specific precise voltage, (0.1 V or whatever), plug that into the guitar input, and adjust the gain of that input so that the AT5 sim input shows 0dB at 0 gain?

Playing devil's advocate here...how many guitar players that use AT5 have signal generators that can precisely emit 0.1V? I don't see any other way for the kind of published spec that's being discussed here to be useful. Is that how this spec would work?

[machinated]

Hope that makes sense, if it doesn't I can try and explain it another way perhaps.

Excellent - now let's keep going. At multiple places in this thread I've read the request for IK to publish or document A NUMBER that will solve this problem. What is that number? I don't mean what numerical value, I mean what will that number represent?
[/quote]

Sure! Again, these questions are all in the right direction, and (hopefully) have easy answers.

The number we are looking for is a way of showing the relationship between the analog signal (from the guitar before its inside the computer) and the digital signal that is inside of the computer (that Amplitube is working with).

The reason we need this number, is converters aren't built to the same specification.

If you imagine a 1V sine wave, oscillating without being plugged into an interface.

We can plug this same analog source of a known voltage, into 2 different interfaces, and each one could show a different digital level in your DAW. Just using random numbers, but lets say one interface gives a -10dBFS signal, and the other is a -15dBFS signal. This is the same source, but the interfaces give different digital levels.

Now imagine in context for Amplitube. The designers don't know what interface everyone is using, and there is never going to be a "one size fits all" solution for this, because interfaces vary so much. All they can do is design the plugin so that its accurate for one interface calibration, and anything that differs can adjust the input level. So this is why they have an input control - you can make one interface louder (or quieter) to match the response of another.

This is basically where Amplitube is at now. What we are trying to determine is WHERE we want to correct this calibration to. We don't have a target to compensate for this difference because we don't know what definition of Voltage->dBFS Amplitube is using.

Maybe everyone who cares enough about this issue will have to obtain a signal generator that can emit a specific precise voltage, (0.1 V or whatever), plug that into the guitar input, and adjust the gain of that input so that the AT5 sim input shows 0dB at 0 gain?

Playing devil's advocate here...how many guitar players that use AT5 have signal generators that can precisely emit 0.1V? I don't see any other way for the kind of published spec that's being discussed here to be useful. Is that how this spec would work?

This is definitely a pretty easy way of doing it - every DAW comes with a signal generator plugin, and if you don't have one, there are plenty of free ones available too.

All you have to do is send out a sine wave to your soundcard output, plug a cable in the output and use a $15 multimeter on the other end. You then adjust the output volume until you get a known voltage level. As long as your meter is accurate enough, you're good - you have a precise voltage sine wave.

and then, as you say, you can run that into your input and look at what level it appears in the DAW. If you know what Voltage->dBFS relationship amplitube needs, you use the difference as your input level.

There is an easier way too though, that doesnt require any multimeter or measuring. All it requires is you are using the intstrument input of an interface.

Most specs give a value for "maximum input headroom before clipping" for the instrument input. Its typically written like 12dBu=0dBFS. This means you have 12dBu (approx 3V) of headroom your input can take before you are clipping. You can just set your interface to 0 gain and use that value from the specs.

If you know your instrument input has 12dBu of headroom, and Amplitube needs (say) 7dBu of headroom, you set the input to +5dB and you are done.

[LuisAngel1980]

If you know your instrument input has 12dBu of headroom, and Amplitube needs (say) 7dBu of headroom, you set the input to +5dB and you are done.

Exactly, it is a matter of calibrating your system once to the optimum level for Amplitube, then different guitars will respond differently based on their output level like they would do if you connect them to the real amp. This is true for Amplitube and any other plugin.

The other issue is that it seems that different models seems to need different calibrations, this is a guess, but it looks likely if you compare some models with the real amps. For example, I think the Brit 8000 need much less gain to sound accurate than any of the Fender models. This situation makes really hard to have a consistent experience when using Amplitube.

I find Amplitube worth it just for the Fender models, which I use a lot, but when I try to experiment with other gear/sounds it is a little frustrating because I am not that familiar with the tones they should produce. I have Amplitube 5 max and Tonex max too btw.

I think the modeling tech behind Amplitube is great and you can get really accurate tones out of it, but I need some help to make the most out of it. So please, can we have an answer?

[Matthias Mueller]

Wow thats too fast for me now XD ... i like to see that

I have a strong feeling that 1V peak=0dBFS is used sometimes in Amplitube, but certainly not all the time. I would be surprised if several models didn't use this reference point, but there are significant number which don't. Its hard to recommend values when it could be that, but it could just as feasibly be -3dBu, -6dBu, 7dBu. 10dBu, 12dBu etc. Only IK can say for sure. Sometimes its easier to guess that its closer to say -3dBu than it is to 12dBu but its basically impossible to get anything certain without more information.

Exactly But i think to calibrate to this 1V=0dbFS is a good starting point. From my experience and what i hear from other people its very unlikely that you have to go lower than that (but who knows...). But because its not consistent and different for each amp/pedal, there are some models where you defenetly want to boost more. By how much? When IK dont tell us, we can only assume and reference it by ear and feel to the real amp.

The biggest problem for me I think is in fact the connectivity, firstly how to set the interface input gain (I use a Roland Octa-Capture) with Hi Z input for guitar.

In the manual of the octa core i found this (sorry its in german, its easier for me..):
"1. Combo Input-Buchsen (INPUT 1–4)
Dieses sind analoge Audioeingänge mit Mikrofon-Vorverstärkern
(Mic Preamps). Sie können entweder XLR- oder Klinkenkabel
verwenden sowie zwischen “symmetrisch” oder “unsymmetrisch”
entscheiden.
XLR: -56–-6 dBu
Klinke -46–+4 dBu
Für die Klinkenbuchsen der Eingänge 1 und 2 können Sie die Eingangsimpedanz umschalten:
“Hi-Z” für eine E-Gitarre bzw. “Lo-Z” für andere Geräte (S. 36)."

Normaly you have someting like: headroom: +xdbu (0dbFS)
but i asume that +4dbu means at 0dbFS in your case...
Now you just have to follow the steps i posted in my last post

One thing I keep on coming back to is: what exactly is being asked of IK in terms of documenting the "proper" input level?

An amp sim is a non linear plugin, that means that it behaves different when you feed it with different input signal. When you create such an plugin, you need kind of a reference point, because in digital its not feeded by any voltages anymore (but instead zeroes and ones).

This means what we want from IK Multimedia is something like xVoltage refers to xdb in digital world.
Because it seems that every amp/pedal has different values in this regard we have to send different inputs to it to get realistic behavour from the model.

Suppose you're working with a '65 Twin Reverb sim and you want the "proper" input level for a stock Stratocaster. OK - which Strat? Mine? Yours? My buddy Frank's? They all have slightly different output levels. Or do you want the proper level for an ES335? Again - which one? Or do you want the proper level for a PRS Custom 24?

My point is, all of those present a different level when plugged in to a real Twin. But in every case it will sound like a Twin (a Twin with a Strat, or a Twin with humbuckers, etc.) because it IS a Twin. Different sounds for different levels, sure, but it's a Twin. Would we not expect the same of a totally accurate and faithful simulation of a Twin?

So, its a bit confusing because this is actually a different problem but it corresponds with the other one. To give you an answer on your first question: Your strat! (but when your buddy frank plugs in his guitar, than buddy franks strat... you get the point )

No one here wants to make his guitar sound like the one from anyone else, thats exactly the point. We want the opposite.
But when you do what ik tells you in the manual, then you actually make all guitars sound the same.

I mean you kind of already said it for yourself but imagine:
your buddy has a really cool clean tone with his fender and his twin. He tells you, you should try this with your metal guitar. You dont change anything on the amp, plug your guitar in, now its distorted because your metal guitar has more output. Thats what you expect from an real amp.

But when you do the same in amplitube and do what the manual tells you... as soon as you touch your input on the interface, it is actually the same as you put a volume pedal in front of your amp. In this case you would lower the input gain for the metal guitar, otherwise you would clip your interface. But by doing this you turned down this volume pedal.. the outcome is that your metal guitar is now also clean.

And because not all interfaces are equal and have different headrooms it can also happen that its like this - imagene again:
you made a really cool crunch preset in amplitube at home (with your interface). Now you want to show it to your buddy, you even take your guitar with you to show him as accurate you can. Now you plug into his interface and....oops, its totally clean and no crunch in sight... (could also be the opposite... oops, now its way too distortet)

both of this problems are just natural when you plug in an pc and play with amp sims. Every brand has to deal with this and also ik cant deny...

Bear in mind that every audio interface I've plugged a guitar into has gain or trim control for the guitar input. Those gain controls are generally abstract, they don't have any calibration or units on them. So I'm presumably going to have to set that gain so my Strat sends the right level to the AT5 Twin sim. How am I going to do that?

Yes, most interfaces dont show you in numbers/db how much you boost. But not all, eg. my rme does when i do it inside of totalmix. but thats no problem. you can always download a free/demo daw and use the metering there.
But the way i/we would recomend is, to set your gain on the interface to "0 gain" (no boost, all the way left) You schould only touch it when you get problems with noise floor.

Exactly this (to not touch the input gain) is the important step to keep different characteristics of different guitars and dont make them the same!

[Boogieshoes]

No matter what, it seems to me that the limitation is that the interface is effectively a 'volume pedal' and we can't loose that link.

So for me in real amps each will have a different feel, but I don't have anything except a cable and some pedal between the guitar and the AMP. I can set the pedals so that I can switch them in or out and know, because of testing, how they will work in conjunction with the amp. What you can notice is buffering effects between and across the selection of pedals. Sometimes you want a buffer and sometimes actually they can be a pain with some pedal combinations.

Putting all the pedals in a daisy chain and switching them in and out 'within the daisy chain' can be when buffers get a problem so setting levels as they can become cumulative and interact in ways that are problematic on real pedal board. So much as you are seeking info on the AMPs I think there is probably a reason to aim that question at the pedals too.

In truth, if I believe some of the Youtubes I have watched some people do seem to manage to get something out of modelling packages, I simply have not managed to get something I really like in about 2 years with AT5 Max. I can't be bothered to update the software now since version 5.3 and have just about given up with it. This particular discussion caught my interest as I could palpably feel the pain with others trying to sort things out so I thought I'd join in with you. As a radio amateur I probably do have things I could start testing the interface and software with by injecting signals. I just can't be bothered as I think there are many others things that in the digital world simply don't work like analogue, as Joe Walsh sang 'I am an analogue man, living in a digital world', I think he got the balance of it all for me.

[machinated]

No matter what, it seems to me that the limitation is that the interface is effectively a 'volume pedal' and we can't loose that link.

It’s 100% possible to remove the uncertainty from different interfaces, but only with more information from IK.With the right information from IK, the volume pedal is removed, it’s perfect amp response for all.

HW modellers like Axe FX/Helix/QC etc have their own instrument input where they know the voltage->dBFS relationship. With plugins and different interfaces we just have to adjust gain a little, but it’s quite straightforward for any user to do.

It’s possible to have the exact experience these offer, once we calibrate the interface. I have done this for many plugins with perfect accuracy, because they have provided the information I am trying to get from IK.

[mod redacted: unnecessary and please read the rules of posting thank you]

[Matthias Mueller]

No matter what, it seems to me that the limitation is that the interface is effectively a 'volume pedal' and we can't loose that link.

Yes, exactly... we have to deal with this fact but we (more precisly Ik Multimedia) can find a solution thats easy to use. maybe its one little step more when setting up amplitube but the end result is also more enjoyable/realistic.

So much as you are seeking info on the AMPs I think there is probably a reason to aim that question at the pedals too.

Yes totally, i actually asked the same question to machinated.
If you put your chain together in amplitube and every model (amp or pedal) needs a different input level to work correctly it gets quite complicated.

i think its okay for him when i quote his answer from the private chat here. (If not, please let me know)

From: machinated
To: Matthias Mueller
Yep, you are correct with everything there. If you want to use pedals and combine them with different amps and expect a realistic gain response, it gets very messy/impossible. If you want to swap things and compare, even worse. And if you want to throw ToneX into the mix? even messier still.

I already sent this idea in my support ticket. i came up with this in the private chat so i also just quote it here

Sent: Sun Aug 06, 2023 1:28 pm
by Matthias Mueller
- Someone came up with the idea of a pedal that manages the input...
Good idea, but i can imagine it like this:
You open amplitube/tonex the first time and some start page asks you how much headroom your interface have + guides you through the setup.
(Maybe in the manual are the most common interfaces listed). After that, all different models get boosted the amount they need under the hood/ or they get an equal gainstructure across the board and get boosted/cut to this level...doesn't matter for us end users witch way.
This is now the new deafault. New users get an amazing result right after this short setup. Old users too, but if they dont like to do that they can choose "use legacy mode" on the start page and/or later under the menu to use all the old presets.
For tonex.... when you create a capture, you have to enter your headroom. Tonex stores this data inside the capture file and automaticly adjust for other users. (maybe more or less the same, when you upload an AT preset to tonenet...)

What do you think about this idea?

[dmitch]

Old users too, but if they dont like to do that they can choose "use legacy mode" on the start page and/or later under the menu to use all the old presets.


What do you think about this idea?

That legacy mode is indeed crucial. Without it, some/many of one's existing presets would suddenly sound different when the new system is implemented. Actually, ideally, existing presets would by default have "legacy mode" enabled, and then you can select "legacy" or "new way" on a per-preset basis.