Input Voltage/Gain Calibration for Guitar Plugins

[aciddose]

Burillo do you assume the gain knob on the audio interface does not raise the analog noise floor?
Have you actually tested it?

Why does it need to be on the interface?

Your argument reduces to "any increase in gain adds noise, therefore there is no signal. Ever."

Yes that is totally metal... but not true.

[Burillo]

Burillo do you assume the gain knob on the audio interface does not raise the analog noise floor?
Have you actually tested it?

my interface has 40dB of variable gain on line input according to the manual, and the dial goes -20 to +20, confirming that.

test:

1) plug in a TS cable into a line-in (without connecting it to anything - just a dangling cable)
2) record a track with just thermal noise, with gain knob set at -20
3) record another track with just thermal noise, with gain knob set at +20
4) adjust gain on the first track to be +40dB

result: first track has noise peaking at -48dB, while the second track has its noise peaking at -84dB.

my instrument input has 80dB of variable gain on its instrument input according to both the manual and the dials. do the same as above, except for adding 80dB of digital gain to the first track.

result: first track has noise peaking at -9dB, while the second track has its noise peaking at -36dB.

so, yeah, i just did test this, and it seems that i'm correct in "assuming" the obvious: that having your input stage properly calibrated is important for maintaining low noise floor.

[Ilya-v]

I have tested it too and you are right Berillo, adjusting the gain on the audio interface does maximize SNR, it does not linearly add gain as in a simple plugin boost as I mistakenly thought.
BUT this does not mean this thread is obsolete just yet, it means the calibration procedure will be slightly less straight forward.
I still want 1vAC peak at the input to be 0dbFS but the calibration procedure in DAW has to be done after guitar+gain adjustment.
Yes, this approach is not straight forward plug and play but requires some know-how, yet it is still very much valid for most amp sim plugins except maybe Amplitube (as far as we know).

[Burillo]

I have tested it too and you are right Berillo, adjusting the gain on the audio interface does maximize SNR, it does not linearly add gain as in a simple plugin boost as I mistakenly thought.
BUT this does not mean this thread is obsolete just yet, it means the calibration procedure will be slightly less straight forward.

which is what i've been saying all along, and i'm pretty sure this is what aciddose was trying to tell you as well. i'm glad you are able to change your mind when you're proven wrong.

I still want 1vAC peak at the input to be 0dbFS but the calibration procedure in DAW has to be done after guitar+gain adjustment.
Yes, this approach is not straight forward plug and play but requires some know-how, yet it is still very much valid for most amp sim plugins except maybe Amplitube (as far as we know).

that i don't care about. do you. just don't give bad gain staging advice.

[michaeltn86]

did a step by step following ur instructions on 1st page. I was kind of understanding wrong, but my 1st and 2nd post were spot on on what I think.

I'm kinda of glad that what I'm thinking makes sense to me at least. it was a kind of waste of time, but I'm kinda of glad that I did the test as I finally could make a direct comparison with some OTB gear.

that said it was an interesting test and for those interested I converted in mp3 and will also share the whole project. I will keep them available for one week only.

https://www.dropbox.com/sh/xpumurbz7szx ... vC9sa?dl=0

please note that I splitted my DI signal. please note u are also free to use my files to test with ur amp sims if u feel like it.

here are the details:

1)hi-z's cable is a LONGER and a different brand cable so NOT a 1-1 comparison. my interface is a MOTU 1248. the cable is going straight to the interface.

2)the other signal is a DI, going into a compressor (max 1-4 GR db), then to the line level in the interface. I used the compressor in all clean guitar files. only the "distorted" signal I didn't use compression to avoid the noise floor increase (though I did use compression on the clean noise file).

3)my guitar is a Squier classic vibe heavily modified. I use 0.11 Ernie ball. the strings is about 2-3 weeks old. when there is no noise, it's position 2-4. with noise is generally neck, but I also used it bridge in 1.

4)I'm not a consistent player, and my hand is kinda heavy. sometimes I put too much strength on purpose, as when u are recording, many times the best takes are the ones u are not thinking. and this was a test.

5)it's just delay + amp sim. it's not about sounding good and it's not about my playing or tone I believe haha

well, in the test I 1st create a sign generator - 20 dbfs -> route it to a hi-z gain into my Motu 1248 interface. my interface hi-z can go up to -12dbfs. u can see that in the logic project.

I recognized the sinal as -17 dbfs. went + 17 db with logic EQ as in my project.

in the first example the song is "I'm not in love". the other is a well know chord progression.
https://www.dropbox.com/s/7z9aocyvcm8ff ... z.mp3?dl=0
https://www.dropbox.com/s/o806rnc2htfuw ... n.mp3?dl=0

in the 2nd example the song is "I'm not in love" and an original blues. by this time the latency was getting longer due to many plugins, so I apologize for my playing
https://www.dropbox.com/s/kv6qe9zpnh2fu ... z.mp3?dl=0
https://www.dropbox.com/s/2gy3bxqjcooj5 ... n.mp3?dl=0

in the 3rd example the song is still smoke in the water and an original. the latency was really difficult.
https://www.dropbox.com/s/1iun9sepvmydp ... e.mp3?dl=0
https://www.dropbox.com/s/mnp80tv50k89p ... e.mp3?dl=0

I do think the noise ratio does increase in the hi-z, but it's not something that would bother me, even though I'm very sensitive to noise. I do think the transients sounded different to me, and that would bother a little bit more.

the files are not "level matched" but are in the ballpark.

I do think that setting up the guitar, and setting up ur recording chain comes before the "gain staging". it really makes me wonder what the op's guitar sounded like to make such a bold claim in the first post.

probably, it was not about gain staging, but that the level that was hitting the converter was just too much.

that said, I also checked some other amp sims, and I think that some amp sims definitely sounded better with levels peaking higher than -10 dbfs, KUASSA being one of them. also I found out why I like the bass amp of theirs better, it's because of the compressor.

so, all in all, at least, I'm not imagining things and my choices are making sense at least for now.

I don't advice testing the way the op suggested in the 1st post.

hope this is helpful for someone out there :]

cheers

[michaeltn86]

omg, I'm sorry Ilya-v, I was really understanding wrong what u wrote.

after dinner and having a shower I took the time to re read everything again.

I also noticed the sound example on page 2. for me it was an average sounding guitar sound, and that's totally fine and good sounding to me.

I also understood Burillo and aciddose messages, finally. I was reading too fast and just Ilya-v's posts. that was confusing me too much.

learned a good lesson today. thank u everyone.

[JardCrocker]

Hi...i am a new user here. As per my knowledge without calibration a 50mV single-coil Telecaster can be as hot as a 3V EMG, that is 60 times the voltage.If you adjust the gain pot on the audio interface, all guitars hit the red light making the equally hot.

[Burillo]

Hi...i am a new user here. As per my knowledge without calibration a 50mV single-coil Telecaster can be as hot as a 3V EMG, that is 60 times the voltage.If you adjust the gain pot on the audio interface, all guitars hit the red light making the equally hot.

the point was not to make all pickups equally hot, it was the opposite - to represent "real" voltage in a DAW, so that your 50mV Telecaster doesn't hit the virtual amp as hot as a 3V EMG does. the problem is that the way the author chose to do so degrades the signal, so a better way to achieve the same effect would have to be devised.