Me: 
That’s some news I can use – I am leaving the Grace dials alone and saved the setting in the Lynx as a recallable “scene”
Thank you again - going to check out the videos and subscribe to your YouTube channel
good god. not this nonsense again. here as well?
Glad you’ve found some correct information (directly from NDSP’s developers no less) on gain staging for accuracy with NDSP plugins. I hope you enjoy your guitar tones more than ever now!
@MirrorProfiles I decided to test my Quad Cortex myself to see if this matched and I appear to be getting different results. With the impedance set to 1M I am sending a -13dBFS sine wave out of output 3 back into input 1 and measuring and the incoming waveform measures at -13.2dBFS which is significantly less of a drop in signal than is suggested here.
Is it possible that a recent update to the Quad Cortex adjusted default gain levels?
At what frequency should i generate this sine wave?
1khz?
Also, i think that the process should look something like this:
-
Gain stage the guitar so that the ADC in the audio interface are hit properly (i.e. push just before clipping on the preamp)
-
input a 1 Vp = 0.707 VRMS = -0.79 dBu into the same input
-
Read the peak meter inside the daw
-
Adjust the digital gain of the digital signal accordingly until it reaches -13 dBFS (boost or attenuate digitally)
-
Route this into the Neural DSP Plugins
I’m looking forward to doing this but i don’t know how to generate the sine wave.
1kHz is a pretty standard frequency for audio test signals as long as your multimeter works well in that range.
As far as gain staging, I’d only really consider adjusting the gain from 0 if:
- your signal is extremely quiet and near the noisefloor
- your noise level is very high and near the quietest signal from your guitar.
For most modern interfaces, at 0 gain you’ll have a signal that is already suitably loud and well above the noisefloor of your pickups.
Optimising further still isn’t really possible, as long as your interface noise floor is lower than the noise of your pickups. All you are doing is raising the background noise, your signal to noise ratio will not get better.
The benefits of leaving the input gain at 0 are:
- it’s repeatable if you use your preamp for recording mics and other signals, you can just go back to 0 and have a consistent level for all guitar tracks
- you can keep the same proportional headroom for each guitar rather than having to optimise and adjust for each (with no real benefit)
- You can reamp guitars through real gear without having to recalibrate your reamp box each time
- you can switch between plugins which require different levels easily by saving presets with the levels adjusted already. If they’re always different you’ll always have to calculate the different amounts.
The only reason not to do this if your signal is so quiet that noise is an issue. Most interface manufacturers are designing their products so it’s solved by default - that’s why they typically have a maximum of 12dBu or so and not 30dBu. Something around 12dBu is JUST enough headroom to capture a loud guitar DI without clipping. The Quad Cortex uses 15dBu, and I believe an Axe FX is 18 or 19dBu, so more can be absolutely fine, although you’re unlikely to get close to clipping those (while 12dBu MAY clip occasionally).