proper method to compare input level plugin and hardware version ?

[Sounddiy]

I am currently modeling a VST plugin.

I am wondering how to test and compare the impedance and input level between the plugin version and the hardware version.

For now, I am sending a 1000Hz signal at -18dBFS to two different outputs on my RME UCX audio interface. Output 1 goes to my effects pedal, which then connects to an input1 on the interface; the other output is routed via loopback to input 2 which add the vst plugin.

I am testing the behavior this way, but what is the proper method?

thanks!!

Math

[griffinboy]

@Sounddiy

It's not possible to know the internal levels unless you probe it electronically. You can make a guess if you know the schematics though.

However, you can work it out from audio measuremnts, if the hardware is nonlinear (which I am assuming it is).
If you use volume sweeps on various test signals, passing into your hardware, you can look at the harmonics and nonlinear behavior at different volumes. And then you can look at your digital model of the same effect, and figure out how to gain stage your digital effect such that it behaves the same as the hardware.

This is something you have to fine tune as you keep developing.
As for impedance, if you have the schematics of the hardware, you can work out the theoretical impedance (of course you won't fully know for sure unless you measure it electronically), and your recording interface will have published impedance specs, so look those up.
In the end, assuming you won't open up the hardware and probe it, this is something you will have to fine tune over time, by looking at the frequency response (particularly high frequency roll off) after you've already done a good deal of the modelling.

TLDR do this stuff while you develop. You can't really do it at the start, so you have to start doing it after you've already done lots of the analog modelling.

The process to follow for all serious good analog modelling is to make a theoretical model first (based on schematics and known facts and basic measurements) and then once you've developed this, you compare to the hardware and now the real work starts: why does it sound different? Why is it so asymmetric in real life? Maybe there is hidden bias in certain places in the circuit? Are certain published "facts" about the hardware design actually not true in practice? Do certain components behave differently when placed in this specific circuit than their published specs and measurements?
This is where the fun begins. But make the theoretical "simple" model first.

Some people mistakenly stop at the theoretical stage. You will sometimes see plugins making claims like "100% created from the real schematics and component values". Not realizing that this is actually nothing to do with accuracy to the hardware and is a gross simplification!
: )

Happy modelling.

[Sounddiy]

thanks @griffinboy These are exactly the steps I was planning to take. I’ve realized that there is, in fact, a world of difference between the theoretical values ​​for resistors and other components and their actual values—sometimes you have to change the values ​​specified in the schematic. Thanks for your advice. I’ll keep on learning. Thank you.