How to Normalize Output Levels?

[watchnerd]

Devices:

Luxman L-590AXII integrated amp

Schiit Mjolnir 2 hybrid (SS/tube) headphone amp

Key features:

1. Luxman has pre-out, main in, and a Separate button that allows you to drive the pre-out to something else and feed it back to the main in.

2. Schiit has pre-out

3. Both have volume controls and headphone jacks

4. Speakers on Luxman can be turned on or off independent of whether something is inserted in the built-in headphone jack or not


Testing hook up with Separate button engaged:

Source -> Luxman analog inputs -> Luxman pre-out -> Schiit SE input / output -> Luxman main in -> Transducers


I've tried level matching by ear using test tones and fiddling with the relative gains of both and I think it's close, but I know even tiny amounts of level difference can impact ABX testing.

What's the best way to normalize the volume of the two using instrumentation?

 

[DVDdoug]

Make a test tone. You can create a test-tone file with Audacity.

Use a multimeter to measure the voltage. You don't need an expensive meter. The readings don't have to be accurate, just repeatable. You'll (probably) be reading less than 1V and some inexpensive meters don't have a low AC voltage range so make sure it's got a low-enough range to read with a few millivolts resolution.

but I know even tiny amounts of level difference can impact ABX testing.

Right... The main thing is, if you can hear a difference in volume you can always tell if X is A or B and it invalidates the test (unless volume differences is what you're listening for).

As you may know, an ABX test only confirms that you can reliably hear a difference. Then you can investigate the nature of the difference or choose a preference.

If the Schiit introduces audible noise (or changes the noise level) that's a valid difference but it makes it harder to "prove" any other differences. And you probably don't need an ABX test to demonstrate or prove a difference in noise.

 

[watchnerd]

Make a test tone. You can create a test-tone file with Audacity.

Use a multimeter to measure the voltage. You don't need an expensive meter. The readings don't have to be accurate, just repeatable. You'll (probably) be reading less than 1V and some inexpensive meters don't have a low AC voltage range so make sure it's got a low-enough range to read with a few millivolts resolution.

I have a Siglent DMM.

I can also create a test tone via my function generator.

Should I just stick the DMM probe into the headphone out jack?

Right... The main thing is, if you can hear a difference in volume you can always tell if X is A or B and it invalidates the test (unless volume differences is what you're listening for).

As you may know, an ABX test only confirms that you can reliably hear a difference. Then you can investigate the nature of the difference or choose a preference.

If the Schiit introduces audible noise (or changes the noise level) that's a valid difference but it makes it harder to "prove" any other differences. And you probably don't need an ABX test to demonstrate or prove a difference in noise.

The goal isn't to differentiate a difference in noise between the two, as this is empirically obvious.

The goal is to see if I can hear the difference between different tubes I stick in the Schiit by recording them to stick in ABX comparator software.

The native output of the Luxman is relevant just to keep it in the range for what I consider a typical comfort level under normal circumstances.