Level matching for dummies?

I'd like to perform a more careful comparison between two different amps+headphones fed by the same DAC. I don't have specialized gear (microphone, measuring equipment) for that purpose, otherwise I wouldn't be asking here. What's a minimum procedure, maybe using phone/tablet apps, that would not be embarrassingly misleading? Looking for knowledgeable advice. Thank you!

 

One relatively simple way is to use a multimeter. This technique is often used for speakers but should work for headphones too.

Set the volume at a comfortable level (the level you plan to use for critical listening). Now play a 1khz tone and measure the voltage across one of the channels and ground. Record this voltage and mark the position of the volume knob.

Now switch dacs and play that same 1khz tone, measure the voltage and adjust the volume control till it is the same as the previous voltage you measured. Mark this position of the volume pot.

Now you can fairly easily match conditions for the 2 dacs. (not as simple as speakers as you will need to access the headphone jack wiring in the amp)

This will not work if you plan to use multiple headphones etc. Ideal for the mantra of one element at a time.

.. dB

 

@dBel84 For dacs or amps, I totally agree. But, how can the suggested method ensure nearly identical "loudness" (as opposed to sound pressure at 1khz signal)?

To my understanding loudness level is a function of sounding body's FR and music's frequency distribution. The latter is even a function of time..

It's actually one of the areas I've struggled with 1khz single tone calibration. These days I prefer to use multiple wide-band testing materials. And measure B or C- weighted loudness to get calibration levels for each material. Then take an average for final calibration level for each setting.

Of course this is sorta imperfect (I am thinking it's inherently impossible to perfectly match two speakers or headphones' loudness level). So, what I'm trying to address this issue is create evaluation sequences with random assignment of +1/0/-1 db extra attenuations.

In AB comparison, I'm heavily depending on digital attenuations (with setting analog pot at 12pm) that enable me to adjust levels by floating number decibel resolution.

 

Yes, this is the compromise of trying to apply objective measures to a subjective task. The ideal would be to standardize across frequency (as an aside, this was always the conversation I would have with Alex when we discussed measuring how good a product is - the things that bring us back to listen to certain equipment time and again cannot be measured, they have to do with phasing and harmonics across the frequency spectrum)

This method is simply a minimal tools required - a $3 multimeter and a streamed or downloaded tone.

Personally I don't bother and I set speaker levels roughly using the audiotool app and an external mic that plugs into the audio jack of my old phone (thanks technology for doing away with headphone jacks) with the test track in question (no good for headphones unless you rig up a dummy head.

.. dB

 

The actual listening bit of the test is always subjective, isn't it? The set up is just to provide a level playing field, eg so that .25dB* difference doesn't get interpreted as "fuller richer tone"

(Disclosure: my ears are so much the weakest link of any chain that none of this matters to me in practice any longer --- but I am still interested in the principles)



*Or whatever --- a difference that is not big enough to be heard as loudness.

 

Thanks, for my purposes, the multimeter method you described is enough; if I got obsessive I might try multiple tone frequencies. The goal is indeed to get the same level out of the two amps so that I'm not super-fooled in describing differences between the two chains. To be clear, I have one DAC switched into two different amps, I've been enjoying both a lot with different headphones, but I want to be able to describe the differences without falling into the "louder sounds better" trap.

 

Oh My bad. I misunderstood you were going to compare two distinct sets of headphone-amp combos, which I often do myself lol.. If both dac and headphones are fixed, then nothing to worry even if you use the method you said.

One minimalism tip that worked for me is to additionally try with slightly lower (and/or higher) volume with amps you try. Say, you like Amp A over B. Then you can lower A's volume by tiny bit (to the extent B sounds clearly louder) and repeat listening to the same track with both amps again.If your finding persists, your evaluation is more likely robust. If not, we can suspect some volume-dependent bias.