Flat gain vs volume in headphones amp?

[Jprime84]

I am enjoying newly acquired Hifiman Sundara headphones with my also newly acquired Schiit Magni Unity. This is the new model with a built-in DAC.

The Magni Unity offers a 3 stage gain switch on the front:

• +15dB
• 0dB
• -10dB

On the Sundara, I can use them at full volume on the 0dB gain setting and they are right at the edge of approaching uncomfortable listening. Obviously, the +15dB setting gives me more volume range if I really want to crank it, but I presume that also introduces more noise? Maybe asked differently, is the same volume output combination (knob plus gain switch) equal or different as you change gain?

Schiit lists an APx555 test suite for the Magni Unity w/ DAC but I still have a lot to learn to make sense of most of it.

Their specs page does list separate values for the low/high gain settings, so maybe that's an answer right there.

THD+N

Low Gain: Less than 0.0001% (-116dB) at 2V RMS into 300 ohms
High Gain: Less than 0.0003% (-108dB) at 2V RMS into 300 ohms

IMD

Low Gain: Less than -110dB at 4V RMS into 300 ohms, CCIF
High Gain: Less than -108dB at 4V RMS into 300 ohms, CCIF

SNR

Low Gain: Greater than 122dB, referenced to 2V RMS
High Gain: Greater than 112dB, referenced to 2V RMS

In layman's terms, is it better to run at lower gain with less volume range to play with, or are the differences negligible to run at higher gain for more volume range?


Side note, running Oratory's Optimum curve EQ but finding that I prefer to pull the treble down a bit more with an additional 13500 peak filter, -3 gain 6 Q.

 

[DVDdoug]

Generally, it's better not to have excess gain.

If you have a choice you usually want a strong signal through the "chain" for the best signal-to-noise ratio, and then attenuate the signal and noise together at the last place you can.

but I presume that also introduces more noise?

If you were hearing a difference, you would have said so.

All active electronics add SOME noise and high-gain amplifiers will usually add more noise than a low-gain amplifier, and of course an amplifier amplifies everything including any noise. And often, "gain adjustment" is actually attenuation.

It depends whether the noise is coming before or after the gain/amplification/attenuation. For example, if you have a preamp and a power amp with a volume control in-between, lowering the volume will reduce any noise from the preamp but have no effect on the noise from the power amp. Since noise from the power amp remains the same, turning-down the signal gives you a worse signal-to-noise ratio out of your speakers. (Which doesn't matter if you can't hear the noise.)

 

[Jprime84]

Gotcha so in a more complex system if you had one component with particularly bad characteristics, you would want it to get whatever its purpose was done with minimal amplification at that stage, and do your primary amplification later down the line.

 

[Jimbob54]

For most modern (read performs at the level of the good stuff reviewed on here) headphone amps just use the gain that gives you the most useful volume control range for your types of music and loudness preferences but the penalty for additional noise from a higher gain is negligible and likely inaudible if its half decent. If you are normally happy with less than full volume on mid gain, use that.

 

[AnalogSteph]

What kind of levels are you running on the digital side? Having plenty of volume with rock / pop etc. material played with no further processing should not be a challenge at all, but things may start looking a bit different once volume normalization starts being used (e.g. I'm running ReplayGain in Foobar with a pre-gain of -3.3 dB, i.e. a -17.3 LUFS target, which tends to drop levels by 10-13 dB) or on classical orchestral recordings or movies. EQ likewise costs some headroom although you do not generally need a worst-case pre-ccut setting when both are used at the same time.