Question about graph of THD+N Ratio vs Measured Level

[duyu389]

Hello friends! I've read a lot of reviews about audio devices, and I've found that some devices can get the power output to a level where THD+N rises dramatically, while others cannot, as shown in the figure.

Then the question arises: how to overload the output of the device? Is it simply a matter of turning the volume knob? If the output cannot reach the level of THD+N rising sharply, and it is not yet the upper limit, why is the measured maximum output power still valuable?

 

[Multicore]

The Apple USB dongle is a digital device so the largest signal (bits) you can send to it is a full-scale swing (roughly speaking). The Schiit headphone amp has an analog input so the largest signal (volts) you can send to it is not limited like the digital. So the headphone amp can be overdriven and that's the typical kick-up in distortion we see when you hit the device's intrinsic limits. And Amir's measurements and graphs are designed to reveal where those intrinsic limits really are.

 

[duyu389]

The Apple USB dongle is a digital device so the largest signal (bits) you can send to it is a full-scale swing (roughly speaking). The Schiit headphone amp has an analog input so the largest signal (volts) you can send to it is not limited like the digital. So the headphone amp can be overdriven and that's the typical kick-up in distortion we see when you hit the device's intrinsic limits. And Amir's measurements and graphs are designed to reveal where those intrinsic limits really are.

Thank you for your quick and efficient answer! Does this mean that the Apple USB Dongle will never output a severely distorted signal? It just turns out that it's possible that the volume isn't loud enough, right?

 

[DVDdoug]

Is it simply a matter of turning the volume knob?

Yes. With the Apple device, at 0dBFS (the digital maximum) the amplifier is not pushed to it's limit. Your digital data will clip before the analog amplifier.

You'll find the "opposite" in audio interfaces where the built-in preamp has enough headroom that the analog-to-digital converter (for recording) clips before the preamp. Both can be clipped if you push it far enough but you'll never know the analog is clipping.

 

[duyu389]

Yes. With the Apple device, at 0dBFS (the digital maximum) the amplifier is not pushed to it's limit. Your digital data will clip before the analog amplifier.

You'll find the "opposite" in audio interfaces where the built-in preamp has enough headroom that the analog-to-digital converter (for recording) clips before the preamp. Both can be clipped if you push it far enough but you'll never know the analog is clipping.

Thank you for your answer, it opened up my mind. But I don't fully understand that why analog-to-digital converter is beyond the preamp for recording... I thought it was "sonic wave-pre amp-adc-digital signal" ...

 

[Jimbob54]

Thank you for your answer, it opened up my mind. But I don't fully understand that why analog-to-digital converter is beyond the preamp for recording... I thought it was "sonic wave-pre amp-adc-digital signal" ...

Think there is a translation issue there. The adc will clip earlier than the preamp. Not that it is before the preamp in the signal chain.

 

[duyu389]

Think there is a translation issue there. The adc will clip earlier than the preamp. Not that it is before the preamp in the signal chain.

Oh I see. My Audient's line-in will get clipped when I turn knob to about 11'o clock with 0 dbfs sinewave playing. Does that mean for some weaker sound from mic, it still got the ability to zoom up the signal?