[raduray]

I just downloaded SoundSource so that I can test parametric equalization on my HD6xx's. But I find that when I turn on either the Headsets Preset or the Parametric Equalization using settings from Oratory1990, there's a significant drop in gain. Note that Oratory1990 calls for a -9.3db in pre-amp gain, so that explains it, but my question is why is this necessary? Why is this better than leaving the pre-amp gain alone and not having to up the gain on the amplifier stage?

 

[VNandor]

The negative pre-amp gain is matched to the maximum boost of the EQ to prevent the track from clipping. If you are curious about what clipping sounds like, just create an EQ profile with a 10dB bass boost while leaving the preamp at 0, play some track that has bass, and you'll hear what I mean by clipping.

 

[raduray]

The negative pre-amp gain is matched to the maximum boost of the EQ to prevent the track from clipping. If you are curious about what clipping sounds like, just create an EQ profile with a 10dB bass boost while leaving the preamp at 0, play some track that has bass, and you'll hear what I mean by clipping.

So the clipping would be in my 2019 Mac's pre-amp? Is the assumption that increasing the gain on my external amplifier to offset the pre-amp gain loss would not clip the amplifier?

 

[VNandor]

Not exactly. The audio samples sent to the DAC is what's getting clipped by the EQ if you don't apply some pregain. Let me show you a visual example (you can click and zoom in the picture to see what's going on):

Here are 3 waveforms of the same track that has some bass on it with different EQs. The first waveform is the one with no EQ applied. The second one has a 10dB bass boost with no pregain. Notice how the top (and bottom) of the waveform gets "clipped". It looks like the waveform can't get above a certain threshold and gets flat-topped (and flat-bottomed). The EQ boosts the samples beyond what they could go, the peaks get shaved off at +1 and -1 which are basically the maximum and minimum values an audio sample can hold. The third waveform has the same EQ profile applied but before adding the boost, I reduced the gain first. That way, once the EQ boosts the bass, the samples still won't try to go above +1 (or below -1) so the pregain prevents the clipping.

 

[raduray]

I do understand clipping. I think you misunderstood the question in my initial response. Assume a laptop with an external DAC/Amp, and I want a certain volume (SPL) coming out of my headphones. For a given external amp volume setting, when I apply the headset EQ the headset's SPL drops. In order to get back to my desired SPL, I then have to increase the amp's volume. So is the assumption behind dropping the headset's EQ gain in order to avoid clipping the laptop's output that the external amplifier will have no such clipping problem when it claws back the lost gain? I guess an amp with sufficient headroom should have no such problem.

 

[VNandor]

Well, you asked why it's better to lower the pre-amp gain than leaving it alone when using an EQ profile. It is better because you don't clip the digital signal (and in turn, the analog output of the DAC) if you use the pregain correctly. Boosting with EQ without pregain almost always causes clipping. Turning up the volume on the headphone amp would only cause clipping when an especially high output voltage is combined with a low impedance headphone.

 

[raduray]

Thank you!