Quote:
Originally Posted by
Head Injury 
The Clip players use digital volume attenuation to adjust the volume. I think the voltage output is always the same (I could be wrong there) and volume is instead adjusted by decreasing the volume of the digital files before they're converted to an analog signal. What this means is that adjusting the volume below 0 dBFS (the maximum digital volume) costs you bit depth. Every -6 dB loses a bit. Doesn't mean anything for normal use, just that the noise floor will be higher relative to the signal (because the noise floor is constant at a maximum -96 dB for 16 bit files).
When you amp the Clip's signal with an external amp that has an analog volume knob, it's usually a good idea to set the digital volume to 0 dBFS so you get the lowest possible noise floor. Don't set it above 0 dBFS because that is the maximum volume a digital file can be and any higher may (in the case of most modern songs, will) cause clipping. Analog volume attenuation is usually superior, because it lowers the noise too. However, some amps might have a problem with channel balance at low volume settings, which is arguably much more audible than a higher noise floor. If one ear is louder than the other at low volumes but not at higher volumes, lower the Clip's volume too.
Impedance is only one concern when determining how much a headphone needs an amp. Sensitivity is just as important. Ignoring volume, low impedance headphones will benefit most from a high quality amp, because amps tend to perform worse into low impedance loads.
So sorry I took so long to get back to your response on this. I remember seeing it and meaning to but my mind blanked. Very informative post. Thanks! I usually have my Clip+ (Rockboxed) set to about -2 without the amp. Am I still losing bit-depth even though I am not lowering the volume all the way to -6? Also, I am looking to get the KRK KNS 8400 and drive them with my Clip+. Do you see any problems coming from this?
Again, thanks for the informative post. I appreciate the new knowledge! :)
Edit: I also have another question regarding the bit-depth, volume and distortion. My pre-cut is currently set at -6db (to avoid distortion - especially when EQ'ing). Furthermore, I have my highs and lows (up to 150hz and above 10 000 hz) set at -5 and -8 decibels. I have this crazy EQ'ing done because I hate the native response of my Pro 900. Hence, I am looking to purchase the KRK KNS 8400.
Would this reduction of decibel value in my precut and EQ lead to bit-depth deterioration? Furthermore, are there any other sonic issues this could cause?
I am looking to purchase the KRK KNS 8400 but apparently they are a bit hard to drive even though they are 36 ohms. Of course, I will set all my EQ values to zero (or close to it) or even turn off the EQ once I get the KRK KNS 8400. Its native sound signature looks appealing. I looked at graphs on inner fidelity. I am thinking I can just turn off the precut too, that is; if I turn off the EQ.
Thoughts?
Edited by Sonic Atrocity - 1/2/12 at 7:49pm
