I’m using RoPieeeXL to a Benchmark DAC3 via USB.
I’ve noticed that the DSP volume is actually in the RoPieee side, but I was expecting that it was on the Roon side, alongside with DSP convolution and volume levelling.
If I select Device Volume I’m getting an Enhanced audio path, but I’m not really clear where the “Hardware” part is taking place. Someone can explain it to me?
I’ve also noticed that in RoPieee config, it states that “hardware mixer is not supported”. How does this work? Is it the hardware mixer used only for Roon, or it’s not supported from the DAC? If the later, what does “Device volume” mean in Roon?
@spockfish thank you for you answers!
So, in terms of sound quality, the best solution would be to use Roon using DSP volume, and with this setting calibrate the system to 75/77 db SPL with Pink Noise full range -20db using the DAC’s volume? Or is preferable to use Roon’s Device volume?
It highly depends on how the volume control is implemented on the DAC side. Is in in the analog domain? Then I suggest you go for that route (general rule of thumb: analog volume control is better then digital volume control).
If it’s digital (which I suspect in this case) I tend to choose Roon: purely because it’s done in software, on a powerful platform (your Roon server) and thus is more advanced than the digital in your DAC.
Thanks!
FYI this is an excerpt from the DAC3 instruction manual Page 6:
“The HGC system uses an active analog gain control for analog inputs and a 32-bit dithered volume
control for digital inputs. Both types of inputs leverage the low-impedance passive analog attenuation system at the XLR outputs.
The dual-domain HGC system combines the high dynamic range of Benchmark’s HDR analog control (used in the DAC1 HDR) with the low distortion and accuracy of a digital control. HGC outperforms traditional analog or digital volume controls, including the two- stage DAC1 HDR system. Musical details are preserved over a very wide range of output levels. Analog inputs are controlled in the analog domain. Digital inputs are controlled in both domains.”
It seems that the DSP volume is done on the endpoint and not on the core, therefore requiring another 24bit -> 64bit -> Volume -> Dither to 32/24bits process. I’m not sure if this double volume levelling (i.e in core for the LUFS and in the endpoint for final volume) is compromising the sound quality. If I’m needing a fixed volume control (e.g. -5dB, which is what I need for 75db FS with the DAC at unity gain), would it be better to set the LUFS to a relative -5?