I have an SMSL VMV D1se DAC which is capable of full MQA decoding and rendering.
Is it better to let the DAC do the whole process, or let Roon do the first decode, then the DAC do the final render?
I stream using ethernet from a MacMini Core.
(Hoping to avoid any unhelpful “Don’t use MQA in the first place” replies)
1 Like
The first decode is identical whether it is done in software (Roon) or in the DAC. Doing the decode in Roon allows the use of “MQA-compatible” DSP which is just about everything in the PCM realm except sample rate conversion.
The MQA team, when they were all at Meridian, used to say “get the audio up-sampled as early as possible in the chain”. I’ve always understood the rationale behind this to be that a 2x rate SPDIF transfer has half the inherent jitter vs. 1x rate SPDIF. But streaming of audio over the network has probably superseded that lore.
There’s probably nothing in it, but there are those who swear that doing everything in their expensive DAC is better. I’m not convinced.
FWIW, I do both in different systems, one of which requires some room correction DSP (so decode in Roon) and one which does not. Network streaming in both cases.
If you set your endpoint to “Decoder and renderer” in Roon, then Roon will not decode unless DSP is also applied.
4 Likes
A very helpful explanation. Many thanks.
