SMSL RAW-MDA 1 DAC & Headphone Amp Review

[tccalvin]

Not all CS 43131 devices have the Cirrus hump.

True, but all CS devices have added distortion as long as DRE is enabled. That said, the DL100 almost certainly also has the Cirrus hump, since I've owned and tested both a DS100 and a M300SE by SMSL and they both had the hump (i.e. the maximum amount of distortion that you can get with DRE).

Amir's tests do not show this because the suite of tests that he performs simply elude the issue.

Again, this doesn't mean that it's terrible or that you shouldn't enjoy your unit. I was simply making the point that, all else being equal, it wouldn't make much sense not to go for the better measuring product.

 

[nick_l44.1]

I think in Amir's tests the DL100 was almost as good as the RAW-MDA1. Not all CS 43131 devices have the Cirrus hump.

Not at all. Compare the THD+N ratio versus frequency. Aside from the fact that the DL100 has this noise, measuring this noise masks a sharp increase in distortion as frequency rises. This increase is particularly pronounced when the volume is lowered.
The hump is simply a DRE artifact. The fact that it’s usually audible imho isn’t the main problem with the CS431**. These are very outdated chips and, by all accounts, are based on an even older architecture. Their main issues are high noise and low processing power. Well, it’s actually hard to say exactly how modern chips differ in terms of noise. In the case of the CS, the biggest problem stems precisely from the combination of these factors. As a result, these chips are unable to filter out all noise outside the operating frequency range. Incidentally, this is precisely why support for sampling rates above 196 kHz is purely nominal. Also, with a dense spectrum, the noise shaper’s efficiency approaches zero.