Sure Mojo is more mid-centric, in that it has a softer or fuller bass; and this is down to the use of a single coupling capacitor (digital DC servos would be too expensive to implement for Mojo). Electrolytic capacitors add an unnatural bloom to the sound, which can sound superficially attractive. I have talked before about yin and yang with audio - the battle between refinement and on the other hand transparency. More refinement equates to a warmer presentation; more transparency gives a much brighter presentation. Unfortunately, people are extremely sensitive to this balance; and when you employ listening tests in product development it's crucial that very sensitive and accurate listening is used, as a superficial improvement may just be down to better balance (one distortion balancing another). So I have to consciously be extremely careful when using listening tests, because you can never be 100% sure; that's why a null listening test (when one can hear no difference) is extremely valuable.
I want to give you an idea of the tricks that are available. I talked about capacitors - an electrolytic will impart a softness to the sound (tantalum caps are better - more accurate - and I use that in Mojo), and this is due to dielectric absorption effects, and straight forward low frequency distortion from the capacitor. What happens is a low frequency error is created, and the brain detects this error, and the bass then sounds soft and fuller. Removing the capacitor will make a much tighter, leaner bass; also it will be faster, with better pitch definition. But superficially one just notices the lack of bloom due to the absence of the low frequency error.
Another source of low frequency errors are due to the reference circuitry. Now actually a DAC is simply a reference voltage where the output is digital data modifying the reference voltage. That's simply how a DAC works. The reference voltage is buried inside the DAC chip, but it is crucial to performance - what happens is the DAC draws current from the reference voltage, and this current subtly modulates the voltage directly - and this changes the gain of the system. This is normally overcome by adding a decoupling capacitor; but then what happens is the OP will amplitude modulate depending on the music envelope - and this creates another set of LF distortions, again giving the sense of body or bloom. Unfortunately, this set of AM distortions are not measurable with conventional measurements, and so developers and chip DAC designers aren't aware of the consequences; but it does provide an opportunity for a developer to tweak the sound by changing the decoupling. This is one reason why I talk about my discrete reference circuitry, as I can eliminate this issue. The third trick one can use is straight forward low frequency second harmonic distortion, and I see lots of developers using this to artificially fatten up the sound.
So given that Hugo 2 has a digital DC servo, so no coupling capacitors in the signal path, and the reference voltage is rock solid with zero music envelope modulation, and there is zero low frequency distortion, in a normal DAC this would make it sound impossibly bright and lean - even if the bass is extremely tight and tuneful - the yin-yang balance would be wrong. So why would a conventional DAC sound un-balanced with this approach?
In three words - noise floor modulation. This, where the noise floor pumps up and down with signal - is the principal problem. All other DAC's suffer from measurable noise floor modulation, and this makes things sound hard and bright (there are other errors that do this too). By eliminating the tricks that will add an artificial bloom to the sound means I have to work very much harder to eliminate this problem; and this is an issue that has taken me 30 years to resolve, so it is not easy to eliminate, as there are myriad sources of these errors within a DAC.
Moving on to the filter selection. Now providing the option of adjustable filters very much goes against my purist design philosophy; the 256 FS filter (white or green) is the technically more accurate filter, in that the reconstruction of transient timing is more accurate to the original analogue signal in the ADC, and in the past I would have simply given that. But I included the option for two reasons; firstly as it provides a simple way of getting yin-yang balance, even if we are using an inaccuracy to enable that, and that's OK so long as people are aware of that fact. When you get better headphones, then ideally the incisive filter options will be the best. The second reason was to demonstrate that actually very subtle technical differences can have a profound subjective effect. The technical difference between the 256 FS filter and 16 FS is very subtle; both options employ digital filtering up to 2048 FS, but the 16 FS option is just IIR filters (analogue type filter) and the 256 FS replaces the 16>256 IIR filter with a WTA filter (an FIR filter) which will re-construct the timing much more accurately. But the difference technically is very small. Note that conventional chip DAC's employ no filtering whatsoever above 8 FS or 16 FS (that's 384 kHz and 768 kHz max respectively).
Now although the technical difference is very small, to me the sound quality difference is very much not small. And it does sound very different to a normal WTA filter - you do not get the usual changes in sound quality, but you can perceive the starting and stopping of notes more easily - and because it's like snapping everything into focus, it sounds sharper and more incisive. When you can't hear the starting and stopping of notes, things sound soft and warm. Indeed, poor timing reconstruction is another way where you can soften up the sound - but again it's unnatural, as everything sounds soft, even sharp percussive effects that in real life sound sharp and fast.
To me the change is not small - so I was disappointed that a lot of posters were initially saying that they could not hear much of a change. Maybe I am a hyper-sensitive listener - or more likely the change is much more apparent when you are not doing AB tests, but are listening long term to music, and your brain has learnt about the better accuracy that Hugo 2 offers. But it's good to hear posters starting to report the change that I hear.
Just to wrap up - I have talked a lot about yin-yang balance, and how distortions are used to balance other distortions. But because Hugo 2 is very different to other DAC's, and distortions or errors are fundamentally lower, the downside is that headphones deficiencies will be thrown into relief; transducers that were before too bright may actually become acceptable; and conversely transducers that had "nice" or optimizing errors become unacceptable - it becomes very much easier to hear problems. So bear in mind that your headphone collection abilities will change in sometimes surprising ways.
Happy listening, Rob
