Relating sound quality to technical performance is very complex, and I will try to explain, but I could talk for days about it and completely confuse everybody. But here is a quick answer to your questions.
"I hear what people refer as "musical" and "emotional," but I want to understand why" Musicality and emotional is complex, but in a nutshell its about removing distortions that interfere with the brains ability to understand the music. Conventional DAC's have a number of distortions that make it much harder for the brain to perceive the sound. Now we underestimate what the brain does with hearing, and simply make the assumption that the ears convert sounds into nerve impulses, and that's that job done, the brain simply access's the nerve signals. But that's not what happens - audible reality is an illusion created by the brain, and a considerable amount of brain processing is employed to create that illusion. So for example, you listen to a guitar and a singer for example. The data the ears feed the brain is a jumbled up mess of information, and the brain separates this mess of data into two distinct entities - the guitarist and the singer, and you perceive this as two separate entities. Not only that, but the brain very cleverly calculates where in space those entities are, and it does this from subtle timing, amplitude and resonance cues from both ears. But this requires considerable calculation. Moreover, small and subtle distortions (by saying distortion I mean anything that changes the original signal in any non linear way) interferes with the brains ability to separate sounds out into distinct entities, and interferes with the brains ability to place entities in space. This has two consequences for being able to enjoy music - firstly the brain is struggling to process the data, so has to work harder - which means you get listening fatigue, and so you can't enjoy the music. Secondly, being able to enjoy the music means being able to perceive what is going on - and there are many distortions that disable the brains ability to perceive the music. This is where it gets complex, as there are a myriad of different distortions that upset the brains processing. That's why Mojo has the WTA processing, why it filters and over-samples at 2048 times, why its got noise shapers that are a thousand times more resolving than conventional noise shapers - I could go on.
"The iDSD sound more flat, more monotone almost compare to the Mojo" The perception of depth information is down to very small amplitude differences of small signals. Now the brain calculates depth from a number of different cues, but most of it comes from the reverberant sound from the acoustic the recording was made in (or depth is added by adding artificial reverb). Now reverb is very small signals, and the amplitude accuracy of these small signals is crucial for the brain's ability to calculate depth. Now there is something very strange about depth perception - and that is the brain needs these small signals to have perfect amplitude linearity. If a small signal is slightly larger or slightly smaller than it should be, then the brain gets confused and can't calculate the depth properly, and things then sound flat. But the amazing thing is, there appears no limit to how accurate these small signals need to be in order for the brain to not truncate or flatten depth. In order to accurately reproduce depth you need extreme small signal linearity. You can't do this with R2R DAC's, as the resistors can't be matched. With DSD or delta sigma (Mojo is delta sigma too) the problem is now how well the noise shaper functions. As a signal gets closer to the noise shaper noise floor, the levels get smaller, as a signal that is smaller than the resolution limit of the noise shaper is truncated. To overcome this you need to have very high resolution outputs, with a noise shaper that has very high resolution - in Mojo's case, the noise shaper has a thousand times more resolving power than conventional high end noise shapers, and ten thousand times more resolution than DSD 64. But there is another source of error that can upset sound-stage depth and this is digital noise adding to the analogue signal. This applies to all DAC's, and is a big problem with chip DAC's, as there always exists a path from the digital noisy part to the analogue part, and this noise corruption will degrade the small signal non-linearity. But with Mojo the actual analogue parts are discrete, so its possible to eliminate digital noise from corrupting the signal. There is another mechanism for depth to be truncated, and this is with metal to metal interfaces. When you have a soldered joint, or any metal to metal interface, oxides and impurities concentrate at the interface. This oxide barrier is non-linear in that the resistance to small signals is larger than with big signals - so again we have small signals being attenuated. To reduce this problem you can only do this by reducing the number of passive components in the signal path. Conventional DAC's (delta sigma and R2R) have very complex analogue components, due to the need to convert from differential to single ended and to filter the high amounts of RF that comes out of a conventional DAC. With pulse array (my DAC technology within Mojo) this is not an issue as I can get single ended to work, and it runs at 104MHz, so little analogue filtering is required.
"The iDSD sound more flat, more monotone almost compare to the Mojo" I think here you are referring to timbre - the tonal colour of the instrument. Now timbre is an issue with timing reconstruction, as the brain uses transient information to infer the timbre of an instrument. Now conventional digital has uncertainty in the timing of transients (does a signal cross through zero just after a sample, or in the middle or close to the end of a sample?) and the only way of recovering the timing information perfectly is to use an infinite amount of processing on the interpolation filter. With the use of the WTA filter, which has been optimised to recover timing, and 500 times more processing than conventional DAC's, I can reduce the timing uncertainty - which results in much better timbre variation, so things don't sound monotone. There is another aspect in that noise floor modulation also affects timbre reproduction, but this is answered in your next question.
"Then there's the matter of warmth. The iDSD is known for its warmth, yet that warmth has a bit of way veiling the sound. Mojo also has a hint of warmth, but that warmth is applied without extra veil. I'm in no way dismissing the iDSD, as it is absolutely superb, and perhaps more neutral" Warmth or smoothness can be can be artificially created - for example with a dollop of 2nd harmonic. Mojo has very low levels of distortion, so its warmth is not down to doing this or other things. The key to true refinement with DAC's is noise floor modulation. This is where the noise pumps up and down with the signal, and all other non Chord DAC's have large amounts of noise floor modulation. Now noise floor modulation is a scary issue with DAC's, and there are countless ways that a DAC can suffer. Mojo, on the other hand has zero measurable noise floor modulation - the noise floor is at -170dB and it maintains this whether its output is 2.5v or zero, the noise is completely static. Now the issue of neutrality is a very complex thing, as increasing transparency will make it brighter and sharper, and increasing refinement will make it smoother and darker, and its possible to use distortion to create the impression of warmth or brightness. To be honest, I (or anybody else for that matter) do not know what the tonal balance of a perfect (and hence neutral) DAC is. And neutral cam mean different things to different people and with different gear!
Mojo's musical performance is down to lots of technical things - way to complex to talk in detail with - but there are solid reasons why you hear what you hear, and why other DAC's can't do this.
Rob
