I can't remember if there ever was a BT function envisioned for Dave as it's a feature I would have strongly reacted against. And I can't remember why an SMA connector was put down on the prototype; but I left the SMA in in case I wanted to add another clock or have a port for testing purposes. Remember Dave took nearly 3 years to develop with plenty of prototypes on the way. You always add just in case extra functionality, with plenty of redundant bit switches and spare IO.
I honestly don't know why other reviewers do not get the same results as I do - and every Dave I have tested measured the same. I once was very perplexed with the Stereophile review of Dave, as some measurements looked fatally wrong. When I got the unit back it measured perfectly, with a DR of 127dB (A Wt with the reference set to the max 1% OP level of 6.75v) and the measurement of 16 bit levels absolutely perfect. John Atkinson 16 bit plot gave awful results - and I never got to the bottom of why.
Sure other DACs have managed to get closer to Dave over the last 7 years and some measurements are better (noise for example).
Noise is irrelevant if it's inaudible and fixed - which will be the case for Dave.
But - and this is a big but, it depends what you measure and how you interpret the importance of those measurements - in terms of noise floor modulation (nobody else measures for) Dave I firmly expect to be still the no 1. And interpolation filters that are min phase, NOS or show slow roll off (all of the filters in all other DACs) to me are fundamentally broken and not fit for purpose.
I am an it then? Dave (excepting the metalwork) is my baby.
And on the same week-end that ASR was published, What Hi-Fi (a magazine which Chord do not advertise with) published their list of 16 best DACs. Chord had five of them - all of Chord's DAC product range. And Dave has continued to achieve countless awards in respectable professional websites and magazines. Have all of these journalists been duped too?
My Dave measurements are -127dB Awt referenced to the 6.75v max output level. That would make it 21 bits.
Given that transients are an essential cue for the brain to create the audible illusion, it's no great leap of faith to state that a DAC must reconstruct transients to eliminate timing errors - that is ensuring that transients are reconstructed without being seemingly randomly too early or too late. It is also obvious that interpolation filters will modulate the timing of transients and create these kind of errors. The only contentious aspect is the level of transient timing reconstruction accuracy that is needed, and how the interpolation filter is optimised. And that depends upon huge amounts of listening tests, something that filter designers do not do at all.
Well after you and I have left this mortal coil, I am convinced that in the future it will be de rigour to design high end interpolation filters with transient reconstruction accuracy being the no 1 priority. Just because I am the only one to talk about it or realise the importance of it does not make it wrong.
My designs are about making the performance as accurate and transparent as possible to the original performance as I hear it. If others do not perceive as I do then so be it, I am not bothered. Thankfully there are very many people who agree with me and are prepared to spend their own money on it - something that I am grateful for, which is why I spend so much time on these threads.
Another aspect is that often thousands of specific and detailed listening tests are involved, that is simply not possible with a panel.
And I am not a testee for Chord - I am entirely independent and own all of the IP that goes into my designs.
Actually SINAD is worthless as a useful measurement parameter - at least 0 to 60 does give you one important idea for a car. It's more like measuring the length of a vehicle, as a two seat sports car is much better than a long truck, and then ranking cars solely on their length.
The reason why it's useless is that a DAC with a completely fixed noise level of say -100dB (giving SINAD of 100dB) measured in the context of the way it's being used, would give inaudible noise in reality. If this DAC was perfect in every other regard (no distortion etc) and was compared to a DAC at -120dB noise (giving 120 SINAD), and also perfect in every other regard, the listener would not be able to hear any difference whatsoever. But a DAC that had SINAD of 120 dB but with copious amounts of higher order and anharmonic distortion, noise floor modulation and poor reconstruction filters, would sound terrible by comparison to the lowly 100dB SINAD DAC.
