Yes disclosing Mojo's tap length of 38,896 was inadvertent. I had intended not to disclose it, but actually for very complex reasons.
Hugo 1 was 26,384 taps, and Mojo is 38,896 taps, so that means Mojo must be 1.5 times better?
No.
It's not as simple as that. I have set the narrative that increasing tap length will improve accuracy and will improve sound quality; and that I will keep doubling the tap length until I can hear no further improvements (hence to million tap M scaler).
And this is completely true. But - only if the algorithm is the same, and only if the oversampling rate of the filter is the same. With Hugo 1 and Mojo the WTA was the same.
But Hugo 1 is an 8 times filter, and Mojo is a 16 times filter and clearly 1.3uS resolution (16 times) is better than 2.6 uS (8 times).
But to further complicate matters, Hugo 1 has an extra filter to get you too 16 times; but this doesn't sound quite as good as a single stage filter.
But on the other hand a 26,384 filter works on 3,298 samples; but Mojo's only works on 2,431 samples - and this is a significant sound quality disadvantage.
So I had a very complex situation comparing the two filters, and it isn't as simple as the tap length suggests. And of course Mojo is intended to a different audience than Hugo, so we need to make it simple story that everybody can understand - hence the 500 times more processing power than any other DAC.
There was talk also about it being half speed. There are more DSP cores available on Mojo's 15T FPGA (40 against 16) but - I could not run the DSPs at the usual 208 MHz, due to power limitations - Mojo's DSPs are run at 104 MHz, and this is where the half speed comment came about. Running at half speed halves dynamic power, but halves the available processing.
On a more personal note, I have been upset and very frustrated by some posters on Head-Fi arguing that product xyz is better than Chord; now I am not going to argue with a genuine and honest opinion - after all some people like muddled soft sound and distortion; but I do worry that some posters are working to another agenda. Facts are facts, and for sure Mojo with it's 38,896 tap 16 times WTA filter will reconstruct transients to a far better accuracy than any other non Chord DAC out there at any price. And if you want to be convinced about the importance of transients, take a look at the Hugo M scaler thread.
Moreover, measurement wise, something that can be replicated with the appropriate test equipment (APx 555) Mojo is the only non Chord DAC (at any price point) that has zero measurable noise floor modulation; and this is important for refinement and timbre variation (it's the reason why Mojo sounds so natural and refined).