Rob Watts
Member of the Trade: Chord Electronics
- Joined
- Apr 1, 2014
- Posts
- 3,146
- Likes
- 12,451
Is Blu 2 the end though? I doubt it.
If DAVE's output stage can resolve the performance of a noise shaper with 350dB of resolution, then Blu 2 seems to be far short.
Now playing: Stina Nordenstam - Fireworks
You are mixing up differing errors; there is the small signal non-linearity error, which the 350 dB figure refers too, and the transient timing error, for which the Blu 2 is trying to reduce - that's what the 1M taps are for.
The small signal non-linearity error for Blu 2 actually exceeds Dave's performance:
The noise floor up to 16 kHz comes from the test tone.
As to the transient timing error, then I do not know how much further one needs to improve it; all I know is that against the ideal sinc function (which will perfectly reconstruct the bandwidth limited signal inside the ADC before sampling) the 1M taps WTA exceeds 16 bit accuracy. Whether our hearing needs more than this remains to be seen; Davina will tell me how far short we are from perfectly reconstructing the original un-sampled signal from the subjective POV.
Interesting times ahead.
Rob