Rob Watts
Member of the Trade: Chord Electronics
- Joined
- Apr 1, 2014
- Posts
- 3,141
- Likes
- 12,370
Rob,
Thanks very much for your detailed explanation as to why and how the M Scaler is superior to PC upsampling. You have certainly convinced me!
Just one more question, and forgive me if it has already been answered or if I am misunderstanding the way your DACS work.
Presumably (and I'm sure I'm oversimplifying), your other DACs (the Hugo2, HugoTT2, the DAVE, and the Qutest) use a similar method to process audio, but with lesser FPGA processing power and far fewer taps. My question is, after the audio signal goes through the M Scaler and it does its magic, and the signal then goes to the connected Chord DAC, does the Chord DAC attempt to upsample or otherwise process that signal using its lesser power and taps, or is that somehow bypassed?
Thanks very much!
Yes it does - and Qutest, Hugo 2, TT2 and Dave all share the same process. The M scaler is the first stage of converting a sampled digital signal back into the continuous analogue signal before sampling; and it does this in a twofold process. First another WTA filter (WTA2) that takes you from 768k (16FS) to 12.288 MHz (256FS); then a third order filter that takes you to 104 MHz. Thus the DAC noise shaper, which also runs at 104 MHz, is fed with a digital signal that is very close to the continuous analogue signal that was in the ADC.
So if I understand what you're saying, a 16bit 44khz source file sent to the M Scaler gets upsampled to 705khz and the Hugo2 treats it the same as if a 705khz source file was sent directly to the Hugo2. And any taps, etc. the Hugo2 applies is cumulative to what the M Scaler has done.
Yes; the intent is that we are converting the redbook recording to as if it was a 705.6 kHz recording; indeed from an interpolation it will be accurate to better than 16 bits.
Rob,
Since the WTA filter is now identical to the sinc function filter, how much room is there for future improvements to CD sound quality? Have we approached a maximum, and so future sound quality improvements will have to come from other areas, e.g. the DX amplifiers?
Thanks
Obviously not identical, but identical to 16.6 bits to an ideal sinc filter. Of course, that leaves open the question of whether 16 bits is good enough, and how much more we get from longer tap lengths; but for sure there must be a point of diminishing returns. But the DX amps will offer a big step forward in transparency, judging from my listening tests so far; being able to effectively remove the sound of a pre-amp and power amp is no small change in transparency.