Metrum Musette

[Jud]

That could be the NOS DAC thing, as mentioned in the Metrum paper - NOS do not measure well, but sound better possibly. AFAIK, Rega DAC don't do oversampling as claimed by them - doesn't matter which on -R on non-R.

 

Rega states "We decided not to use a sample rate converter and process the data at the incoming sample rate which keeps the signal processing to a minimum."

 

The Rega DACs do oversample, hence the filter selection, you need to oversample to use those filters. What Rega means here, is they don't upsample prior to oversampling like some dacs do or allow you to do - i.e. 44.1 to 192k

 

I hate marketing talk, it's so confusing.

 

The second page here is the block diagram for the chip used in the Rega: https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=2&ved=0ahUKEwijyOiU1NvLAhWJbD4KHflXAGoQFggkMAE&url=https%3A%2F%2Fwww.cirrus.com%2Fen%2Fpubs%2FproDatasheet%2FWM8742_v4.3.pdf&usg=AFQjCNELkLoPo9QCzAI5fzt8e3PSPXAivw&sig2=QQvzFKfArXZNKte7yvk_Wg

 

Note the blocks for the "PCM digital filters" and the sigma-delta modulators.

 

When you select the filters on the DAC-r, you are selecting the "PCM digital filters" you want to use in the process of sample rate conversion.

 

So why does Rega claim not to do sample rate conversion?

 

It has to do with the history of DACs.

 

The DAC-r has asynchronous USB input, which is a means of minimizing jitter. Prior to the popularization of async USB input, a jitter minimization method commonly used in USB DACs was "asynchronous sample rate conversion" (ASRC). After async USB input became popular, the old ASRC DACs developed a reputation for not sounding as good as the new DACs with async USB input. So what Rega's marketing people are actually trying to say in the statement Krzysztof quotes is "We don't use no stinkin' old-fashioned ASRC to minimize jitter, we've got async USB input like a good modern DAC should."

 

The DAC-r, like every other delta-sigma DAC, interpolates and filters to obtain a PCM sample rate of 352.8 or 384KHz, then runs the bitstream through sigma-delta modulators to get a DSD-like format at MHz sample rates before the final conversion to analog.

[Jud]

Thanks Jud, now it is crystal clear to me ;-) They used XMOS in the USB implementation as asynchronous. Does it mean for sure that is independent of the PC clock?

 

 

Sent from my iPhone using Tapatalk

 

Yep, that's what it means.

[Jud]

Would not spend $$ on a converter unless you knew for sure it would be better and I think that's doubtful.

 

Intona - maybe. Depending on your budget you may also be interested in something like an Adnaco S3B (optical USB) with an LPS to inject clean power to it.

 

 

Sent from my iPhone using Computer Audiophile

[Jud]

HI JUD,

 

do you know for certain if usb on musette is definitely galvanically isolated? the manual says coax inputs are which is another reason i wanna go that way. the email from metrum was a little confusing so not sure about usb here. funnily enough i will get a musette in april funds allowing but i only have the budget to implement one input well. my confusion about the usb is making me think sell intona go fully coax. some reviews seem to favour coax plus my dealer told me usb will give me 95% of coax. something tells me coax here is more suited to high quality music reproduction. any thoughts please. if i knew the usb on musette was not isolated for sure i would probably use my intona but i dont want to risk double isolation as that can cause unwanted distortion so i read. im thinking of hiface2 convertor if you know it?

 

What is the highest resolution the Musette will accept, USB and coax?

 

 

Sent from my iPhone using Computer Audiophile

[Jud]

Thanks, pabe12. musickid, what sort of computer do you use with your audio system and what operating system(s) is it running?

[Jud]

The S/PDIF implementation in the Metrum DACs is specified for 75Ohm; if you use a cable with a different impedance then the transmission between source and receiver is not perfect and you risk the occurrence of signal reflections on the transmission media (the cable). These reflections look like (attenuated) signals, and could blunt the capability of the receiving side to properly detect the bits with correct timing and value. Thus leading to increased jitter and other nastiness.

 

The S/PDIF implementation is specified for 75 ohm, period, I believe. It's good that Metrum follows the spec.