[chesebert]

I use the Optilink connection with my DA2, which is about the best you can do for electrical isolation. Streaming via ethernet is the noisiest but, I think simply having the processors inside the DAC for ethernet, and to a lesser extent USB, probably generates noise.

Too bad NS1 didn’t work that well for me. Transients were too sharp which suggested there could be digital noise on the signal - probably already riding on the signal before being converted to optical.

 

[IanB52]

Too bad NS1 didn’t work that well for me. Transients were too sharp which suggested there could be digital noise on the signal - probably already riding on the signal before being converted to optical.

I've noticed the NS1 does change its sound with different power cables, conditioners, grounding etc, ethernet cables, switches, even with the Optilink. I still find it to be the cleanest connection, though, so it's more likely noise from (or to) the NS1 via the AC line.

 

[chesebert]

I've noticed the NS1 does change its sound with different power cables, conditioners, grounding etc, ethernet cables, switches, even with the Optilink. I still find it to be the cleanest connection, though, so it's more likely noise from (or to) the NS1 via the AC line.

I tried all kind of power cords even bought an Emm Labs one. I tried different conditioners including a full isolation transformer. Had to give up after some level of craziness

Probably some AC noise issues elsewhere. I wished NS1 had a better PSU so I didn't have to deal with all that. So probably my issue not NS1.

 

[Whazzzup]

so frank? getting closer to 500 hours? how's the mpd8 burn in

 

[F208Frank]

The MPD8 is likely my favorite item in the chain, if not that, then the rowland 555 amps. System is sounding absolutely wonderful.

If you are ever in NY, swing by.

 

[skids929]

The MPD8 is likely my favorite item in the chain, if not that, then the rowland 555 amps. System is sounding absolutely wonderful.

If you are ever in NY, swing by.

I might be getting a chance to hear the MPD8 against my TotalDAC Triunity in the same room same system in the next month or so. This could be very interesting.

 

[chesebert]

I thought I knew what I was doing after all these years and all the time spent learning this stuff but I learned something new today about PBD and Emm Labs. The best way to actually use these native DSD dac is to feed it DSD whether native or converted. But not any DSD. DSD64 doesn’t quite cover auditable frequencies and DSD256, 512 and 1024 actually degrade sound because of the need to employ highly sophisticated and complex noise shaping algorithm by the dac. The best input file format is actually DSD128 and all PCM would likely sound much better if first converted to DSD128. My streamer does this conversion in real time and my Emm Labs and PBD have never sounded so real. I learned this by accident when I was half listening to Andreas Koch drone on in an interview.

 

[Whazzzup]

Easy enough in roon. Chord it’s the opposite, it converts dsd to pcm so dsd is an extra conversion. However the effect is negligable, much like don’t use xlr out to balanced amps, it also an extra conversion, and also a non event

 

[chesebert]

Easy but counterintuitive. I vaguely remember I read somewhere that MPD8 converts all PCM to an intermediate DSD as a first step and then upsample to 50mhz (45.2mhz to be exact). But at the time I never understood why the initial PCM to some version of DSD conversion mattered. I still don’t really understand why it makes a difference but this initial conversion step absolutely works. Would be great for someone else to verify if they have Emm Labs or PBD (other than MPD8 because it’s already doing this intermediate step).

 

[IanB52]

Easy but counterintuitive. I vaguely remember I read somewhere that MPD8 converts all PCM to an intermediate DSD as a first step and then upsample to 50mhz (45.2mhz to be exact). But at the time I never understood why the initial PCM to some version of DSD conversion mattered. I still don’t really understand why it makes a difference but this initial conversion step absolutely works. Would be great for someone else to verify if they have Emm Labs or PBD (other than MPD8 because it’s already doing this intermediate step).

I've heard similar things, but it is surprising that some software would do a better job than Playback/EMM's algorithms, since that is their primary game. Could be that it moves the processing noise out of the DAC.

I do know that not all DSD upsamplers are equal. The much lauded HQPlayer never failed to sound sterile and PCM-like with my old Mytek, no matter what filter/setting. Not the case with the SoX algorithm in Audirvana.

 

[chesebert]

I've heard similar things, but it is surprising that some software would do a better job than Playback/EMM's algorithms, since that is their primary game. Could be that it moves the processing noise out of the DAC.

I do know that not all DSD upsamplers are equal. The much lauded HQPlayer never failed to sound sterile and PCM-like with my old Mytek, no matter what filter/setting. Not the case with the SoX algorithm in Audirvana.

Do you know if DA2 does 1 step conversion or they also use an interim DSD encoding and then feed they into a separate DSD upsampling algorithm?

I am using Shanling ET3 - I liked it better than NS1

 

[IanB52]

Do you know if DA2 does 1 step conversion or they also use an interim DSD encoding and then feed they into a separate DSD upsampling algorithm?

I am using Shanling ET3 - I liked it better than NS1

My understanding is they upsample to a really high PCM rate, do some proprietary filters, then to DSD and to the converter at a super high rate. It is a bit opaque, but I know that much is true.

Playback seems mostly the same, but slightly different filter algorithm, and they downsample back to 2x DSD instead of outputting 16x.

 

[chesebert]

My understanding is they upsample to a really high PCM rate, do some proprietary filters, then to DSD and to the converter at a super high rate. It is a bit opaque, but I know that much is true.

Playback seems mostly the same, but slightly different filter algorithm, and they downsample back to 2x DSD instead of outputting 16x.

PBD MPD8 will convert PCM to DSD128 using dynamic filter and then that interim DSD will then be upsampled to DSD512 for processing. I didn’t know they then down sample that to DSD128 but I don’t think it matters TBH given the a simple LPF will be sufficient in either case.

Anyway, would be great if you could try convert PCM to DSD128 outside the dac and see if you get better sound.

 

[chesebert]

I thought someone here would find this amusing. I am not stupid to buy 2 of the same dac for giggles. The one on top is used as streaming interface with Playback Link output and the one on the bottom is the dac.

The streamer and PC are connected to the one on top via AES and USB and then the top Merlot is connected to the bottom Merlot via Playback Link (fiber). I also use the top unit as headphone amp, which is very good as I said before. Just to be clear fiber sounds the best.

I don’t know if the fiber link also works between Merlot and MPD8 but I bet it would work as well.

Pretty neat stuff

Edit: For the couple people with Emm Labs gear, I also tested the fiber with Emm Labs and the encoding of the two brands are not compatible.