GUSTARD DAC-R26 Balanced Decoder R2R+1Bit Dual Native Decoding Music Bridge

[BShaw]

Yes. I leave my OCK-1 on all the time.

This has completely solved the issue, thanks guys. My GF, used to seeing me crawling around the rack, often muttering, asked me the other day “you know when you just go to the store and buy a regular stereo for $200….why do those just work without ever having a problem but you are always fixing things?”

 

[BShaw]

Quick wit that I am, I said “that’s life on the edge, baby.”
She didn’t seem too impressed by my explanation.

 

[Forgisound]

This has completely solved the issue, thanks guys. My GF, used to seeing me crawling around the rack, often muttering, asked me the other day “you know when you just go to the store and buy a regular stereo for $200….why do those just work without ever having a problem but you are always fixing things?”

$200 Sony, Pioneer, etc. SACD players work forever without fail. But all expensive SACD players break down. True, except Accuphase, but hardly any other can last 5 years.

 

[camrector]

This has completely solved the issue, thanks guys. My GF, used to seeing me crawling around the rack, often muttering, asked me the other day “you know when you just go to the store and buy a regular stereo for $200….why do those just work without ever having a problem but you are always fixing things?”

Ugh

 

[Odezra]

Another review … sort of..

 

[camrector]

Received a follow message from Gustard answering a few more questions I had and( @labeelen)ignoring another. The language barrier is tough.
This really was the only useful info I got out of the email.
Confirms what we have gathered so far.
Still a lazy way out for DSD, when we have a FPGA present also and everything could be done in that. Just convert to 6bit PCM, upsample, dither and noise shape and DSD out. Exactly how Denafrips does it.

 

[Odezra]

Received a follow message from Gustard answering a few more questions I had and( @labeelen)ignoring another. The language barrier is tough.
This really was the only useful info I got out of the email.
Confirms what we have gathered so far.
Still a lazy way out for DSD, when we have a FPGA present also and everything could be done in that. Just convert to 6bit PCM, upsample, dither and noise shape and DSD out. Exactly how Denafrips does it.

Really useful diagram.

confirms the logic we had on PCM.

However, I think it changes things up a wee bit on the DSD side. My reading is:

- If file is detected as DSD, and DSD Direct is off, the file goes through the 1 bit DAC and into output stage
- If file is detected as DSD and DSD direct is on, it bypasses the 1 bit multidac and goes straight to output stage

Am I right here?

If so - there are three pathways to the output stage. PCM (whether native or upsampled) through the R2R ladder. DSD (with DSD Direct off, native or upsampled) going through the 1 bit DAC, and DSD (where DSD direct is on, native or upsampled) which bypasses the 1 bit DAC altogether.

 

[camrector]

Really useful diagram.

confirms the logic we had on PCM.

However, I think it changes things up a wee bit on the DSD side. My reading is:

- If file is detected as DSD, and DSD Direct is off, the file goes through the 1 bit DAC and into output stage
- If file is detected as DSD and DSD direct is on, it bypasses the 1 bit multidac and goes straight to output stage

Am I right here?

If so - there are three pathways to the output stage. PCM (whether native or upsampled) through the R2R ladder. DSD (with DSD Direct off) going through the 1 bit DAC, and DSD (where DSD direct is on) which bypasses the 1 bit DAC altogether.

Yea that part is a bit confusing
I’ll explain my best.
Pcm or DSD enters the DAC. The first stop identifies if it’s PCM or DSD. At the bottom is the “and logic” circuit. That tells the first stop if DSD is detected and the DSD Direct is on, to send it to the DSD dac.
If that’s not switched “on” DSD is converted to PCM and follows the PCM path.
The “and logic” circuit also connects to the 2:1 multiplexor The 2:1 mux needs a control signal to tell it which analog signal to output.

 

[camrector]

@labeelen An “answer” to your question.

“It could be that the DSD upsampling brings too much out-of-band noise and is refracted back into the audio band on the back-end equipment”

This could be true if you were not upsampling to a high enough rate or possible from any added filters or noise shaping HQplayer is doing.
What I think is happening though is more from Gustard using a 1bit Dac. SINAD measurements were very bad.
This paper explains many reasons why.
https://timbreluces.com/assets/sacd.pdf

 

[Odezra]

Yea that part is a bit confusing
I’ll explain my best.
Pcm or DSD enters the DAC. The first stop identifies if it’s PCM or DSD. At the bottom is the “and logic” circuit. That tells the first stop if DSD is detected and the DSD Direct is on, to send it to the DSD dac.
If that’s not switched “on” DSD is converted to PCM and follows the PCM path.
The “and logic” circuit also connects to the 2:1 multiplexor for linking input and output stage.

Thanks Cam - this is helpful and makes sense.

So that leaves two paths - PCM through R2R ladder (for PCM native, PCM upsampling, or DSD native / upsampled which is converted to PCM when DSD Direct is off), and DSD direct through 1 bit dac (for DSD files, native or upsampled, where DSD Direct is on).

Interestingly, i installed HQ player remote over the weekend on my phone (as a/b testing Roon upsampling and HQ filters was a real challenge when the Mac Mini is on other side of house). With HQ player, I am not getting any pops for DSD files when DSD direct is off, per our discussions previously.

Also - HQP remote is a super handy way to quickly change between DSD / PCM and any range of filters to determine what you like best. Have ended back on PCM with this (for awhile at least). Instruments are more realistic, however, it comes at the expense of a slight veil which is not there with DSD upsampled files.

EDIT: it is so good (for me at least) to understand how exactly the DSD direct function is working. thanks @camrector for chasing gustard on this!.

 

[camrector]

Thanks Cam - this is helpful and makes sense.

So that leaves two paths - PCM through R2R ladder (for PCM native, PCM upsampling, or DSD native / upsampled which is converted to PCM when DSD Direct is off), and DSD direct through 1 bit dac (for DSD files, native or upsampled, where DSD Direct is on).

Interestingly, i installed HQ player remote over the weekend on my phone (as a/b testing Roon upsampling and HQ filters was a real challenge when the Mac Mini is on other side of house). With HQ player, I am not getting any pops for DSD files when DSD direct is off, per our discussions previously.

Also - HQP remote is a super handy way to quickly change between DSD / PCM and any range of filters to determine what you like best. Have ended back on PCM with this (for awhile at least). Instruments are more realistic, however, it comes at the expense of a slight veil which is not there with DSD upsampled files.

Correct.

The lack of a pop noise with DSD direct off is a dead give away that your signal has been converted to PCM.

Good info here from HQ creator and Roon tech

https://community.roonlabs.com/t/loud-popping-noise-when-starting-new-track-native-dsd-only/11559

 

[camrector]

Thanks Cam - this is helpful and makes sense.

So that leaves two paths - PCM through R2R ladder (for PCM native, PCM upsampling, or DSD native / upsampled which is converted to PCM when DSD Direct is off), and DSD direct through 1 bit dac (for DSD files, native or upsampled, where DSD Direct is on).

Interestingly, i installed HQ player remote over the weekend on my phone (as a/b testing Roon upsampling and HQ filters was a real challenge when the Mac Mini is on other side of house). With HQ player, I am not getting any pops for DSD files when DSD direct is off, per our discussions previously.

Also - HQP remote is a super handy way to quickly change between DSD / PCM and any range of filters to determine what you like best. Have ended back on PCM with this (for awhile at least). Instruments are more realistic, however, it comes at the expense of a slight veil which is not there with DSD upsampled files.

EDIT: it is so good (for me at least) to understand how exactly the DSD direct function is working. thanks @camrector for chasing gustard on this!.

Totally
I’m happy we can put this one to bed. Not to happy about what we found though. Up until two weeks ago I thought I was listening to the purest DSD recordings NativeDSD has to offer.
But no. R26 was converting to PCM.
Its Sneaky too that the DSD sample rate is still shown on the R26 screen after a PCM conversion has taken place.

I read more on DSD in the past few days than really anyone should.

 

[Forgisound]

R26 gives the possibility to choose DSD direct, where the DSD track is played. Also, if we turn off DSD direct, the DSD track will be converted to PCM. I don't see a problem. Many other R2R DACs convert DSD to PCM without exception.

OK, with DSD64 there is noise, but with higher resolutions there is no noise. That's what Gustard says, with a recommendation for DSD upsampling.

 

[Odezra]

Totally
I’m happy we can put this one to bed. Not to happy about what we found though. Up until two weeks ago I thought I was listening to the purest DSD recordings NativeDSD has to offer.
But no. R26 was converting to PCM.
Its Sneaky too that the DSD sample rate is still shown on the R26 screen after a PCM conversion has taken place.

I read more on DSD in the past few days than really anyone should.

I hear ya. I think the marketing blurb could be much clearer as to what's going on.

The implementation of the DSD direct feature feels a bit slapstick, given the SINAD results, in my view.

On the other hand - the selling feature is the R2R dac, so to use Direct DSD (and bypass the R2R DAC altogether) seems to negate the point of purchasing the machine.

That all said - it's amazing that the Gustard has the R2R ladder and the 1 bit DAC - it allows us to try all the permutations and figure out what best suits our tastes / systems. Sure the 1 bit DAC could have less noise - but the differences are still on the small end of town (for me anyway upsampling to DSD512).

 

[camrector]

I hear ya. I think the marketing blurb could be much clearer as to what's going on.

The implementation of the DSD direct feature feels a bit slapstick, given the SINAD results, in my view.

On the other hand - the selling feature is the R2R dac, so to use Direct DSD (and bypass the R2R DAC altogether) seems to negate the point of purchasing the machine.

That all said - it's amazing that the Gustard has the R2R ladder and the 1 bit DAC - it allows us to try all the permutations and figure out what best suits our tastes / systems. Sure the 1 bit DAC could have less noise - but the differences are still on the small end of town (for me anyway upsampling to DSD512).

100% agree
If we ever have a sticky here “Upsample your DSD” or “DSD256 native bare minimum” should be near the top of the list.
I’m going to have to dive deeper now into HQplayer. I’m intrigued