When playback NativeDSD or HQplayers' DSD512, R2R ladder (Holo spring dac) vs pure 1 bit converter (T+A DAC8 DSD), which is genuine?

http://www.computeraudiophile.com/blogs/miska/building-new-server-multichannel-processing-702/

Also, check Jussi’s advice.

No it’s not quiet. That’s why it’s in a closet in another room connected by Ethernet to a microRendu, which is silent.

Yes I see.

The Holo Audio Spring had 4 resistor networks. Two handle PCM and are R2R. Two handle DSD in a similar fashion to Jussi’s SC-1. Within each group one network performs the processing and the second is used to correct linearity.

The surface mount PCB with the resistor networks is double sided, repeating on the underside:

2 Likes

Interesting, So mircrorendu supports DSD512 in Holo spring DAC?

For T+A users, Amanero drivers of microrendu still cannot support T+A’'s DSD512.

So I have no chance to use the mighty microrendu but Intona+Uptone regendu instead.

Well done, the picture you upload is valuable for Jussi to make further comment :slight_smile:

Guess he is more active on CA :wink:

Yes. Using the NAA app. Both native DSD and DoP. I’ve been listening to native, but it’s important to mute the pre-amp before pressing Play, nasty pop at volume.

Wow! Now I’ve more reason to up grade my Zotac i5 fan-less Mini PC to a more powerful ones. I was thinking of this:

Not sure it is up to its task to up sample to DSD256 or even DSD512 from Redbook?

The loud ‘pop loud’ seemed to happen when Holo Spring is used in Linux OS, I read from Computer Audiophile, but Windows with the Holo Spring driver shouldn’t have any loud pop, probably some low level clicks when switch tracks.

If I turn off CUDA I get some pops with the i7 7700 alone. CUDA makes all the difference, so a case with a slot made sense for me.

I’m not sure about Rockna wavedream DAC, there’s no mention in the website that they can do direct DSD conversion in the R2R converter. Is there’s a possibility that Rockna convert DSD to PCM to the the R2R ladder DAC? If there’s the case, there will be little SQ difference between different DSD sampling rate. Again this is just speculation but it will best to ask Rockna directly.

What is CUDA? It is some kind of hardware accelerator?

I can confirm that Rockna uses FPGA and cannot do direct DSD512

(since it is mentioned by someone on some thread in China, the person sounds like an technician-ish guy connected to Rockna)

There are even some nasty commercial review regarding Rockna vs Holo spring dac on Chinese Hifi thread. (Better not discuss it further, becasue it’s quite sensitive)

That’s why I tell you this is a typical example of meaningless DSD512 on R2R dac. Because in practice, DSD512 brings no SQ improvement.

There is no DSD512 recording, so upsampling is the only way to get DSD512, but does no-SQ-improvment DSD512 is the purpose from HQplayer? I don’t think so.

CUDA offload, is a feature when your CPU is not powerful enough, by using the Nvidia graphic card’s CUDA floating point to ease the load from CPU.

So in my case, i7 3770 cannot run DSD512 alone but with the help of my graphic GTX 780 which has CUDA feature, can smoothly run the DSD512.

Thanks @Wei_Xia1, looks like I need to build a custom designed PC with i7 and Nvidia GTX-780 graphic card to do DSD512. This is going to be tough. Any off selves ready made PCs that may contain the above components that I can purchase directly?

Sound gallries SGM is a beast audio PC for T+A DSD512, but for that price, I’m not going to pay for it even my wallet is heathy enough. It’s just my personal opinion or my personal attitude towards what kind of things I should splash out or not.

GTX 780 is not good enough I’m afraid, flagship serious double-precision graphic card like Titan is a good choice. But please note that HQplayer relys on CPU more than GPU, so a powerful CPU is crucial. (Jussi is a nice person, mentioned the setup lots of times since so many users ask him the same questions from time to time)

The Noise from the fan is also a factor to consider, CPU water cooling system or low-speed, low noise CPU cooling fan is the way to go.

Check the link of Jussi’s recommendation.

http://www.computeraudiophile.com/blogs/miska/building-new-server-multichannel-processing-702/

Ive heard many water cooling systems are made inching and not of great quality or leakproof. I would stick to passive or large area slow fan options.

the newer 1070 cards are just off the high end prices of the 1080 cards and should be adequate I was told. YMMV

You will also be needing a good smart fan power supply 550-600W+ and a fairly sizeable case to fit all of this in too…no SFF options I’m afraid.

1 Like

I agree re T + A DAC8 DSD and higher rate DSD. Loving it!

Good to know, thanks.

BTW, any idea how to build a fan-less PC? :slight_smile:

You might consider opening a new thread for that

@David_Christie I’m curious what Dsd you are using and if its native files you are playing or upsaming ?

I’ve started a thread here…I hope others will benefit and or chime in

1 Like

Hi Jaspal. I upsample PCM to DSD 256 into the T + A DAC DSD.
I don’t upsample native DSD files as Roon converts DSD files to PCM then back to DSD.
I don’t have much DSD material and this “passes through”.

I also use HQPlayer an upsample to DSD512. The HQPlayer NAA is a bit flaky, however… loses the DAC if the DAC is put into standby… have to restart the NAA (and then often HQPlayer) which is annoying…