HOLO Audio Spring DAC - R2R DSD512

[luisma]

21 hours ago, barrows said:

But the DAC does matter in this for sure.  This difference is more marked in DACs designed really well for DSD, like ones using DSD direct mode on an AKM chip, or really good discrete DSD converters with no processing other than conversions

This indeed matters, and I just have to add that in addition to the internal digital conversion / processing, the last output analog conversion stage on the DAC is critical as well.

 

21 hours ago, barrows said:

This kind of noise may be responsible for some of the differences people hear between different DACs: the noise goes to the amplifier, or preamp, and may create problems there,

As you well say once noise is injected you can't get rid of it, and this is why the last analog conversion stage (including linearity, analog filtering, signal amplification etc) is so important. I know I will be accused of "simplistic" but pre processing the digital signal using computer software (developed by a known Finland based developer) and feeding this to any DAC ($300 - $20000) will yield excellent results as long as the analog stage is well designed. The Holo products it is my understanding are excellent well designed around this analog last stage.

 

21 hours ago, barrows said:

this could be an answer to those simplistic DAC "reviews" at ASR, where they just measure the output of the DAC, and then declare it "perfect"

This comment is brilliant because it addresses exactly that difference between measurements and actual SQ that people notices, but I would not say the entire review is simplistic, I think at ASR they (he) does a good work with the process and his methodology, now the results are sometimes misinterpreted or simplistic and pragmatic conclusions are drawn, yet I personally enjoy reading these reviews and learning a few things along the way and filtering (what I consider) the crap out of it. And my last comment (this one) might be (again) simplistic but I think from the marketing standpoint ASR "has" to offer another "flavor" to its customer base, hence that divide that mostly happens between users at AS and ASR.

 

And I'm completely OT here, sorry

[luisma]

5 hours ago, hols said:

I have been using HQplayer(Taiko extreme Windows) and also HQP embedded(Linux system)to upsample to PCM1411 and I used to set it to 24 bits as T+ is a 26 bit DAC and there is no problem playing PCM1411 with T+. When I changed to May it returns some clearly distorted sound. The sound is normal if I play 705.6 but once I set to 1411 it returns some distorted sound. I tried setting it to 20 bits but sound is still distorted and finally I read through the threads here in AS and realized that Miska has illustrated that one has to turn it down to 16 bits before one can play PCM1411. And that is not the end of the story because if I use the Taiko extreme (set to 16 bits)to play PCM 1411 there is cracked high frequency sound but there is no such sound if I use Linux HQPe. We all know that there is a Windows Thesycon driver that comes with Holo but no Linux driver(usual practice). What's intriguing is that why with the Thesycon driver that comes with T+ there is no problem playing 24 bit 1411 with T+ but with the Thesycon driver that comes with May it can't even play PCM 1411 at 16bit in May? So is it a driver problem or is it the capability of the DAC ?  I still don't have an answer. But it seems it is a real pity that May cannot play PCM 1411 smoothly because when successfully played it gives an excellent sound as I shall briefly outline my impression below.

Thank you @hols for this report and what you describe, me without being an expert or owning the May or T+ but being an owner of a T- :) and HQPe I remember the inconsistencies playing especially when source content switched resolution or format with former Amanero and then Terminator's new board which eventually after some firmware updates were smoothed, now I don't know how frequent are firmware updates on the May but this looks like a good question for Holo's development team.

 

Thanks again, I have to say quite a few people not having the chance to compare all this equipment side by side and throwing in HQP in Windows and Linux is definitely a very good review.

 

[luisma]

5 hours ago, Miska said:

Yes, as you can see above based on measurements.

LNS15 preferred over TPDF? or the difference in the graph is made by 20 vs 24 bits?