Best native DSD DACs for use with HQPlayer?

HQPlayer doesn’t pass any clocks through, so as long as the input side can lock into the source, it works. IIRC, RME’s PLL can lock into up to 100 ns of jitter. DAC runs on it’s own clock, unaffected by the input.

Of course one can also have for example input at 96k rate and output at 44.1k x256 DSD for example.

1 Like

No, I understand the limitations of the analogue output and therefore would never use a(ny) WiiM device as a DAC for anything but a casual application and have little interest in analogue performance. That being said, ASR’s test of the Mini using Toslink output looked very good: ASR WiiM Mini Test

The tests that flagged my attention, revealing substantial periodic jitter when the digital outputs of various WiiM devices (particularly the Mini) were routed into a Holo Spring 3 (with no PLL) are here: WiiM Jitter J-Test

My results show how good the cheap AH90 DAC does with incoming jitter rejection

It is not bothered by the Wiim Mini at all

Would be the same if you turn PLL on with the Holo with the Wiim Mini.

Well, I’d say it shows how well HQPlayer does incoming jitter rejection. :sweat_smile:

From DAC’s point of view it is no different whether you play from WiiM or from a local file, or from Qobuz.

If you connect WiiM Toslink directly to the AH90 then it may be different case…

Still not as good as through HQPlayer… :wink:

3 Likes

I have this now. It is hot garbage. I’m sending it back to the owner and telling them to get an Amazon refund.

Its DSD Direct input is broken

I hope SMSL get back to you and fix it

They should be testing DSD256 input not DSD512 . It is obvious they are not

I spent 20 minutes setting up for this measurement. I’ll never get this time back :sob:

1 Like

Are you using the headphone amp of the AH90? If yes, do you like the amp section?

Edit: No need to answer, found the answer in a different thread.

Question:

I wonder how you experienced guys would consider Gustard A26, compared to RME ADI-2 DAC fs and NAA.

I cannot afford the T+A. Since both RME and Gustard are chip based DAC’s I am curious about how A26 sounds as HQP endpoint with DSD Direct DSD256. Is it same character, do I gain anything with Gustard in terms of the extra resolution. Will SINAD improve with DSD Direct from HQP as RME do?

I will off course run it with the NAA. Therefor I wonder about how far Gustard have planned the updating of NAA. Anyone knows?

You should be checking out the Holo Cyan 2 then, imho. I have an ADI-2 Pro FS R Black Edition and I think it sounds better than the A26 I demoed. Direct connection not using NAA.

2 Likes

No, I should not. Jussi have already explained the difference between my RME and Holo, and R2R is not my way. I want the chip DAC, as warmth is not my cup of tea. I want transients like shots fired, like Jussi desribed the T+A.

Thanks for suggestion though :slight_smile:

Can you post a link to where Jussi said what he said?

It will take some time to go through the threads. I recall it was in conjunction with Jussi posting measurements of the HOLO Cyan. The SINAD was very equal to T+A, some -140 dB, and I asked if they were competitive, but Jussi replied they were different in sound. While Holo was a bit warmer as most R2R, the T+A was lean and very tight.

Audalytic AH90 (it is Gustard). You can use your own preferred 12Vdc linear PSU (DC input).

The built-in powersupply is also clean.

Features built-in ethernet NAA, like A26. Less than half the price.

AH90 is cleaner under 100kHz than Cyan2. And definitely no early roll-off for warmth.

Like A26, AH90 will not come from factory with DSD Direct support. You get the firmware from Gustard and DIY.

Like A26, AH90 true DSD Direct is limited to DSD256 (AK4499EX)

I think you will like it, if you liked RME ADI-2 (AKM). It is cleaner than the RME under 100kHz.

No need for USB isolators. Just use the ethernet NAA input.

2 Likes

Cyan 2 is definitely better than RME or Gustard. Don’t worry about what Amir says about Sinad.

I switched from the RME ADI-2 Akm4493 DSD Direct to the Cyan 2.
I don’t regret my choice, it’s better in any case and I absolutely don’t find it warm…
I use Hqp DSD 256

2 Likes

If transients are your thing then you should try the Border Patrol DAC. @ipeverywhere has both that and the T+A 200 and while the T+A is better in most every way, the BP DAC brought transients as he’s never heard before. I tagged him so he can either add to or correct me. :smiley:

1 Like

I wouldn’t go that far :slight_smile:

I said it excelled at transients over everything else its supposed to do as a DAC and that’s probably why the thing gets high praise. I still prefer the ADI-2 (AKM) over the BorderPatrol overall. BorderPatrol is built like a tank though and is a nicer looking bit of kit. I’ve not tried the new -fast modulators with the ADI-2 but they have certainly brought new insights into what I’m hearing on the DAC200. I don’t think it makes any sense to use HQP with the BorderPatrol (although I will try). The BorderPatrol won’t do DSD and somewhere near 16 bits of depth at best.

If you want to use DSD, and everyone should so call me a fanboy, stick with the DACs in this thread known to do DSD correctly.

4 Likes

I’ve shared measurements, Cyan2 rolls off early and does sound too warm to me. Maybe the early roll off.

I never used DAC Corrections though, probably Jussi has helped this.

Cyan2 has a much bigger IMD vs level hump than RME and my AH90 (4499EX like A26), which I don’t think can be helped with DAC corrections

AH90 has better measurements than Cyan2 under 100kHz. Basically all measurements

Under half the price

1 Like

This rolloff is due to NOS operation/lack of oversampling.

If you’re using HQP @ high sample rates you won’t have rolloff (unless the filter you’re using does)

1 Like

Nope its due to a fixed frequency analogue filter :wink:

We’re in the HQPlayer thread so I obviously measured with DSD256, not HQPlayer in NOS mode.

All my measurements are above, I measure frequency response with sinc-S filter…

It is very good but AH90 even better at less than half the price. With ethernet NAA input. And use your own linear PSU if you want.

1 Like

I do not worry about Amir at all. The measures I am refering to were Jussis own. The measures do not say everything about SQ, but it is an important part of the puzzle. The problem is noise at complete system level. Devices interacting with noise, impedances, resonances and ground management. I try to choose devices that themselves have very good SINAD, and prepare transmissions between them in the most noise cancelling way possible. My Hypex Nilai also display excellent SINAD, and cutting ground lead between them, ie pin1 and chassis ground was very beneficial. Now the track noise is so easy to spot, as the digital black between tracks is dead silent. I will further explore the common mode block over the balanced interface by skipping ground and shield completely and try a stripped antenna coax cores twisted pair as interconnects. That would lower the cable capacitance to a neglectible level, I hope, providing for even faster transient response. HiFi is fun … :wink: