Meh - if D900 is like D90 III Discrete that Jussi measured, then both don’t belong in this thread lol
1 Like
At the moment D900 has same issue as D90 III.
It is probably an issue with the format conversion chip they use to produce output for the D/A stage.
3 Likes
Very disappointing on the D90 Discrete and D900, lots of positive reviews on Sound quality. They were on my list for a new DAC to pair with my Denefrips ARCE and start up sampling to DSD using HQ player. You would think Topping would test this before making claims. Holo Cyan 2 back to top of the list. Has anybody tested to see if Audio GD Nos is bit perfect with DSD?
Not exactly alone though are they. Everything seems to arrive with poor firmware now with an immediate update necessary once you plug it in. Even then there seems no guarantee that any device will be stable or meet its marketing claims.
Well it may not be ideal for HQP with DSD Direct but that doesn’t mean its sound quality cannot be excellent.
Us here will pass but other people can probably enjoy high fidelity
Sorry to flash back to your late 2023 postings about the Gustard A26. I have three A26’s built in late 2024 with v1.44 firmware that I use for three-way stereo active crossovers, feeding 24-96 PCM to the DACs via AES from an interface. I forget which NAA is included in the firmware. But I recall @jussi_laako telling me last year that it is not up-to-date enough to send directly from HQPE to the three DACs separately via NAA. Certainly, on my German AVM-Fritz modem-router, only one of the DACs appears at any one time anyway.
I see from a September 2023 post by @Stef_Dahl that they know how to update the NAA of the A26 separately.
Any advice on how to do this? And which NAA file (link) I should use for my purposes?
Of course, advice like, “Don’t bother, won’t work or might just brick your DAC”, would be followed first!
Note: Another reason for wanting to NAA to the three DACs at the same time would for surround DSD256 playback (5.1=2x3).
Thanks much!
PS: Jussi, I am still working on Inferno with its developer… Not there yet though…
Reason for this may be that they probably have the same name. Each NAA device within the same network must have a unique name. On NAA OS this is auto-generated. Also Sonore’s rendu-series does similar thing. But many other third party solutions use a static name, while they should at least let user customise it. If name is not explicitly defined, device’s hostname is used by default.
Unless the DAC clocks are somehow synchronized, this may not work properly…
If you use AES for example, they are all slaved to the same source clock. I don’t know if A26 supports DoP over AES, but if it does, you could do six channels of DSD64 from HQPlayer this way to ensure all DACs stay in sync. If there would be multichannel I2S output device, you could do more (I2S is slaved to the source device just like S/PDIF and AES/EBU).
2 Likes
Just as a FYi, When I was choosing DACs, the UD-701N was the best I heard at remotely its price point.
1 Like
I have UD-701N on my office rack along with Spring 3, Gustard A26 and Musical Fidelity M3x DAC.
Is this a DAC swinging contest? 
2 Likes
Remember to also enable DAC correction for the UD-701N. It is also important to pay attention to the DAC settings. DSD D/A conversion filter should be set to FIR1 or disabled, FIR2 is objectively so bad that I would recommend avoiding it (one could call it a “tube-amp simulator”).
Managed to find my 4.4mm to XLR cable (from AliExpress for $30 
).
Most measurements are the same as my RCA (with super careful grounding…) except multitone which has around 105dB ‘range’ - nice improvement over RCA ! Surprised me
One thing I didn’t show with RCA since I wanted to wait for XLR measurements, was that there is a little bit of noise floor modulation. This is with DSD256 , gauss-xla, ASDM5 and ASDM7 both are the same in terms of noise floor modulation.
Here you can see 0dBFS signal vs -120dBFS
Its small but this is absent in the AH90 with built-in NAA input.
Still the iFi Zen DAC 3 is easy to recommend - supports 44.1k and 48k base rates. Good performance out of the box with DSD256.
I suspect measures the same as Neo iDSD for much less money? I haven’t measure Neo iDSD yet
5 Likes
Thanks for the effort. I picked one up, sounds good to me. I’m trying its headphone output, variable switch is on.
I’m using the volume knob on the unit.
I wonder if technically its better to set it to “fixed” and use HQP for volume control with the zen 3 headphone instead.
SDM 256, ASDM7EC-fast, gauss-hires-lp/gauss-long
1 Like
I returned the DAC and didn’t measure the headphone amp - I’m using one of the best measuring headphone amps with my AH90, this one .
The DAC’s built in amp has output impedance < 1 ohm, which is a great start.
iFi’s specs for the DAC match my measurements (which is nice), so if their amp spec is accurate (the amp of the DAC I am talking about), it will degrade the DAC’s performance.
The amp I use has lower noise and distortion than the iFi DAC, which is not something I ever thought I’d write. What a time lol
I had the newer Topping L30ii head amp. Lots of power and meaured as good as yours (L50). It ran hot, both ps and the unit, but not too hot. I returned that and got the Zen 3. It might not measure as well as the Topping gear, but so far the Zen 3 works fine with HQP using its head amp or fixed out to my Naim pre amp.
I like the fact it is no fuss direct DSD and sounds good to me.
The path I’m using is: Ultrarendu NAA/Intona isolator/usb c input of Zen 3.
Interesting, have my hand on my L50 right now and it is barely warm to touch.
Good passive cooling inside.
From the well respected “solderdude” on ASR forum "No encapsulated stuff.
No ESD issues. Properly functioning DC protection." - which he points out cannot be said for some other Topping and SMSL amps…
Yes easy to recommend it based on my measurements , although I did update its firmware to latest - there was something about 44.1kHz/48kHz rate change improvements which I thought maybeee could affect my DoP measurements (I connected direct to my Mac + USB isolator for measurements)
I should have measured with old firmware but had limited time with it.
You should ask Jussi for a DAC correction for it 
I might actually buy one later myself for travelling, connected to Macbook
What is powering the ultraRendu?
Since you have an Intona connected (literally isolating the Rendu) you could probably sell it and get an RPi4 and run the more optimal NAA OS
The rendu is powered by a Sbooster lps. I have 2 rendus, 1 optical rendu, and a pi 4. I picked the rendus up second hand over the years. I have used the pi4 with naa os, never noticed a difference between them when using an isolator. I have the pi4 in use now as a upnp server running Asset Upnp.
The Zen 3 is on latest firmware 1.10. The optional firmware 1.1c, is something about rate changes. Not sure it matters with DSD direct usage?
Well my measurements were DoP ( = DSD over PCM - which is still DSD Direct in this case). I used firmware 1.1c. Maybe not necessary? Didn’t have time to overthink it , had to measure quick lol
Maybe doesn’t matter for your use with Rendu/Linux
I’m not sure. I will read the 1.1c update docs again. Right now it is working and sounding great which is the main thing. And, as you’ve mentioned you don’t have to mess with special firmware, DSD direct just works.
Yeh since you are not using DoP, nothing to worry about.
Enjoy !
1 Like