Understood, but this suggests that the Intona is not mitigating all noise artifacts going into the Dac, even though measured output looks the same in either case - though maybe someone/@jussi_laako can correct me I am wrong here. To the point where it’s still audible to you. Otherwise the upstream choices in endpoint/power/noise would be irrelevant. This is what I am curious about.
Upstream devices or usb cables have been irrelevant to me when having an intona in the chain. tested the rpi4, holo red, fitlet3 with a farad lps, ipower elite, allo shanti, cui medical grade switching power supply recommended by Jussi.
It is due to the functional and quality problems in their products. Similar with SMSL.
Sorry, but given that I have spent a few thousand EUR already in Topping/SMSL products that don’t function properly, I’m not particularly eager to spend more…
1 Like
There’s me thinking they sent you DAC’s to try.
I think Jussi is the only one here interested in my measurbating but I’ll post here anyway 
For those that have been following Jussi’s generous knowledge sharing for 10+ years (lots of good stuff on Audiophile Style), we know that the un-correlated noise produced by oversampling to 1-bit SDM is not a problem, provided you can’t actually hear it (like a hiss from your speakers/headphones/IEMs).
With my Gustard/Audalytic AH90 one measurement I didn’t do previously was to see run my linear sweep test files created in REW: 0dBFS, -6dBFS, -12dBFS, -18dBFS, -24dBFS etc input signals.
Since I have the same Picoscope, I can see what Jussi has been saying for over a decade, in action for myself.
You can see that as the input signal reduces from 0dBFS to -24dBFS there is no change in the fixed noise humps.
With my very sensitive IEMs, I hear no hiss going all the way down to -80dBFS input signal. So no worries!
In short, as music signals change in levels you don’t get any noise modulation ! No noise floor modulation. Its just a fixed noise that doesn’t change with music signals (hence its called un-correlated noise).
Fixed noise is significantly less problematic than modulating noise floor, assuming you don’t hear any hiss
Like before, all these measurements using the AH90 LAN NAA input.
Btw Gustard confirmed they will be updating the NAA 4.4.0 to v5 - I assume it includes A26 obviously.
6 Likes
You are making me want to buy an AH90 to replace my Zen Stream/Zen DAC combo in the man cave. I already have two other listening setups in that room so I really don’t need to do it. But you are not being at all helpful with this 
4 Likes
I think you mentioned that the external ps could be 9v to 15v. It says 12v on the back of the unit, have you tried lesser voltage? I have a Sbooster 9 - 10v, wondering if it’s ok to use.
Its there on the main product web page.
https://www.gustard.com/?post_type=products&page_id=17831
I only measured with 12V 2A output powersupply.
No SMPS noise visible in all the < 100 kHz measurements.
I might check how the built in power supply looks
Sorry if this question has been addressed before. Is AK4493 by default direct dsd, optional to bypass dsp or simply not direct?
Thanks
I guess, it is DAC (the whole device) dependent. For example, ADI-2 DAC lets you select if you want direct DSD (with no DSP features like digital volume control, PEQ, etc) or non-direct DSD (you have DSP features) but with a PCM conversion in the way.
1 Like
By default it is not in direct mode, it needs to be specifically configured to enable direct mode.
1 Like
Hi!
I’m also a happy Spring KTE L3 user!
I don’t know about the Mac M4, but it will definitely give you a great start to explore many filters and features of HQplayer.
if you’re curious and like to discover the potential of this DAC, you might end up with an Ubuntu build with i9000 14th gen with a 4070 Super GPU! Well, just in a couple of years!
The DAC correction is another feature on HQplayer, is not an upgrade, this software includes all features discussed here.
We have a desktop and an embedded version, the latter one takes more technical steps, but you can build up your knowledge in this forum, as I did too, but still a bit newbie in many features.
Is highly recommended to use a RPi4 as NAA, so you can get most of your DSD capacity (unlike DoP) and also reduce noise on the audio chain. I’m sure can find where you are!
1 Like
Good news. I get a chance to measure S.M.S.L RAW-DAC1
AK4499EX (like my Gustard AH90) and confirmed to have DSD Direct mode.
Lets see if SMSL have finally implemented AK4499EX DSDD mode properly. And see if >100kHz is cleaner than my AH90 !
<100kHz is already very excellent in AH90 , beating my Cyan2 in most measurements
AH90 has the advantage of built-in ethernet NAA input
I’d be curious to know if you encounter the same issue I’ve been getting. I have the RAW-DAC1 and RAW-PRO-DAC here currently.
On the RAW-DAC1 (the AKM one), on DSD512 all the time, and on DSD256 about 90% of the time, either one or both channels shows massively higher noise throughout the audible band.
These are at DSD512 for example, and the behaviour changes depending on the ‘Soundcolor’ setting for the DAC:
At 256 I can sort of sometimes get it to work on soundcolor 2, and one channel is fixed but not the other for the three other settings:
I did let SMSL know about this but it seems that they can’t replicate the issue, they said they’re seeing everything working fine:
I know @jussi_laako has mentioned he’s had trouble with SMSL/Topping DACs before, not sure if this is a similar issue to one you may have come across before?
I’ve tried various different USB sources to no avail, and the ESS variant doesn’t have the same issue at all so I’m not sure if this is just a dud unit or a wider issue.
PCM playback is totally fine
I wouldn’t bother measuring AK4499EX at DSD512 when the SMSL user manual and AKM data sheet (and Jussi has mentioned numerous times) states (see gray box):
If you see my AH90 measurements above in the thread (also AK4499EX) you will see I didn’t bother doing DSD512 measurements, since the above screenshot is well known. Waste of time.
And I’m personally not interested in measuring the Sound Colour modes.
I’ll be measuring DSD Direct and DSD256x44.1kHz only (I will check if x48kHz is broken but I assume it will be).
As I showed above, AH90 is really excellent under 100kHz in DSDD mode at DSD256x44.1kHz. And has NAA input built-in.
No need to worry about USB isolators with AH90 NAA input (just use unshielded eth cable of course…)
@GoldenSound can you share some of your DSD Direct measurements at DSD256 x 44.1 ?
Select DSD filter = wide
And ASDM7 modulator
sinc-S or gauss-xla filter for measurements (don’t need EC modulator for measurements)
Multitone, IMD vs level (any IMD hump?), jitter.
I can do, but it’s hard to get too much useful info due to the dominating noise issue shown above (which also happens at 256x. The 2nd set of graphs is all x256)
@GoldenSound Your noise issues are at DSD512 (expected) and DSD256 with non-DSD direct mode
Those Sound Colour profiles are only in effect in non-DSDD
Can you try DSD Direct and share those measurements I mentioned above?
I found an issue with SMSL D-6 in DSDD which SMSL fixed with firmware updates
I wonder if we’ll have to do the same here
Disappointing they dont check (if thats the case)
Ah I see what you mean, sorry I should have said, the DAC was in DSD bypass mode. Converting it to PCM instead works fine. But the native dsd mode does not (and the sound color still seems to be doing something). So I’m really confused what’s going on
Interesting if Sound Color modes affect DSD Direct output
Not looking forward to measuring now haha