Fosi Audio DS2 Portable DAC & Amp Review

[mc.god]

Many thanks for your measurements, great job.
The output shutdown after 3 second is the cause of the ramp-up effect when coming back on, my Sonata BHD behaves in the same way, I suppose they are almost identical implementation. I can add that the Sonata, when in "sleep" mode, catches up audible emi interference from my laptop, probably from the wifi or the ssd, producing a low but audible buzz noise that is greatly amplified if i tighten my hand around the iem cable while reducing if i touch the mtal part of the dongle body or the metal chassis of the laptop. Ironically it happens only if the dac is set into low or medium gain, disappearing in high gain (it has 3 level). As soon as something start playing coming alive from "sleep" mode, no more noises at any gain setting, so I can imagine how it can mess up measurements.

p.s.
I just ordered this other dongle, quite different dual CS43131 implementation, we'll see if it shows the same behavior.

Interesting one. More options than Fosi (I tried that one's key combinations but none of them worked with DS2), so it is a different design. Awaiting your report!

The Kuang Pai Player 3 arrived today, and despite the different implementation, it ended up using the same Savitech usb bridge than DS2 and Sonata BHD, showing the same behavior about the "ramp-up" after going in "powersave mode".
Just testing now, in Linux audio properties panel -> speakers test , clicking on "front left" you can only hear "ont left", then after 5 second clicking on "front right" you can hear "ront rght", so right channel come up a fraction earlier than left one (panning effect), both truncating the very first samples.
My complete impressions on the dongle here
https://audiosciencereview.com/foru...ed-to-cirrus-logic-cs43131.23138/post-2140861

 

[mc.god]

Sorry for the little OT, but just wanted to say that the Kuang Pai made me check in details the buzzing noise issue I mentioned earlier about the Sonata BHD, since with the KP it was much much worse and didn't go away when start playing, making it almost unusable (probably due to the extra amplification stage, who knows...)
Well, i finally found out that it has nothing to do with the dongles themselves, the culprit was a crappy after market psu i recently got for my old laptop for reasons Disconnecting the psu leaving the laptop in battery, all the electrical buzzing went away, reconnecting it buzzing returned, connecting the original laptop psu again no hint of noises. It is a cheap aliexpress 90W psu for HP laptops, clearly sh***y made, probably unsafe in the long run for the laptop itself, so maybe these dongle saved me from future worse problems.

 

[MC_RME]

It might just miss the ground.

 

[Bongó]

Hola a todos,

Estaba muy feliz con el fosi audio ds2. Y ahora mismo empezó a fallar. Lo conectado al celular y no lo reconoce, parpadea una luz azul. Tal vez el cable tipo c no funciona. ¿Alguno de ustedes tiene información?

Muchos saludo!

 

[mc.god]

It might just miss the ground.

It's what I think, even though it has standard 3 pins AC cable and is plugged into a properly grounded AC socket. I went back to the original psu so I'm ok now, I'm seeing to return this crap, if I could'n then maybe i'll dismantle it and try to see what mess they did inside with my limited electronic knowledge.

 

[MC_RME]

You should first measure the connection from DC barrel ground to AC PE. It might be that PE is only used on the primary side inside the PS, or not at all connected.

 

[mc.god]

You should first measure the connection from DC barrel ground to AC PE. It might be that PE is only used on the primary side inside the PS, or not at all connected.

I tested continuity between the barrel negative and the AC ground pin and there is none. By the way the original psu is the same, no continuity, but it produces no noise.

 

[mc.god]

I tested continuity between the barrel negative and the AC ground pin and there is none. By the way the original psu is the same, no continuity, but it produces no noise.

Just to close the OT, and maybe it could still be relevant to the DS2 since its similar configuration, i have to say that the buzzing noise, although only present with the crappy psu, is ggenerated only with the Sonata BHD (faint) and the KP Player 3 (unbearable), while nothing strange is audible with internal laptop phones jack and with all my other dongles (Ugreen HiFi Pro and Jcally JM20 both with CS43131, Grave audio A06 with CX31993), so i really don't understand, the only things that come to mind are that both Sonata and KP are balanced dual dac and with detachable cable, while the others are all single ended single chip with integrated cable.

 

[nscrivener]

Putting aside the issues people are reporting with use on Windows 11, how would the sound quality of the DS2 compare to the RME ADI-2 DAC Fs when driving Sennheiser HD600? Is it likely that I will notice any significant difference?

 

[tccalvin]

Mettendo da parte i problemi segnalati dalle persone con l'utilizzo su Windows 11, come si paragonerebbe la qualità del suono del DS2 a quella dei DAC F RME ADI-2 durante la guida del Sennheiser HD600? È probabile che noterò qualche differenza significativa?

With all likelihood, no.

 

[Mort]

how would the sound quality of the DS2 compare to the RME ADI-2 DAC Fs when driving Sennheiser HD600?

Indistinguishable as long as it was loud enough for you.

 

[Sangram]

I purchased mine a month ago and it seemed all right to me, but I hadn't run any tests on Windows. I did notice that my OnePlus 12R did not recognise the DAC, and needed a different cable, which I borrowed from my USB SSD. Since then it seems many users have been having issues with Windows. After reading some of the feedback here, gave it a whirl again, this time on Windows.

I tried three different systems, two running Windows 11 and one running Windows 10. Foobar 2000 was used for the tests. Control software was SM Player and control device was AK4377-based SevenHertz 71. Using HD600 for the tests, though audio quality is not the subject of the investigation.

1. Asus Zephyrus G15 running in low power mode

2. Ryzen 3600XT desktop.

3. HP laptop running in full power mode, Windows 10

The stock cable was tried, as well as the alternate cable from my SSD. Ports were 10gbps, so no issues on the power front. All tracks were 44.1KHz, as I didn't want resampling to affect the results. I will say that I got poor results on all systems, with both sets of cables.

WASAPI/exclusive: unusable. Stalls for three seconds while seeking, unpredictable pauses in the middle of songs, balance shifts between ears as the fade-in happens, and takes about 4 seconds to actually start the song. I tried multiple tweaks, setting the buffers to 5ms in the advanced menu, setting the process to realtime priority, and enabling and disabling Event mode. I think what this needs is the old KS driver, but I'm not sure if it works on W10 and 11. As it is WASAPI is unusable.

ASIO: Also unusable. First phrase of all songs are lopped off completely. Regardless of sample rate and buffer settings - you can actually see the visualisation start a second before hearing any sound, sort of like lightning. Seeking behaves the same way as exclusive, silence for two seconds and then starts with a fade in and the two channels come in at different times.

Shared mode: Very interesting in that Foobar seems to work properly with none of the above issues. SM Player, on the other hand, has a noticeable delay in starting at all. No pauses noted in either software in shared mode.

Control device did not exhibit any anomalous behaviour is shared mode or exclusive mode. Worked perfectly during seeking and played the first note of each song perfectly..

On Android One Plus tablet, using stock cable (also using Foobar 2000)

First note of songs are occasionally lopped off, just like Windows. Seeking is fine even while micro-seeking to provoke stupid behaviour regardless of cable used and resampling. No skipping or pausing observed at any samplerate. Sanity check was carried out on alternate device at 176k samplerate, again it was OK with no issues.

I have not run any tests with the iPad yet, but maybe I'll do that later.

I'm happy to blame Microsoft for the WASAPI issues because Windows is a messy codebase - but ASIO at least should work properly. Interestingly ASIO driver does not seem to recognise the device fully. The ASIO control panel shows this when device is connected:

And this when unconnected:

Fosi dropped the ball and should simply state the device is not compatible with Windows, because it isn't. It does not exhibit any traits of Windows compatibility in its current form. A friend of mine bought one on my recommendation and his is unusable because he only uses Windows. I've offered to buy it from him to offset his loss - we live in a country where customer support is poor to non-existent, and the manufacturer has stopped replying here, one of their prime target markets. It is much worse for us but it's a lesson learnt, never trust these small manufacturers. I made the same mistake with the DX3 Pro, which was also a lemon and a half. Had much better luck with Gustard, and I'm beginning to think they're the exception.

 

[melowman]

The ASIO driver needs to be loaded (e.g. by playing audio (in ASIO mode)) in order for the Control Panel to update accordingly.

 

[Sangram]

The first picture is with audio playing.

 

[Labjr]

I use Windows 10 almost exclusively and it's working pretty well after installing a Savitech driver. I only hear the uneven fade-in for a split second if I pause and restart the media player. When I first started using it, the audio would briefly dropout about once per mInute while playing DSD files. That seems to be resolved. I also have a Windows 11 machine which I have not tried.

 

[melowman]

The first picture is with audio playing.

Play in ASIO mode and you’ll get what you expect.
The control panel is ‘working’ fine here.

 

[Labjr]

I just tried the DS2 with my Windows 11 machine with pretty much the same results as Windows 10. Everything PCM worked well without any changes. Dropouts about once per minute with DSD files. Again, I installed the Savitech driver and now it seems fine.

I do sometimes get that brief fade-in problem when I change tracks. I'm not positive but it seems to happen with the same track if I switch back and forth a few times. For me it's minor. Considering the performance of this device, I still find it to be a great value.

 

[mslim]

How important is knowing output impedance? (This was tested for the Meizu Pro but subsequently I don't usually see this testes anymore).

 

[Chr1]

Jeez. Hope I don't see testes either...
Put you right off your music for sure.
Scary thought.

 

[Chr1]

Sorry