SMSL SU-2: a dumb, wasted opportunity

[rcstevensonaz]

But in the end, I just feel more comfortable with the quality of the US made Schiit Mimir.

The quality of the manufacture (regardless of country) is always an important factor, along with functionality and price.

The other consideration — where country does matter — is whether you can obtain service if there are any issues with the product; without needing to incur high cost of shipping it to a foreign country for service. However, that is usually just a matter of buying from an authorized in-country retailer.

 

[Roland68]

Topping D70 Pro is next cheapest.

Obv with S/PDIF is dime a dozen

Schiit Mimir is $299. AI says it probably does not have auto mute. Schiit hasn’t confirmed. Otherwise the tests I’ve seen on Mimir all look good.

There are several other—and also cheaper—alternatives with an AES input. For example, the Loxjie D40 for under €170, the D40 Pro on sale for €225, the SMSL DO400 and DL400, the D400 Pro, etc.

 

[Red_Red]

Finally, the review by Flo, NTTY: https://www.audiosciencereview.com/forum/index.php?threads/smsl-su-2-dac-review-ntty.72921/

 

[NTTY]

Sorry I completely forgot to tell you I published it!

 

[stuartkohl]

And maybe the silence just prior and during the Trinnov calibration measurements bursts also have dither, which would keep the SU2 from auto mute. But in the end, I just feel more comfortable with the quality of the US made Schiit Mimir. There are reports on the Trinnov threads on AVSForum stating that the Trinnov works well with the predecessor to the Mimir. Fingers crosses the Mimir works well also.

The Mimir is a nicely made DAC that's assembled in the US. It also has a nice feature set. But don't buy in to the marketing hogwash about it "sounding better" because of some algorithm that Mike Moffat has been hand-rolling on the thighs of virgins for decades.

 

[PGAMiami]

Putting the virgins aside, I had them both side by side. The Mimir is more refined physically and operationally. Never pops. Doesn’t have the goofy mislabeled LEDs. Has more input choices (I just need AES and XLR). Can easily be upgraded to a full featured digital preamp with DSP. It is nicer than the $99 SMSL. But it’s 3x the price. Overall though, both are amazing values that should perform as well or better than products where the sales tax is multiples of the total cost of an SU2 or Mimir.

 

[Ruediger]

Putting the virgins aside, I had them both side by side. The Mimir is more refined physically and operationally. Never pops. Doesn’t have the goofy mislabeled LEDs. Has more input choices (I just need AES and XLR). Can easily be upgraded to a full featured digital preamp with DSP. It is nicer than the $99 SMSL. But it’s 3x the price. Overall though, both are amazing values that should perform as well or better than products where the sales tax is multiples of the total cost of an SU2 or Mimir.

Maybe a bit apples and oranges.

Expensive full featured DAC with chip from 10 years ago vs. competetively priced DAC with reduced features and today's chip.

This comparison makes no sense to me.

 

[rcstevensonaz]

Putting the virgins aside, I had them both side by side. The <my preferred DAC> is more refined physically and operationally. Never pops. Doesn’t have the goofy mislabeled LEDs. Has more input choices (I just need AES and XLR). Can easily be upgraded to a full featured digital preamp with DSP. It is nicer than the $99 SMSL. But it’s 3x the price.

The virgins say "thank you".

No doubt there are a lot of DACs that very favorably compare against the SMSL SU-2... many that are "better in so many ways" at only 2x the cost, let alone 3x the cost (or even those at 5x the cost).

May I also suggest this thread isn't the best place to compare those hundreds (thousands? tens of thousands?) of more expensive DACs against the SMSL SU-2? Instead, can we get back to real issues with the SMLS SU-2 — such as whether the fade-in is actually an issue for calibration systems and the strangely implemented blinking LEDs?

 

[PGAMiami]

I haven’t heard back from Trinnov on whether the fade in matters for their calibration. But even if it doesn’t today, it may in the future.

I really wanted to like the SU2 because I wanted something really simple, with excellent performance and digital AES in and analogue XLR out. Maybe the fade in does not matter but there are other DACs that don’t click or pop without having a fade in from a locked signal and digital silence. I didn’t get really loud pops on turn on, but I did get pops that were loud enough to be bothersome. And the LEDs, although more cosmetic, also bothered me. The blinking was distracting and labels are just incorrect.

Funny thing is I was about to just cover front with black tape, but then read about the fade in. That’s when I packed them up and returned them to Amazon. Also the subjective impression the SU2 left me with was the SMSL didn’t give too much thought to this product and it worried me because I have them connected straight into power amps and speakers that are much more expensive than the SU2.

That said, if you’re on a budget and want a very high performance DAC with AES, there aren’t many options at this price.

 

[rcstevensonaz]

I haven’t heard back from Trinnov on whether the fade in matters for their calibration. But even if it doesn’t today, it may in the future.

I really wanted to like the SU2 because I wanted something really simple, with excellent performance and digital AES in and analogue XLR out. Maybe the fade in does not matter but there are other DACs that don’t click or pop without having a fade in from a locked signal and digital silence. I didn’t get really loud pops on turn on, but I did get pops that were loud enough to be bothersome. And the LEDs, although more cosmetic, also bothered me. The blinking was distracting and labels are just incorrect.

Funny thing is I was about to just cover front with black tape, but then read about the fade in. That’s when I packed them up and returned them to Amazon. Also the subjective impression the SU2 left me with was the SMSL didn’t give too much thought to this product and it worried me because I have them connected straight into power amps and speakers that are much more expensive than the SU2.

From my perspective, those seem like very good reasons to look for something other than the SMSL SU-2 DAC.

 

[Frunse]

The mimir was also the next DAC in my List i was looking for, but after counting it by 8x it became to pricy for my first Setup

Also than my top tier wish to have Product would be an HAPI MK3 16 Channel DAC / Interface but thats around 6000€

 

[john61ct]

There are LOTS of very clean pro audio D/A format converters, 8/16/24/32 analog ports out

that cost a tiny fraction of that.

Using lots of consumer DACs two channels at a time is just bizarre when you need lots of channels.

Even if you "really need" AES input that only doubles or triples price per port.

But, do you **really** need AES?

The other penalty-inducing (not actually helpful) spec you are imposing is transfer at high resolutions.

So that increases pricing too, needing to SMUX your ADAT connections between primary interface and format converters.

Solved by (reducing transport resolution or) looking at primary interfaces with the D/A port count you need.

Or look at legacy pro gear that has pristine SQ, used to be crazy pricey but now goes for pennies on the dollar.

My main point is, don't lock yourself into spending thousands unnecessarily, design your overall system so it can be implemented at reasonable costs.

...

However, wrt this SU-2 device, have you yourself experienced any actual problems wrt your use case? Or are you buying into the idea that this "fade in problem" is going to impact anything besides specific / obscure measurement issues ?

It seems likely that the popping issue can easily be worked around, or get fixed with firmware updates

 

[Frunse]

I have no problem'S with SU-2 at all only had poping noise at first test switching it on or off while the Amp was on!

I have not even seen any Problems with that Feed in and out while i setuped my first aktive Speaker with REW all looks wery well and worked like it should

The Poping stuf will went away if i build up my Relaise Volumecontrol after the DAC, because it would be aktivated with the DAC's and the Output Relais would need longer to switch on than the DAC so anyway.

The POPING was there because i removed the Powercord from DAC while the AMP was on, i don't used the Power off butten before, i guess, not sure anymore, need to test it again soon.

And please there Peopl out that paying much more to get an Trinov or Hypeion Device or even that Mini DSP Tide 16 for 4600€ LOOL

I have spend 1800 € for one RME Card and 8 SU-2 DACS so what?

2x MOTU ul mk5 cost 1500€ for 16 Channels before, now i keep one and sell one soon

An MOTU 16A cost also 1800€ and has 16 Channels but will it run on Linux already?

I waited for an MOTU 24AO with this new 16A Design and newer ESS DAC Chip's but after more than a Year nothing happen in this Direction.

Also MOTU Devices need to be configured with Windows before you could use it on Linux thats also sux.

I am absolutly fine with my decision, it is open for all i want without any Restriction

I need XLR Outs after the DAC'S because i use Hypex Amps!

AES was my decision over DANTE or Ethernet AES67.

ADAT or Toslink was never on my wish List and never will be.......

Cheepest option would be Ferrofish AE16 990€ but only for ADAT with an DIGIFACE USB 415€ .....but not on my wishlist and never be!

thats around 1400€ so what i loose 400€???? because i want AES .....LOOOOOL


Don't forget that all New Surroundprocessors have AES Out or DANTE Out and not ADAT or Toslink 16 Channel for Atmos, so i am absolutly fine with my Decision

I can switch it directly to my RME Card an be fine !!!

With an USB Device not that easey

Nuprime HX16 will have USB but thats not soon geting released and i am not sure will that only Input or also Output or Both, the Future will show us what they build up.

But they will have AES by sure so ...already, i am fine

 

[rcstevensonaz]

It seems likely that the popping issue can easily be worked around, or get fixed with firmware updates

Is there a mechanism for the SU-2 to receive firmware updates? This Google Search AI summary says that it is not possible:

It is generally not possible for end users to update the firmware on the S.M.S.L SU-2 DAC at home, as SMSL does not publicly release user-facing firmware update tools or files for this specific model. Any microcode or internal updates usually require specialized hardware headers or returning the unit to the seller or manufacturer.​

 

[john61ct]

I have no problem'S with SU-2 at all

So don't let the naysayers here influence your plans

 

[john61ct]

Is there a mechanism for the SU-2 to receive firmware updates? This Google Search AI summary says that it is not possible

aka Gemini. I would not rely on that, but if true then only those delaying purchases would benefit, unless "local" sellers can provide that service.

One point in favour of USB based devices I guess

 

[Frunse]

So don't let the naysayers here influence your plans

I don't, there was already some people whining, oh man, the DAC'S wil have Phasedrift'S and so on.....Thats will not work at all...

Itried it and it works very well so far, also i tested this Devices already before anyone as writen something about Fade in and i even didn't noticed it while using REW

No Measuring shows anything bad on my Measurings that could afected by this even on my Stepresponses from Speakers didn't show it.

I listen to music and was not noticing it at all....at 96kHz, 10ms fade in, thats not much thats 10 of 1000 counts of an Second try to count to 10 in a second thats already fast .

 

[john61ct]

Phase getting out of sync with clock drift is a genuine valid concern.

Maybe not for your specific application, AES3 being self-clocking like ADAT and S/PDIF, and all identical DACs, no other DSP processing taking place, short listening sessions...

But the principle holds otherwise always best to BNC WC sync all the A/D & D/A conversion devices to eliminate that concern.

 

[rcstevensonaz]

Phase getting out of sync with clock drift is a genuine valid concern.

Only when you have multiple digital sources producing sound in parallel that must remain precisely synchronized in the time domain. For example, live music recording (drum kits, keyboards, synths) or a home theater with multiple active crossovers and speakers. I believe this is the case for @Frunse situation; and could become an issue if all of the down-stream DACs are not being fed by a common clock source.

However, when the digital domain is a single flow chain (e.g., CD -> DSP processing -> preamp -> speakers), then each digital device in the chain handles jitter and (re)clocking before sending the digital audio to the next step in the chain. In this case, there is no audible benefit to having clocks synchronized across all devices in the chain.

And as a side note, in addition to BNC Word Clock option, the S/PDIF protocol also contains clocking (cf. RME USB Digiface clock sync is entirely optical S/PDIF).

 

[john61ct]

Only when you have multiple digital sources producing sound in parallel that must remain precisely synchronized in the time domain.

I am not doing any recording or mix-mastering. 8ch input, maybe 22+ output all analog, and most transport of the audio signal may stay analog as well.

Devices doing the DSPing in between the ADCs and final DACs not yet settled, the transducer EQing vs crossovers / phase tuning / ToA delays, quite possibly and ideally handled separately from DRC.

So, for most precise time domain control overall, rather than relying on self-clocking interconnects my plan is to simply ensure that ALL the DA/AD converters have BNC WC capabilities both in & out.

If I use distributed RPi5s and/or use AoIP, then PTPv2 disciplining each node's system clock is another wrinkle.

I really want to avoid ASRC complications and am wary of USB as a transport except for the first source selector/preamp upstream of the first ADC conversion.

This approach gives me maximum flexibility for device flexibilty downstream for signal processing - all interconnect formats become fungible since that choice does not impact the time/phase domain.