SMSL DO200 Pro DAC Review

[Tell]

I don't think so because we are interested in the intermodulation distortions in the audible range that are resulted from harmonic distortions of the fundamental frequencies in the ultrasonic range, not the fundamental frequencies themselves in the ultrasonic range, that musical instruments do not produce. So, since "“A multitone excites both harmonic and intermodulation distortion mechanisms in a device.", it seems to me what Amir stated in post#91 makes sense, you are getting/seeming the total distortions.

No, because the ultrasonic THD is distortion products of the played tones. IE the 10kHz tone has harmonics at 20kHz, 30kHz 40kHz etc. The 20kHz tone has harmonic products at 40kHz, 60kHz....


Now think about how severe that multitone test is as a distortion test.

Each of those 32 tones is producing its own harmonic distortion at multiples of itself, right the way through the audible band (for the lower frequency tones) and on way up into the ultrasonic band for all of them.

Then all the 32 tones, plus all the harmonic distortion tones are all inter modulating together to produce IMD tones at the difference frequency between every tone and every other tone. Yet all that HD and IMD is all sitting in the grass of that particular chart below -120dB.

Yeah of course you both are correct and I also realised this a few minutes after I wrote my question. Ah well

 

[Tell]

Great review, essentially SOTA for less than $ 400. The sweet spot to have AES/EBU digital input seems to be around $500. I just wonder how it furthers improve the performance of the DAC when you use AES/EBU as an input compared to optical or SPDIF and how much worth is to spend extra $ to purchase a DAC model that offers this option, like the recently reviewed SMSL RAW.

You can get SOTA for far below 400USD. And how and why would you further improve on something that's already perfect?

 

[antcollinet]

Yeah of course you both are correct and I also realised this a few minutes after I wrote my question. Ah well

We've all been there.

 

[JIW]

As the chart indicates, my standard measurement bandwidth for DACs is 90 kHz. Are you still asking for something else?

BTW, whatever the other reviewer was testing, was including simple noise shaping as is the case with this DAC. At that time, noise shaping was much more severe and hence the reason he was seeing such high numbers. Can't imagine any DAC having that kind of real distortion.

I am not. If anything, I was explaining why there is no need for that specific measurement. If 90 kHz bandwidth THD+N is fine, there is no reason to suspect significant high frequency distortion. However, according to DonH56, THD is conventionally measured using the first 10 harmonics which requires a 200 kHz bandwidth for a 20 kHz fundamental.

Total harmonic distortion, THD, is the ratio of the harmonic distortion to the signal itself and varies with the signal (larger signals are more distorted). By convention the sum of the first ten harmonics are used to calculate THD (so, “total” = 10).

 

[sejarzo]

I wonder if they spray out a series of dacs and then look to see which ones sell most. The current plethora may be a lot of older but still fully worthy models not intended to be re-made in a new batch and quietly discontinued once stocks run out? Schiit have been doing this in a smaller scale admittedly, as when I was looking around at their headphone amps, I discovered I was two generations or more out of date, not that it'd matter especially as none of my headphones/iems are difficult to drive at all.

I'm still delighted with my little SU1, but one day, I'd need balanced outputs as well. A display isn't important to me at all as it's all hidden away currently and eq not important in the source as I use analogue sources too.

I would assume that SMSL is not fully integrated and has parts made in lots by various suppliers, SMSL then assembles the final product, and if a model sells well and quickly, perhaps makes a second or third run.

 

[sejarzo]

Great review, essentially SOTA for less than $ 400. The sweet spot to have AES/EBU digital input seems to be around $500. I just wonder how it furthers improve the performance of the DAC when you use AES/EBU as an input compared to optical or SPDIF and how much worth is to spend extra $ to purchase a DAC model that offers this option, like the recently reviewed SMSL RAW.

I wonder how many hobby audiophiles have source gear that can supply an AES/EBU signal. My guess is that it doesn't present a big opportunity for a firm that focuses on the consumer rather than pro side of the market.

 

[antcollinet]

I wonder how many hobby audiophiles have source gear that can supply an AES/EBU signal. My guess is that it doesn't present a big opportunity for a firm that focuses on the consumer rather than pro side of the market.

And regardless - it offers no benefit in terms of sound quality.

 

[JIW]

Bump.
Can't find anything online...

You can check the manuals for settings pertaining to that for the device you are interested in. For this device, @amirm's photo shows the default screen since there is no setting in the manual to change it.

 

[JIW]

Significant high frequency THD can be a problem for downstream devices such as amplifiers and speakers which may produce intermodulation distortion in the audible band or lead to overloading.

 

[DSJR]

In this place, we're so conditioned now to dacs not needing to be more than a few hundred pounds/dollars to have good performance and facilities, but take a look at a modern take at yesteryear's high end dacs still being sold today. I'm sure this one is competent, but exactly how it compares to the ears alone I've no idea (it looks gorgeous though and priced accordingly)

Accuphase DC-37 DAC - Signals Accuphase DC-37 DAC

Accuphase DC-37 DAC with HS Link for CD & SACD data from a suitably equipped transport. Spookily, we have one listed here too.

signals.uk.com

 

[Bwaslo]

Harmonic distortion is just the intermodulation of one single tone with itself. It is a simple intermodulation test. Back when multiple tones and narrow filteris were hard to come by, the single tone harmonic test was more practical to do than multi tone IM types. But it has limitations, such as that a low pass filter or bandlimit in a device can hide the products of higher order distortions; or can downplay nonlinearities that are proportionally worse with low level signals.

But harmonic and intermodulation tests all reveal the same equipment nonlinearities, though with different effectiveness in different circumstances. Each is a test technique with (mostly) standardized setup and numerically quantifiable results -- they are NOT characteristics of audio equipment! -- just test situations.

That products of a tone or two go out of hearing range is completely irrelevant to actual usage. You never listen to single tones or even just 2 tones. Real audio programs have a lot more complex waveforms and spectra. Which is why tests with many non-harmonically related tones (which are now trivial to arrange) reveal more and have complexity similar to real program material (but actually more taxing). The results tend to be graphic though, and don't easily provide a datapoint number you can point to as being better or worse.

 

[Jimster480]

This is really "another one". The HDMI Arc is nice but the price of $300 is also not cheap. Too many mid and higher priced DAC's from the Chinese brands these days. Honestly I can't imagine that the industry isn't saturated now with a million devices at every price point.

 

[Jimster480]

In this place, we're so conditioned now to dacs not needing to be more than a few hundred pounds/dollars to have good performance and facilities, but take a look at a modern take at yesteryear's high end dacs still being sold today. I'm sure this one is competent, but exactly how it compares to the ears alone I've no idea (it looks gorgeous though and priced accordingly)

Accuphase DC-37 DAC - Signals Accuphase DC-37 DAC

Accuphase DC-37 DAC with HS Link for CD & SACD data from a suitably equipped transport. Spookily, we have one listed here too.

signals.uk.com

To me these devices were always massive jokes that I would never purchase. So having DAC's like this SMSL one for $399 is more on the pricey scale for just a DAC. After all my DX7 in 2017 was $300 and it had a headphone Amp. For a DAC alone to cost this much; it really has to have special features.... HDMI ARC I'm not sure qualifies since it would only be 2 channel anyway....

 

[DonH56]

I am not. If anything, I was explaining why there is no need for that specific measurement. If 90 kHz bandwidth THD+N is fine, there is no reason to suspect significant high frequency distortion. However, according to DonH56, THD is conventionally measured using the first 10 harmonics which requires a 200 kHz bandwidth for a 20 kHz fundamental.

Most DACs have steep filters above 20 kHz so THD at 20 kHz is pretty meaningless. Bandwidth will always reduce THD, of course, which is one reason it was sometimes just a 1 kHz measurement in the past. Bandwidth of the device must be taken into account when looking at THD, but an argument can be made that if the bandwidth limits the number of harmonics that contribute to the THD, then that also affects what we hear and thus is a reasonable measurement. It can mask the measured distortion, however, and is one reason I like to see 19+20 kHz IMD testing, so you can see the 1 kHz product.

Class D amplifiers are a case (to me anyway) for wideband spectral analysis, not necessarily for distortion, but to check for oscillator/switching feedthrough and intermodulation products in the audio band. I recently read about an amp blowing tweeters the instant it was turned on, and the root cause was a DIY modification that shorted the output filter inductor because the owner had read how inductors limit high-frequency response. They do, very intentionally in this case, natch. A little knowledge is a dangerous thing...

BTW, to be clear my day job after graduation was never audio, so there are likely better experts to quote regarding all things audio. The ten harmonics criteria is from my audio days as an IHF consultant (there was a series of volumes to read and tests to pass at the time) and was what manufacturers and reviewers used at the time best I recall. The IHF and some other standards bodies (I am an IEEE member so use them most of the time) specified at least 10 harmonics, but of course if the amplifier does not have the bandwidth for 10, the measurement will reflect that. In practice, second and third usually dominate the distortion, with very little contribution past the 5th or 7th harmonic for real-world designs. Past that they are often buried in the noise floor.

FWIWFM/IME/IMO/etc. - Don

 

[TNT]

Most DACs have steep filters above 20 kHz so THD at 20 kHz is pretty meaningless.

Well, if the distortion is created after that steep filter.... like in the analog output stage?

//

 

[dtaylo1066]

This DAC has a dozen CS43131 chips

View attachment 421750

Is 12 DAC chips in a DAC a new world record? Also has a remote and can function as a pre. What's not to like?

 

[Bleib]

This DAC has a dozen CS43131 chips

View attachment 421750

So, it's actually cheap?

 

[DonH56]

Well, if the distortion is created after that steep filter.... like in the analog output stage?

//

Is this a serious question? Of course it will be included if it is after the filter stages, but typically the output buffer is part of an analog filter network that rolls off as well. Usually not the very high order in a delta-sigma circuit itself, but enough to kill most harmonics of a 20 kHz fundamental. In addition to suppressing images, it limits wideband noise to keep SNR and SINAD high, especially for a delta-sigma DAC that has a rising HF noise spectrum.

Another obvious exception are DACs targeting higher sampling rates for "high resolution" converters with extended bandwidth. A 192 kS/s DAC with 96 kHz Nyquist bandwidth will output 4 or 5 harmonics of a 20 kHz fundamental. Whether you can hear them is up to you and your system.

I am sure you can think of other cases, but by and large even if they are present they would be inaudible and just cause trouble for tweeters (which could potentially modulate the signal back into the audio band). I would hope competent designs would suppress such high frequency content, and trust @amirm to find the ones that don't.

 

[dtaylo1066]

So, it's actually cheap?

Well, if some want to argue that $350 for a quality DAC is expensive, I would submit they are wrong.

 

[Bleib]

Well, if some want to argue that $350 for a quality DAC is expensive, I would submit they are wrong.

It kind of is,.can get the same performance for half the price and headphone amp as well..