• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Measurements Of SMSL M8 DAC

March Audio

Major Contributor
Manufacturer
Patreon Donor
Joined
Mar 1, 2016
Messages
3,965
Likes
3,864
Location
Perth Western Australia
#1
I currently have a SMSL M8 DAC sat in front of me. This unit is based on the ESS ES9018K2M chip, but I gather it has now been updated to the M8A with an ES9028K2M chip. It has USB, electrical and optical SPDIF inputs and single ended RCA outputs. The unit is powered here by a 9 volt linear power supply, Im guessing thats an optional upgrade and it would be supplied with a SMPS wall wart. The case is a nice compact aluminium affair with a teeny tiny, oh yes tiny display telling you input, sample rate and image filter type (assuming you have your glasses on to be able to read it :) ). Options are fast, slow and Mini (presumably minimum phase).

Available at Amazon for around $250
https://www.amazon.com/ES9018-Optical-Coaxial-Asynchronous-decoder/dp/B00PS0WLX6


IMG_20180312_211930.jpg

IMG_20180312_212440.jpg


Measurements

0dBFS - Output is 1.95V RMS. Pleasantly surprised with the low harmonic distortion! Note noise floor is higher due to use of -20dB attenuator on the QA401 input to minimise its harmonic distortion contribution. Yes I need a notch filter :) !
upload_2018-3-12_21-31-52.png


-6dBFS
upload_2018-3-12_21-34-58.png


-60dBFS - Noise floor drops to very low levels but as a consequence reveals a whole bunch of spuria. Has to be said it is difficult to see this as of any audible consequence at -140dBFS, but still I dont like to see it..
upload_2018-3-12_21-37-43.png


-90.31dB 24 bit Waveform - Very clean due to the low noise floor
upload_2018-3-12_21-41-43.png


Jitter - None of significance
upload_2018-3-12_21-45-0.png


IMD SMPTE - Extremely clean along with the correspondingly low HD
upload_2018-3-12_21-46-58.png

upload_2018-3-12_21-47-41.png


FR and Image filtering - FAST
upload_2018-3-12_21-49-53.png


SLOW - Early rolloff in FR
upload_2018-3-12_21-50-48.png


MINI
upload_2018-3-12_21-52-34.png


Linearity - Pretty good!
upload_2018-3-12_22-30-0.png


I havent listened to this unit yet, but technically its very good. Remind me, why do people buy expensive DACs????
 
Last edited:

stunta

Addicted to Fun and Learning
Joined
Jan 1, 2018
Messages
618
Likes
395
Location
Boston, MA
#3
Remind me, why do people buy expensive DACs????
Too much disposable income? :)

Thank you for posting the measurements. The PSU is around $140 so it takes the cost to around $400 USD. Good to see it measure so well.

I "splurged" on the Topping DX7s and I think its a better buy than this SMSL. The Topping has built-in power supply, xlr outputs and supports a remote control.
 
Joined
Nov 28, 2017
Messages
56
Likes
23
#4
I did the Massdrop thing with this DAC to replace a noisy Behringer 202. Very happy with it but I have no clue what the three filter settings are for and when should they be used.
 

sonci

Active Member
Joined
Feb 18, 2018
Messages
199
Likes
76
#5
Nice!
Is it much complicated to put other dacs measurements in the same graphic?
 

Jimster480

Major Contributor
Patreon Donor
Joined
Jan 26, 2018
Messages
1,143
Likes
623
Location
Miami
#6
Too much disposable income? :)

Thank you for posting the measurements. The PSU is around $140 so it takes the cost to around $400 USD. Good to see it measure so well.

I "splurged" on the Topping DX7s and I think its a better buy than this SMSL. The Topping has built-in power supply, xlr outputs and supports a remote control.
I upgraded my M8 to a DX7. To me the DX7 does sound slightly better in technical tracks.
Also the M8 had pops from time to time which bothered me and also got extremely hot, my DX7 has none of these things.
Tbh the linear power supply isn't worth it, but it would be nice to see it tested with and without.
I paid $180 for my M8 and $270 for my DX7.
 

March Audio

Major Contributor
Manufacturer
Patreon Donor
Joined
Mar 1, 2016
Messages
3,965
Likes
3,864
Location
Perth Western Australia
#7
I did the Massdrop thing with this DAC to replace a noisy Behringer 202. Very happy with it but I have no clue what the three filter settings are for and when should they be used.
Here is what wiki says about filters

The sampling theorem describes why the input of an ADC requires a low-pass analog electronic filter, called the anti-aliasing filter: the sampled input signal must be bandlimited to prevent aliasing (here meaning waves of higher frequency being recorded as a lower frequency).

For the same reason, the output of a DAC requires a low-pass analog filter, called a reconstruction filter - because the output signal must be bandlimited, to prevent imaging (meaning Fourier coefficients being reconstructed as spurious high-frequency 'mirrors'). This is an implementation of the Whittaker–Shannon interpolation formula.


Ideally, both filters should be brickwall filters, constant phase delay in the pass-band with constant flat frequency response, and zero response from the Nyquist frequency. This can be achieved by a filter with a 'sinc' impulse response.

In simpke terms you need to filter the dac output to take away any signals above half the sample rate. So if 48kHz anything above 24kHz. Ideally this filter should be infinitely sharp and deep. However the real world is not like that and filters have limitations and tradeoffs. Im sure you have seen much hifi debate about the sound of filters and their impact on impulse response and things like pre ringing of the signal.

You have 3 filters which have different characteristics. The slow performs very obviously different and impacts the frequency response rolling it off as early as 15kHz. It may sound a little less bright. My experience is that although filters do sound very subtly different it isnt worth getting excited over. Often a different recording will change your preference. Personally I would stick with fast setting.
 
Last edited:

March Audio

Major Contributor
Manufacturer
Patreon Donor
Joined
Mar 1, 2016
Messages
3,965
Likes
3,864
Location
Perth Western Australia
#8
Nice!
Is it much complicated to put other dacs measurements in the same graphic?
It is a challenge. Unfortunately the QA401 software doesn't allow you to load up previous measurements, so to do this you would need to export the data to excel and plot. Not insurmountable of course, but a PITA.
 
Joined
Mar 8, 2018
Messages
12
Likes
8
#9
I used to own a unit of this, a pretty clean sounding DAC. Nice to see the measurements agree with my impressions. Its OLED screen does tend to dim rather quickly over time though. It's another of those Chinese dacs with good performance to price ratio.
 
Last edited:
Joined
Jun 20, 2018
Messages
352
Likes
140
#10
Did you by any chance also measure channel imbalance and channel crosstalk of the SMSL M8?

I'm asking because I'm trying to decide whether to get the topping D50 or the SMSL M8A v2 (stated to even have about 3dB less distortion than the M8), and the SMSL M8 measured here seems to measure slightly better overall than the D50 because of the better SMPTE intermodulation distortion. But the D50 design with two ES9038q2m chips may well have much less channel crosstalk (stated as -116dB by Topping).
 
Joined
Jun 20, 2018
Messages
352
Likes
140
#12
Unfortunately I didn't take those measurements. However based on the other measurements the designer knows what they are doing. The M8 measures superbly. I wouldn't get too stressed about choosing between the two.
Let me take this opportunity to thank you for the measurments!
I'm not too familiar with the technical aspects of DAC design. Is there no reason to believe that by using a single ESS chip in the M8 (and M8A v2) the channel crosstalk may be much higher than the -116dB channel crosstalk of the Topping D50 which uses two ESS chips, one per channel in mono configuration?
I saw that there are rather large differences in crosstalk of the SMSL headphone amplifiers (with the SMSL VMV VA1 having by far the lowest, thinking about getting it but if the DAC has much higher crosstalk this may not be worth it), this made me think it could perhaps go either way with the M8 DAC. And I just recently found out how important imaging is for me :)
 

March Audio

Major Contributor
Manufacturer
Patreon Donor
Joined
Mar 1, 2016
Messages
3,965
Likes
3,864
Location
Perth Western Australia
#13
I think one thing this forums measurements have demonstrated is that it is dangerous to generalise. Individual implementation is more important than theoretically better components. You have to measure an individual design to know how it performs. There's no real value in guessing. Who would have known Schiit products were indeed so shit?

Imaging is going to be far, far more affected by your speakers and room acoustics than the dac, so in this respect either dac will be fine.
 
Joined
Jun 20, 2018
Messages
352
Likes
140
#14
I think one thing this forums measurements have demonstrated is that it is dangerous to generalise. Individual implementation is more important than theoretically better components. You have to measure an individual design to know how it performs. There's no real value in guessing.

Imaging is going to be far, far more affected by your speakers and room acoustics than the dac, so in this respect either dac will be fine.
I'm not worried about using the DAC for my room speakers, it will indeed surely be inaudible. But I'll also be using it with my Beyerdynamic DT1990 Pro and these have fantastic precise imaging (best I ever heard). I just sent a DAC back because it gave hazy imaging. Don't know if this was crosstalk related or if it was because of small channel imbalances or because it had different distortion levels per channel (last two were measured by Amirm, see Melokin DA9.1 review). But after careful listening to actual music I was surprised to find that it was indeed audible and that I didn't like the slightly hazy imaging at all. It took away from realism in some of the music I enjoy most. I found that even though the Melokin had a better "tone" / sound overall, in the end I would have picked a much cheaper Sabaj DAC because it had better pinpoint imaging (which in addition also gave an especially realistic sense of space with certain wet recordings). So therefore my newfound love for precise imaging :) I'm sure the M8 doesn't suffer from the kind of channel imblances and differing distortion per channel that the Melokin does, but crosstalk does relate to imaging width it seems to me and I want to get the imaging thing as perfect as I can. Just my last deciding factor between these two DACs.
edit: btw, have sent an email to SMSL asking for the crosstalk spec of the M8A v2, hope they know and reply.
 
Joined
Jun 20, 2018
Messages
352
Likes
140
#15
Just received the SMSL M8A v3 (ES9038Q2M DAC chip).
My subjective experience is that it is fantastic. Just about perfect actually.
Perhaps it is a little bit on the bright / sharp side, but I switched on quality upsampling to 352.8kHz on my computer and this may have made a small improvement here because I think it sounds fully balanced / filled in now to me but haven't tested this well.

Also got the matching SAP-12 headphone amplifier. My subjective experience is that it's not great. It's a bit on the warm side perhaps but the worst thing is that it has a channel imbalance problem. Left channel is a bit louder, imaging isn't great, etc. Bit dissapointed, I will send it back.
Perhaps it's a better design to do digital volume control in the DAC instead of an analogue volume pot on the headphone amp?
I've decided to order a Topping DX7S and compare it. Hope the DAC will sound as good but with the headphone amp able to keep up this time..
 

Ron Texas

Major Contributor
Joined
Jun 10, 2018
Messages
1,815
Likes
1,275
Location
Equidistant From Everywhere
#16
Just received the SMSL M8A v3 (ES9038Q2M DAC chip).
My subjective experience is that it is fantastic. Just about perfect actually.
Perhaps it is a little bit on the bright / sharp side, but I switched on quality upsampling to 352.8kHz on my computer and this may have made a small improvement here because I think it sounds fully balanced / filled in now to me but haven't tested this well.
This subjective finding that DAC's with ESS DA chips keeps popping up. It is described as bright, overly detailed, high frequency emphasis, forward, glare and so on. The only ESS DAC I had in my prior system was a Dragonfly Black 1.2. It sounded so harsh with B&W 685 S2 speakers that I couldn't listen to it for 5 minutes. Some have opined it may have been plugging an output meant to feed headphones directly into a power amp. When I used it with HD280 phones, it was fine. However the HD280 is often chided for being midrange only.

According to the developer of HDPlayer the "house" sound of AD chips from the most forward to the most relaxed is as follows:
CS
ESS
AK
WM
TI/BB
He also said the differences narrow considerably when converting PCM to DSD before sending it to the DAC. Bear in mind the software he sells is mostly used for that purpose.

At least one member of this forum thinks the ESS forwardness is BS to the point that it is a meme, an opinion which spreads by copying. Arnim and others say DAC's sound mostly alike when volume matched. Designers of DAC's are increasingly picking ESS for the Audiophile market. CS and AK are more popular in pro audio gear. I have become apprehensive about purchasing an ESS chipped DAC, but might try it with an inexpensive unit like the D10.
 

dc655321

Senior Member
Joined
Mar 4, 2018
Messages
495
Likes
446
#17
At least one member of this forum thinks the ESS forwardness is BS to the point that it is a meme, an opinion which spreads by copying.
I guess you're referring to me? Guilty...
Until I see evidence of some physical effect(s) for the subjective claim of Sabre DAC "glare", I will continue hold this theory.
However, if evidence can be produced that shows a cause for this perceived effect, I would be happy to revise my thinking.

Arnim and others say DAC's sound mostly alike when volume matched.
Unless there is some aspect of performance that causes deterioration of audio quality (lots of distortion, substantial roll-off in frequency, etc), yes, there is not much sonic difference between the DACs I've experienced. There are far greater improvements/differences to be had by changing transducers.

I have become apprehensive about purchasing an ESS chipped DAC, but might try it with an inexpensive unit like the D10.
An interesting experiment may be to assemble a collection of DACs, say 4, with 2 of them having ESS chips in them. If all have very low output impedance and are level-matched, have someone swap them in/out while you listen to music of your choice. Cycle through all the DACs multiple times and try to rank them by brightness, bass response, whatever metric you choose.
 
Joined
Jun 20, 2018
Messages
352
Likes
140
#18
Well actually I did a bit more listening and there is indeed a subtle effect when upsampling from 44.1 to to 352.8kHz. It is less subtle than switching to different filters where I actually don't hear any difference. Sometimes I think I might but pretty sure I couldn't pick it out in a blind test, whereas with the upsampling I'm pretty sure I could pick it out with a blind test. (may do one later, not that hard to set up I guess)
I'm upsampling with SRC under foobar2000, band limited sinc interpolation best quality 144dB SNR 96%BW. (edit: this one https://www.foobar2000.org/components/view/foo_dsp_src_resampler )
To me the subjective effect of upsampling this way is that the sound becomes more "filled in" for lack of better words.

Edit: with older DACs the effect of external upsampling was surely a lot more pronounced. Was just switching between upsampling and no upsampling with the M8A with different music and couldn't tell any difference. Very dependent on the music if the effect is audible to me.
 
Last edited:
Joined
Jun 20, 2018
Messages
352
Likes
140
#19
This subjective finding that DAC's with ESS DA chips keeps popping up. It is described as bright, overly detailed, high frequency emphasis, forward, glare and so on..
I've also owned the hifime 9018D which has the ES9018K2M chip and it was also "bright". Sounded similar to the M8A in my memory though less "perfect" perhaps. These are both ESS mobile DAC chips.
I've also owned the Yulong D200 (I think, keep forgetting the model number) which has an ES9016 chip (full version not mobile) and the Melokin DA9.1 which has the ES9038PRO (full version not mobile) and both these DACs did not sound bright to me at all. Though implementation matters a lot of course.
Perhaps part of the bright sound has to do with the upsampling/resampling taking place in the DACs? My upsampling listening experience seems to suggest this. And perhaps the upsampling / resampling in the DACs is different for the mobile versions than the full versions? Perhaps someone here knows?
 

Ron Texas

Major Contributor
Joined
Jun 10, 2018
Messages
1,815
Likes
1,275
Location
Equidistant From Everywhere
#20
I guess you're referring to me? Guilty...
An interesting experiment may be to assemble a collection of DACs, say 4, with 2 of them having ESS chips in them. If all have very low output impedance and are level-matched, have someone swap them in/out while you listen to music of your choice. Cycle through all the DACs multiple times and try to rank them by brightness, bass response, whatever metric you choose.
Well, I am considering getting a D10 because $90 is a bar tab in my world. If it flops, so what? My testing might not be totally scientific. Your suggestion that clipping was causing my Grace M9xx to sound bad was probably correct. I lowered the volume from 90 which is pass through by 6 db and noticed an improvement. Since then I have moved the Grace to my bedside for headphone use. It has a kick ass headphone amp which drives HD650's with insane headroom.
 
Top Bottom