strange SMSL M500 high 3rd harmonic

[jruser]

Can anybody give me instructions on how to test for this issue on a PC? It was mentioned earlier than it can be tested with onboard audio. I'd like to check the Sabaj D5

 

[PeterHans]

Maybe a stupid question, but why didn't amirm find this kind of problem in his testings (e.g. SU-9)?
I think there are also a lot of newbies who just read the first page and will buy the product because of its great results.
On the other hand, I read this device performance in the range of a 150$ device, how is that possible, regarding to the results in the actual review?

 

[3125b]

@jruser
Install Room EQ Wizard 5.20 (I think it's stimm in beta so only available on AVNirvana), select the correct input and output device and run a test tone on only the right channel, the results are shown in the RTA window. The issue might or might not be visible, if you have the DAC connected via an electrically conductive cable you will have a ground loop, but that shouldn't hide a distortion spike that high.

Maybe a stupid question, but why didn't amirm find this kind of problem in his testings (e.g. SU-9)?

Because he didn't use a mono or two different tones in his testing.
There is only so much you can test for and that kind of issue seems uncommon among DACs. And if he tested for it, since this is software related, it might be a bug with only some of the firmware versions and just happen to not affect the reviewed sample.

 

[MusicNBeer]

Maybe a stupid question, but why didn't amirm find this kind of problem in his testings (e.g. SU-9)?
I think there are also a lot of newbies who just read the first page and will buy the product because of its great results.
On the other hand, I read this device performance in the range of a 150$ device, how is that possible, regarding to the results in the actual review?

Because it wasn't tested for. The problem I'm seeing is that it's still not being tested for nor has it been acknowledged in any way by the host. 2nd to the SMSL response, that's the most disappointing thing. Oh well, so much for science.

 

[Massimo]

Because it wasn't tested for. The problem I'm seeing is that it's still not being tested for nor has it been acknowledged in any way by the host. 2nd to the SMSL response, that's the most disappointing thing. Oh well, so much for science.

The problem ASR has in acknowledging that their SINAD testing methodology is flawed (only testing stereo devices with a mono signal) is that it potentially invalidates all the DAC tests in that the SINAD measurements are only certain when DACs are played in mono. In the best interests of audio science the SINAD chart should be marked accordingly.

 

[MusicNBeer]

The problem ASR has in acknowledging that their SINAD testing methodology is flawed (only testing stereo devices with a mono signal) is that it potentially invalidates all the DAC tests in that the SINAD measurements are only certain when DACs are played in mono. In the best interests of audio science the SINAD chart should be marked accordingly.

But that's fine! I've been an engineer for 25 years snd I still miss stuff. Much less than I used to, but still I miss. I've always fully accepted and changed course on a miss. Seems that lesson is missing among our host.

 

[JSmith]

The problem ASR has in acknowledging that their SINAD testing methodology is flawed (only testing stereo devices with a mono signal) is that it potentially invalidates all the DAC tests in that the SINAD measurements are only certain when DACs are played in mono.

Not really to be completely fair, as they're pretty standard tests for audio products. If it was a hardware issue you'd only have one channel working at all and would see/hear this straight away. This is not a hardware issue at all, but a firmware error in the channel mapping (likely) so it should only affect units that use this same topology and firmware for the es9038pro only. I can't say for sure, but I'd imagine Amir would be checking this now quietly, but won't post the results unless a similar issue ever shows up again. We and Amir also don't know if the original unit tested was a golden sample... this is why I posted at some point in this large thread that if SMSL are going to use ASR testing to promote their products, then the door should swing both ways, yet we have no contact here from them at all.



JSmith

 

[MusicNBeer]

@JSmith, but why not accept the miss in measurement coverage, learn from it, and improve. Measurements are great, but ultimately useless if they don't fully cover all use cases. It's a major miss that should be learned from.

 

[JohnYang1997]

@JSmith, but why not accept the miss in measurement coverage, learn from it, and improve. Measurements are great, but ultimately useless if they don't fully cover all use cases. It's a major miss that should be learned from.

You can test all the tests for all sample rates and for left, right, both channel.
IMD tests with different frequencies, SMPTE, and CCIF. With different levels of multitones, different number of tones, adding pink sepctrum slope. THD with N, THD+N. Versus level at different frequencies, versus frequency at different levels.
DSD, direct mode, attenuated.... THD vs level at very small increment, you can see jagged lines with many ess dacs. Also power consumption while doing any of the tests. Also polarity at any of the test conditions. With different inputs.
There are just too many tests. Any one of them can go wrong.

 

[restorer-john]

There are just too many tests. Any one of them can go wrong.

Absolutely, but you do have to ask the question on whether a number of important tests that are currently not being undertaken, should be undertaken, to more correctly categorize and rank the products being reviewed.

Abolute polarity- tick (single 0dBFS sample impulse is fine)
Channel separation vs Frequency- big tick.

Feeding the same signal to both channels and ranking by that, is a best case number. The influence of one channel on the other either by different signals, or deliberately phase offset signals in each channel would be useful for interchannel related issues to manifest.

 

[Shoaibexpert]

I have some thing to ask and something to share on what's going on in my setup. I bought two units of the SU-9 from HiFiGo in the last months. The return period has expired and I honestly thought at the time that I was upgrading from M500 to SU-9 and that it was worth it. Now I want to understand what the issue being discussed means for me as well as other plan users of the SU-9? Is my DAC garbage? Were the measurements reported in Amir's review not repeatable in actual home setups? Basically should I get rid of the SU-9 and buy some other DAC and would that be audibly superior due to the faults in SU-9, if yes...in what way? I feel I am defrauded yet again by SMSL and all I want now is to assess the extent of damage. Finally what's the solution available to me or other users to correct the situation?

Now that sharing part. I have hooked up my SU-9 to a NuPrime A300 D Class amp (temporarily) untill my NAD C298 arrives and am running KEF R7s with it. Below is the REW measurement of pink noise from my system. I notice a very very noticeable clockwise shift in frequency response. Any link to the issue being discussed here? Any possible reasons why this is the case?

 

[mycose]

Is my DAC garbage? Were the measurements reported in Amir's review not repeatable in actual home setups? Basically should I get rid of the SU-9 and buy some other DAC and would that be audibly superior due to the faults in SU-9, if yes...in what way? I feel I am defrauded yet again by SMSL and all I want now is to assess the extent of damage. Finally what's the solution available to me or other users to correct the situation?

1. As far as measurements go, yes, it is absolute garbage, especially if you consider that even something like this ($75) outperforms this DAC, measurement-wise.
2. Yes, his measurements are repeatable, the thing here is that Amir didn't do a quite rare and particular test.
3. I don't think any other DAC would be audibly superior, the human ear can only hear so much and I'm fairly sure that in a blind, controlled test nobody would ever be able to tell the difference between an unit with this bug and another without it. But yeah, we've been effectively defrauded (or should I say scammed?) by SMSL, because most of us bought this DAC because oh mah measurements oh mah SINAD.
4. As of now, there is NO solution. I myself contacted both SHENZHENAUDIO and SMSL only to my disappointment: 1) they don't seem to give a single ___ about this, 2) they haven't even released an official statement about this issue so it literally doesn't exist for anyone outside ASR and 3) I even told them that I was willing to pay both the update tool and other shipping fees only to find that they just ignored my message. Don't even waste your time contacting them: they just don't care about this.

TL;DR: You're screwed up. Just avoid SMSL at all costs. It's infuriating, but we really can't do anything other than avoiding them.

 

[Shoaibexpert]

As far as measurements go, yes, it is absolute garbage

Just trying to understand this measurement part of your comment. I run my DAC fed from Roon via USB balanced to the Amplifier. Amir did the SINAD and other measurement tests... assuming his unit wasn't a golden sample, those results were jaw dropping. Why then now the measurements for SU 9 are garbage and lower even than the Topping D10? Is there a specific setting/use case in which the measurements suffer...coz I'd imagine Amir's tests would otherwise reveal all relevant issues as he does tests dealing with factors affecting real listening scenarios/metrics.

Also, what SINAD value are we looking at as the revised number (estimated)?

And audibly, what subjective ques should be affected, is the darker sound of this DAC due to this issue as well? If you see there is a clear tilt in what seems to be diminishing higher frequencies linearly...again a linearity test passed with full marks when Amir did it.

Something doesn't quite fit well in the story. Why were audible issues not detected in Amir's tests, I wonder!

Thanks for the clarification.

 

[Shoaibexpert]

I think there is also the argument that when both channels are playing and when volume is reduced the 3rd HD reduces significantly. It would be great if for us defrauded and helpless folks, some expert here would suggest the best volume (1-99) to run this thing on to avoid audible distortion and a clarity that people who are using this to play stereo music via USB are affected or not...

 

[aandres_gm]

@Shoaibexpert these issues didn't show in amir's testing because he tested with a mono signal. It's only when the DAC plays stereo signals that you see this issue. Good thing is, no matter how bad the measurements look now, unless you're a golden eared audiophile, chances are you won't be able to hear the difference.

You could actually test this yourself by playing a song normally, and then playing that same song downmixed to mono. Hear any difference? Think you won't.

The issue here is not the audibility of the issue, but the fact the great measurements were part of the marketing for this device, but that performance can't be reached in normal operation. It's almost disingenuous publicity at this point. To make things worse, SMSL hasn't said anything about the issue.

 

[grzlybear]

Hi guys, I've received the package from SMSL today. In it was a USB cable and a Jlink debugger:

View attachment 136009

It didn't come with any instructions or firmware, it might be flashed on the micro controller? I have no way to tell.
View attachment 136010

I opened up my unit and it was quite easy to find the socket for the flat cable.

I connected the flat cable and the micro usb cable to the jlink debugger. I turned on my unit.
Everything seems to be working but there was no UI indication about any process (no progress bars, no update notification, nothing...).
I've reached out to SMSL to see if they can provide me with instructions.
I'll update once I have some more details.

Wanted to update that I heard back from SMSL.
I initially asked for instructions for using the update tool.
They replied that they were extremely sorry but I'll need to return the unit to China.
In the mean time, they'll send me a new unit.

A bit confusing but overall I'm happy with customer service.

 

[mdsimon2]

Just trying to understand this measurement part of your comment. I run my DAC fed from Roon via USB balanced to the Amplifier. Amir did the SINAD and other measurement tests... assuming his unit wasn't a golden sample, those results were jaw dropping. Why then now the measurements for SU 9 are garbage and lower even than the Topping D10? Is there a specific setting/use case in which the measurements suffer...coz I'd imagine Amir's tests would otherwise reveal all relevant issues as he does tests dealing with factors affecting real listening scenarios/metrics.

Also, what SINAD value are we looking at as the revised number (estimated)?

And audibly, what subjective ques should be affected, is the darker sound of this DAC due to this issue as well? If you see there is a clear tilt in what seems to be diminishing higher frequencies linearly...again a linearity test passed with full marks when Amir did it.

Something doesn't quite fit well in the story. Why were audible issues not detected in Amir's tests, I wonder!

Thanks for the clarification.

You would expect a downward sloping response with pink noise. I do not see anything out of the ordinary here.

Michael

 

[aandres_gm]

Seems typical asian experience, you are promised something and given something far inferior...

Ah right. Extrapolating anything a continent of 4,5 billion people can do from...the performance of a DAC made by a small company in Shenzhen.

 

[Laserjock]

The coolest thing about this movie was it was still pandemic atmosphere but the movies here opened up with limited reservation seating.
My wife and I reserved a spot in prime seating area and we were the only ones in the theater for that showing.

 

[JSmith]

this movie

So you watched a movie with a DAC... please, tell me more.



JSmith