SMSL DL200 the best ever value DAC/AMP?

[Veri]

Second, back to the question that started off this blather:
From what I read on that forum, it appears that some guy (IVX, apparently well versed in the topic) tested the chip and found something wrong with the upper frequencies that appears to deliberately artificially increase SINAD and has some negative possibly audible effect. Is this a digital filter? What measured effect did it have and is it even audible? I suppose a better question is: does it mean that the C200 is better?

To be fair, that was the discussion I aimed for in the first place, before it devolved into "9039 sounds best. trust me bro" and the usual dac related blather.
If I reinforced that my bad. But yes let's try to stay to the point of DL200 and 9039-related performance and/or issues. If possible.

 

[nerdemoji]

To be fair, that was the discussion I aimed for in the first place, before it devolved into "9039 sounds best. trust me bro" and the usual dac related blather.
If I reinforced that my bad. But yes let's try to stay to the point of DL200 and 9039-related performance and/or issues. If possible.

do you know what ivx is talking about cuz i dont really understand.

 

[nerdemoji]

ok actually I do understand what this guy is saying... there is better sinad at 1khz but worse sinad at 10-20khz which the guy doesnt like because it is a way of cheating sinad tests at 1k because that is the standard test frequency. He thinks (probably rightly so) that this is a lazy way of making a new chip given that it has worse performance in some aspects. Honestly, this is kind of silly and i feel like -110 db thd+n bla bla is probably inaudible

 

[staticV3]

ok actually I do understand what this guy is saying... there is better sinad at 1khz but worse sinad at 10-20khz which the guy doesnt like because it is a way of cheating sinad tests at 1k because that is the standard test frequency. He thinks (probably rightly so) that this is a lazy way of making a new chip given that it has worse performance in some aspects. Honestly, this is kind of silly and i feel like -110 db thd+n bla bla is probably inaudible

Just for some context:

IVX is currently working on a DAC counterpart to his E1DA Cosmos ADC, and is in the process of evaluating different DAC chips.

The Cosmos line is intended to be precision measurement equipment for all those that either cannot afford an APx555B, or want higher performance than what the APx555B is capable of.

For this use case, the quirk of the ES9039 line that is comparatively terrible THD at 10k and above is obviously not ideal.

I'm not arguing that no one should buy an ES9039 for music consumption because of this, because that would be silly.
I'm just trying to explain IVX's viewpoint from which he writes these statements.

 

[nerdemoji]

Just for some context:

IVX is currently working on a DAC counterpart to his E1DA Cosmos ADC, and is in the process of evaluating different DAC chips.

The Cosmos line is intended to be precision measurement equipment for all those that either cannot afford an APx555B, or want higher performance than what the APx555B is capable of.

For this use case, the quirk of the ES9039 line that is comparatively terrible THD at 10k and above is obviously not ideal.

I'm not arguing that no one should buy an ES9039 for music consumption because of this, because that would be silly.
I'm just trying to explain IVX's viewpoint from which he writes these statements.

Ok thanks for the explanation… makes much more sense now

 

[AncientWisdom]

I guess it's inevitable that if people buy products based on measurements, eventually the products will be developed to beat the tests specifically (see VW Dieselgate as a very bad example of this).

Ideally, the measurement methodologies keep getting improved as well, in this case maybe THD should be measured in 2-3 different places along the spectrum.

 

[Bleib]

I guess it's inevitable that if people buy products based on measurements, eventually the products will be developed to beat the tests specifically (see VW Dieselgate as a very bad example of this).

Ideally, the measurement methodologies keep getting improved as well, in this case maybe THD should be measured in 2-3 different places along the spectrum.

Multitone gives quite a good picture in my eyes

 

[nerdemoji]

I guess it's inevitable that if people buy products based on measurements, eventually the products will be developed to beat the tests specifically (see VW Dieselgate as a very bad example of this).

Ideally, the measurement methodologies keep getting improved as well, in this case maybe THD should be measured in 2-3 different places along the spectrum.

Agree, someone should ask amir if he can calculate sinad for 1k and 10k as an average, to prevent this kind of stuff. I wonder if multitone would be a better measurement (if possible) to generate correct sinad numbers from.

 

[staticV3]

Ideally, the measurement methodologies keep getting improved as well, in this case maybe THD should be measured in 2-3 different places along the spectrum.

As we do:

eventually the products will be developed to beat the tests specifically (see VW Dieselgate as a very bad example of this).

Even with the current battery of tests, there is no way to "cheat the system". A DAC that does good in all tests, will be a great sounding DAC.

Let's assume that similar to Dieselgate, the DAC had some kind of processor inside that detects measurement test tones and only unlocked the full performance then, but went back to default "terrible" mode playing regular music.

For this to work, the manufacturer would have to build the DAC to be capable of transparent audio output anyway, and at that point, he could just sell it as such no trickery involved.

The only scheme of such that I could see as remotely "worthwhile", would be a DAC with subtly bass-boosted or v-shaped frequency response with music, but that reverts to flat response once it detects a synthetic sweep or noise signal.

However, just one simple test with Deltawave/Audacity/Audition using music as test signal would reveal that instantly.

 

[iconoclast]

I guess what I really want to know is if it will be as good as a dx3 pro+. I have that now and it makes my headphones sound better than ever before, but it can’t deal with my tv’s jittery optical out. I need something in the same price range with the same quality that has dpll adjustment for jitter.

I had the same problem with my LG OLED and a Topping DX5. I needed balanced out and dual outputs so I got an SMSL D-6 and that works fine, FWIW.

 

[Bleib]

Well, received mine yesterday, will put it to test today when I get the chance. The LA90 should have no trouble with 5v input...
It feels heavy and sturdy, not very tall and also its footprint is quite a bit larger than expected.

 

[Jutlandx]

Smsl c200 vs dl200

Short:
Size and weight (obvious);
Both are sturdy an well built, but the dl200 motherboard layout is cleaner and feels more premium. No surprise here, because it's larger and there is more space to "work";
Can use dP1 on dl200 in USB 2.0. On C200 it must be on dP4 (same pc system);
Almost the same power output. Tried 4,4mm on both (note that dl200 is balanced). No major diference (at least for my ears). Only compared with the Kiwi Quintet with balanced cable;
Slight problem with dl200 BT antenna. The screw thread is shorter because it stays a bit away from de back panel (see picture),
so the antenna becomes a bit loose because it doesn't press all the way down. However, I didn't noticed any problem with the connection.
Also, removing the hex nut solves the issue. The screw is soldered on the motherboard and is sturdy enough.
This is only a brief impression.

 

[Bleib]

The screw thread is shorter because it stays a bit away from de back panel (see picture),
so the antenna becomes a bit loose because it doesn't press all the way down. However, I didn't noticed any problem with the connection.

I have the same issue here too, a bit strange that this quality is worse than C200 which I've owned (as well as many Toppings)

 

[Jutlandx]

I have the same issue here too, a bit strange that this quality is worse than C200 which I've owned (as well as many Toppings)

Yes... it's a design flaw but doesn't compromise the BT signal. You can remove the hex nut if you are uncomfortable with the loose antenna.

 

[nerdemoji]

Yes... it's a design flaw but doesn't compromise the BT signal. You can remove the hex nut if you are uncomfortable with the loose antenna.

Well this is a goofy problem. Weird that they wouldn’t fix something simple like this.

 

[Catsmeow]

Are the TRS outputs attenuated via the knob on front?

 

[Grooved]

I have compared all AK4490 - AK4499 with ES9039Pro, 9039 clear winners. Their Hyperstream IV is really something special in terms of sound, 9038 sounds much more synthetic.

What instrument did you use to match output voltages during the comparison, how did you assure blind test conditions, and what was your typical confidence score for detecting the right DAC?

I totally agree with you @staticV3 , if nothing of this is checked, it can't be a serious comparison.

Now, I think that there is one thing that can make two DACs sound slightly different, and it's something that is easy to forget to check: do they use the same default filter (or if filter has been changed)?

We are always looking at high frequencies change between filters, but it also changes the impulse responses, and the very little difference I think I hear in some tracks (I would like to measure exactly what it seems to do but don't know what measurement to use to show it) is not in the high frequencies, but is more on the low frequencies response, for example the interaction between a kick drum and a bass.
I tested it these last days with the Tone2 Pro and balanced headphones output, switching between filters, and it looked like it made a little change in the low frequencies (more on transients and length of perceived notes than on overall frequency response itself).

To try to find this in a measurement, I then linked the Tone2 Pro to the Cosmos ADC, and tried a loopback recording test:
the difference found by DeltaWave was different depending on the filters, and if I remember correctly, it was a 8dB delta (I think the difference with the original track was between -46db and -54dB depending on the filters)

So if there can be an audible difference between two DACs, chipsets, devices... I would first check that they are using the same filters before checking the other things and then starting to compare them.

Now, I imagine that between two DACs with different but good enough SINAD and Multitone results, AND using the same filters, they should sound the same.

 

[funkyblue]

I'm looking at replacing a Loxjie A30. Would this be a good replacement? I know it's not been reviewed yet, but I'm willing to take a chance.

 

[Bleib]

I'm looking at replacing a Loxjie A30. Would this be a good replacement? I know it's not been reviewed yet, but I'm willing to take a chance.

So far experiences with it are good, just like with C200. Note though that it outputs 2.5/5v instead of the regular 2/4v which are more commonly used

 

[Tom22]

So far experiences with it are good, just like with C200. Note though that it outputs 2.5/5v instead of the regular 2/4v which are more commonly used

What does this mean if you hook it up to an amp via TRS/XLR - you can lower the gain setting further? Or do you expect problems?