This is a review and detailed measurements of the Wyred4Sound DAC-2v2 SE. It is on kind loan from a member. The cost of the unit is USD $3,800. The one under test also has the "femto clock" option. I don't see a price for that even though it is shown as an option.
The unit itself is narrow but quite deep and heavy for its functionality:
The SE version is an upgraded unit with the following:
Yes, even an upgraded fuse!
I am not too happy to see a small and difficult to navigate interface in high-end devices. For this much money, give a nice display with dedicated buttons for things like setup please. And I like to see all the settings displayed and not crypting dashes and brackets.
There is an included remote which is nice. Not so nice is the rubberized finish that has become sticky. This happens when the finish degrades after many years. I don't think this unit is that old so surprised that it already has that awful feel.
If you are not familiar with Wyrd4Sound, they are what I call "affordable high-end." Their products are frequently reviewed and land a the bottom of tier of what high-end audiophiles shop for. They seem to have gone up in price though relative to what I remember of the company's original products.
Anyway, let's get into measurements and see if the bullet list presented above translates into performance.
Measurements
Since the outputs are variable, I set the balanced XLR to roughly 4.0 volts. Could not quit get there as the notches between volume levels was quite large so had to settle for 4.1 vol. As nice coincident, I got 2 volt output on RCA jacks which is what I like to see for unbalanced outputs. Let's look at our dashboard using XLR balanced connection:
Ouch! Are you kidding me? SINAD (signal to noise and distortion) below 90 dB? ESS specs that at 122 dB by the way so this is a shortfall of 30+ dB from their reference design. At first I thought maybe something is wrong with my instrumentation but then looked and they spec the unit at 0.002% distortion which is pretty close to what I was measuring. Theirs includes a-weighting and may have been done at lower digital levels. I played with that and indeed SINAD improves some but just to 92 dB or so.
The poor performance here lands the Wyred4Sound DAC-2v2 SE DAC in our fourth quadrant of performance (with first being the best):
Low SINAD is not the only issue visible in the dashboard. If you look at the FFT spectrum of our 1 kHz tone, we see a wide skirt under that tone. That usually indicates jitter. Jitter is proportional to frequency so even in devices that have fair bit of them, we don't see it at just 1 kHz. For it to show up there it means jitter must be pretty high. Or else, there are other sources of noise.
Let's run our jitter test and see what is going on there:
The exponentially decaying curve indicates randomness in the spectrum which could either be jitter as indicated, or pollution of the DAC reference voltage with noise. Seeing how it looks mostly the same and hence is not very frequency dependent, it might be voltage modulation and not jitter. Regardless of the cause, this is totally unacceptable performance for a high-end DAC, especially one that makes the buyer think there is ultra precision there by that "femto clock" notion.
The extra power supply spikes and otherwise in 20 to 2 kHz is also something that should not be there in a high-end products.
Fortunately the levels are very low here so not audible but definitely a failed engineering attempt.
BTW, note in the original dashboard FFT how the noise floor is different in the two channel. In an expensive product like this, the two channels better match. Clearly one is closer to the source of noise/interference than the other. We see this clearly in our multi-tone test:
The channel in red clearly has higher noise floor. Both channels have elevated noise floor as we often see in DACs but it is broader here. And again, less tolerable in higher-end DACs.
I talked about how the performance improves if I reduce the input level to the DAC and we see that readily in intermodulation distortion versus level:
Both XLR and RCA outputs start to become dominated by distortion at around -10 dBFS and keep getting worse from there on. And at any rate, both have elevated noise level compared to the Topping DX3 Pro at 1/10th its price.
Dynamic range is OK but way short of specified 120 dB:
One may be able to get there by increasing the volume to max rather than nominal that I have used throughout the test.
THD+N (distortion+noise) relative to frequency with wide bandwidth of 90 kHz, shows really poor performance:
Not only do we have elevated levels of THD+N, but the analyzer cannot get a stable reading below 500 Hertz (symbol "T") indicating lots of variability/noise.
It is not all bad news though. Linearity is very good:
The good numbers come from the fact that linearity test filters out all but a single frequency that is being generated so all the harmonic distortion distortion and broadband noise is eliminated.
White noise test to show how steep the reconstruction filter is, shows very good performance:
We have 75 dB of attenuation by half of our sampling rate at 22.05 kHz or so. I have tested recent DACs that only show 7 to 10 dB here! Theory specifies infinite attenuation by the way.
Conclusions
We run a lot of tests in these reviews but I often say that as soon as I look at the Dashboard, I know more or less where performance lands and that is the case here. The Wyred4Sound DAC-2v2 SE's performance is poor by budget desktop DACs. In the context of high-end DACs, it is totally unacceptable. It appears that they are one of the companies that designs products based on shopping list of bullets from audiophiles and their myths: "femto this," "fancy fuse that." If there is ever proof that these statements have no value, this is it.
This whole "femto clock" thing is such an improper notion anyway. When it comes to DACs, we don't care about absolute accuracy of the DAC. You can see in the dashboard of DAC tests where our 1 kHz tone is produced at slightly lower frequency or higher. None of that is audible to us. Heck, many audiophiles listen to analog sources whose speed is off far more than any digital transport.
What we care about with respect to DAC clock is that it is free of variation, not absolute accuracy. Variations of the clock modulate our music and create distortion products we call jitter. If the variations are random they show up as increasing noise floor. If they are single tone, we see them as spikes. No distortion shows up whatsoever if the DAC runs 0.001% slower or faster. And your hearing cannot detect pitch differences that small.
What this means is that whether a DAC clock is good or not needs measurements of the final output of the DAC as we perform in every DAC review. Variations are trivial to detect there because our instrumentation goes as deep as we want to go to tune of -160 dB or even better! That is nearly 50 dB better than threshold of hearing. So we can conclude two things here: Wyred4Sound either doesn't understand any of this or doesn't measure such. Neither is acceptable. No high-end DAC should ship a product that subjects every tone in your music to randomness.
Before someone says, "ah, but how does it sound" let me tell you that if you think this product "sounds good" then your ears have no sensitivity to small distortions. Our instrumentation is far more sensitive and show that for all the money paid, you got less of a DAC than even budget products at $100. That you don't hear the problems is a blessing as human beings that our hearing is not that good, no matter how much our audiophile ego says otherwise.
There is just no way, no how a designer sat there and said, "let's throw a bunch of unequal, signal dependent noise at the bottom of each tone in your music because it sounds better." They simply implemented the circuit poorly, taking away much of the performance that ESS worked so hard to put in their flagship DAC chip: ESS9038 Pro. The goal of any DAC designer should be to come as close as they possibly can to the performance of the chip as stated by the manufacture of said chip if not exceed it. Trashing it by many dBs is just not right. You are paying for a product whose implementation is worse than the reference design.
My suggestion to Wyred4Sound is to proper measure this DAC and find and fix sources of these problems. They can still make all the claims they want in there but at least it will be based on solid foundation of engineering. No one would sit in a 5 star restaurant and eat food of dirty dishes. Unfortunately that is what we have here.
I like to finish by saying that my impression of Wyred4Sound up to now has been positive based on what I have read in press. So it was disappointing to see such failing performance in their high-end offering this way.
At the risk of stating the very obvious, I cannot recommend the Wyred4Sound DAC-2v2 SE. Buy a much cheaper desktop DAC and put the rest of the budget toward good music, food or whatever other hobbies you have.
------------
As always, questions, comments, recommendations, etc. are welcome.
They say you should save money for a rainy day. It rains for about 6 months here so I need plenty of money!!! Please consider donating funds using:
Patreon: https://www.patreon.com/audiosciencereview), or
upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
The unit itself is narrow but quite deep and heavy for its functionality:
The SE version is an upgraded unit with the following:
Yes, even an upgraded fuse!
I am not too happy to see a small and difficult to navigate interface in high-end devices. For this much money, give a nice display with dedicated buttons for things like setup please. And I like to see all the settings displayed and not crypting dashes and brackets.
There is an included remote which is nice. Not so nice is the rubberized finish that has become sticky. This happens when the finish degrades after many years. I don't think this unit is that old so surprised that it already has that awful feel.
If you are not familiar with Wyrd4Sound, they are what I call "affordable high-end." Their products are frequently reviewed and land a the bottom of tier of what high-end audiophiles shop for. They seem to have gone up in price though relative to what I remember of the company's original products.
Anyway, let's get into measurements and see if the bullet list presented above translates into performance.
Measurements
Since the outputs are variable, I set the balanced XLR to roughly 4.0 volts. Could not quit get there as the notches between volume levels was quite large so had to settle for 4.1 vol. As nice coincident, I got 2 volt output on RCA jacks which is what I like to see for unbalanced outputs. Let's look at our dashboard using XLR balanced connection:
Ouch! Are you kidding me? SINAD (signal to noise and distortion) below 90 dB? ESS specs that at 122 dB by the way so this is a shortfall of 30+ dB from their reference design. At first I thought maybe something is wrong with my instrumentation but then looked and they spec the unit at 0.002% distortion which is pretty close to what I was measuring. Theirs includes a-weighting and may have been done at lower digital levels. I played with that and indeed SINAD improves some but just to 92 dB or so.
The poor performance here lands the Wyred4Sound DAC-2v2 SE DAC in our fourth quadrant of performance (with first being the best):
Low SINAD is not the only issue visible in the dashboard. If you look at the FFT spectrum of our 1 kHz tone, we see a wide skirt under that tone. That usually indicates jitter. Jitter is proportional to frequency so even in devices that have fair bit of them, we don't see it at just 1 kHz. For it to show up there it means jitter must be pretty high. Or else, there are other sources of noise.
Let's run our jitter test and see what is going on there:
The exponentially decaying curve indicates randomness in the spectrum which could either be jitter as indicated, or pollution of the DAC reference voltage with noise. Seeing how it looks mostly the same and hence is not very frequency dependent, it might be voltage modulation and not jitter. Regardless of the cause, this is totally unacceptable performance for a high-end DAC, especially one that makes the buyer think there is ultra precision there by that "femto clock" notion.
The extra power supply spikes and otherwise in 20 to 2 kHz is also something that should not be there in a high-end products.
Fortunately the levels are very low here so not audible but definitely a failed engineering attempt.
BTW, note in the original dashboard FFT how the noise floor is different in the two channel. In an expensive product like this, the two channels better match. Clearly one is closer to the source of noise/interference than the other. We see this clearly in our multi-tone test:
The channel in red clearly has higher noise floor. Both channels have elevated noise floor as we often see in DACs but it is broader here. And again, less tolerable in higher-end DACs.
I talked about how the performance improves if I reduce the input level to the DAC and we see that readily in intermodulation distortion versus level:
Both XLR and RCA outputs start to become dominated by distortion at around -10 dBFS and keep getting worse from there on. And at any rate, both have elevated noise level compared to the Topping DX3 Pro at 1/10th its price.
Dynamic range is OK but way short of specified 120 dB:
One may be able to get there by increasing the volume to max rather than nominal that I have used throughout the test.
THD+N (distortion+noise) relative to frequency with wide bandwidth of 90 kHz, shows really poor performance:
Not only do we have elevated levels of THD+N, but the analyzer cannot get a stable reading below 500 Hertz (symbol "T") indicating lots of variability/noise.
It is not all bad news though. Linearity is very good:
The good numbers come from the fact that linearity test filters out all but a single frequency that is being generated so all the harmonic distortion distortion and broadband noise is eliminated.
White noise test to show how steep the reconstruction filter is, shows very good performance:
We have 75 dB of attenuation by half of our sampling rate at 22.05 kHz or so. I have tested recent DACs that only show 7 to 10 dB here! Theory specifies infinite attenuation by the way.
Conclusions
We run a lot of tests in these reviews but I often say that as soon as I look at the Dashboard, I know more or less where performance lands and that is the case here. The Wyred4Sound DAC-2v2 SE's performance is poor by budget desktop DACs. In the context of high-end DACs, it is totally unacceptable. It appears that they are one of the companies that designs products based on shopping list of bullets from audiophiles and their myths: "femto this," "fancy fuse that." If there is ever proof that these statements have no value, this is it.
This whole "femto clock" thing is such an improper notion anyway. When it comes to DACs, we don't care about absolute accuracy of the DAC. You can see in the dashboard of DAC tests where our 1 kHz tone is produced at slightly lower frequency or higher. None of that is audible to us. Heck, many audiophiles listen to analog sources whose speed is off far more than any digital transport.
What we care about with respect to DAC clock is that it is free of variation, not absolute accuracy. Variations of the clock modulate our music and create distortion products we call jitter. If the variations are random they show up as increasing noise floor. If they are single tone, we see them as spikes. No distortion shows up whatsoever if the DAC runs 0.001% slower or faster. And your hearing cannot detect pitch differences that small.
What this means is that whether a DAC clock is good or not needs measurements of the final output of the DAC as we perform in every DAC review. Variations are trivial to detect there because our instrumentation goes as deep as we want to go to tune of -160 dB or even better! That is nearly 50 dB better than threshold of hearing. So we can conclude two things here: Wyred4Sound either doesn't understand any of this or doesn't measure such. Neither is acceptable. No high-end DAC should ship a product that subjects every tone in your music to randomness.
Before someone says, "ah, but how does it sound" let me tell you that if you think this product "sounds good" then your ears have no sensitivity to small distortions. Our instrumentation is far more sensitive and show that for all the money paid, you got less of a DAC than even budget products at $100. That you don't hear the problems is a blessing as human beings that our hearing is not that good, no matter how much our audiophile ego says otherwise.
There is just no way, no how a designer sat there and said, "let's throw a bunch of unequal, signal dependent noise at the bottom of each tone in your music because it sounds better." They simply implemented the circuit poorly, taking away much of the performance that ESS worked so hard to put in their flagship DAC chip: ESS9038 Pro. The goal of any DAC designer should be to come as close as they possibly can to the performance of the chip as stated by the manufacture of said chip if not exceed it. Trashing it by many dBs is just not right. You are paying for a product whose implementation is worse than the reference design.
My suggestion to Wyred4Sound is to proper measure this DAC and find and fix sources of these problems. They can still make all the claims they want in there but at least it will be based on solid foundation of engineering. No one would sit in a 5 star restaurant and eat food of dirty dishes. Unfortunately that is what we have here.
I like to finish by saying that my impression of Wyred4Sound up to now has been positive based on what I have read in press. So it was disappointing to see such failing performance in their high-end offering this way.
At the risk of stating the very obvious, I cannot recommend the Wyred4Sound DAC-2v2 SE. Buy a much cheaper desktop DAC and put the rest of the budget toward good music, food or whatever other hobbies you have.
------------
As always, questions, comments, recommendations, etc. are welcome.
They say you should save money for a rainy day. It rains for about 6 months here so I need plenty of money!!! Please consider donating funds using:
Patreon: https://www.patreon.com/audiosciencereview), or
upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
Last edited: