This is a review and detailed measurements of the Woo Audio tube WA7 DAC and headphone amplifier together with its companion WA7tp tube power supply. It is on kind loan from a member. The cost of the pair is USD $1,599 so not cheap. As of this writing, shipping is free but I assume you would otherwise have to pay for it.
Note: this is the generation 1 of the WA7. The generation 2 is supposed to have a better DAC (?).
I remember running into Woo Audio amp a few years ago and drooled over its industrial design. The WA7 series is not quite up there with the rest of the units but still benefits from its massive plastic transparent top to give it elegance and drool appeal:
Strangely, the power supply has a rotary control which does nothing. You push it in and it power cycles on and off. A multi-pin short cable attaches the two. Reading the reviews, there are complaints that no AC power cord comes with WA7tp. What the heck?
As you can imagine, these two boxes are heavy and stay where you put them. There are two headphone jacks: the 3.5mm has lower gain so I focused my testing on the 1/4 inch jack.
The tubes in the unit I tested are an upgrade to Russian made, 6C45Pi Electro Harmonix with nice looking gold pins. Woo sells a matched pair for USD $140.
Unusual in tube amp category, there is an ESS DAC in there with USB input. There is also a set of RCA analog in and a switch for high and low impedance output.
I watched a few video reviews for the WA7 and there were unanimously positive. One reviewer complained about the DAC not being good so in my testing, I went both ways: with the DAC input and analog.
Measurements
With combo DAC and headphone inputs, I always try to test the DAC portion first by setting the output impedance high as to not stress the amplifier and set the voltage to 2 volts. Here is what we get with WA7 combo:
Yuck. We have a heap of distortion and a mains peak which I could not get rid of by playing with grounding. The SINAD score (signal over noise and distortion) naturally lands the WA7 at the bottom of the pile:
Is the DAC implementation broken? Let's look at the same performance with analog input:
Nope. The digital input was actually better by 4 dB or so.
Using digital input again, dynamic range is actually decent for a device with so much distortion:
Frequency response was fairly rolled off at both ends:
We are down 3.4 dB at 20 kHz. We are OK at 20 Hz but from there on, we lose amplitude there.
Switching to high z (z standard for impedance), raises the output voltage substantially. Similar roll off exists.
Linearity was very poor indicating perhaps 16 bit implementation or just a lot of noise:
Between this and high frequency roll off previously, don't bother with any "high-res" content for the sake of bandwidth or resolution.
EDIT: was asked to run multitone tests in the thread and here is that:
It is by far the worst results I have seen since I started running this test on DACs. The tall spikes are the 32 tones the DAC is asked to play. It does that but then generates tons of distortion spikes in between tones which can make the music sound bright and reduce effective resolution of music (by masking low level music detail with distortion products).
The all important test for headphone amplifier is power versus distortion and noise. Here is that, with 300 ohm load:
I will give you the good news first: there is plenty of power, especially in high-z mode. It is the highest I have measured at nearly 1 watt in that setting!
Using the more sane low-z mode though, the amplifier starts become distortion dominant at just 1 milliwatt. It is downhill from there up to max of 1/4 watt at 0.6% distortion+noise. This is one noise and distortion factory compared to our reference Massdrop THX AAA 789 (in dashed blue) at 1/5th the price.
Here is the picture with 33 ohm load:
There is still plenty of power and the same soft rise in distortion. At peak power though, we are now in single percentage numbers with max output at 6%. Fortunately you will go deaf well below that level so not a concern there. The worry is that lowest distortion at noise is at just 0.2 milliwatt. Not 0.2 watt but 0.2 milliwatt! Needless to say, at any listening volume you are going to be experiencing distortion and noise with the WA7.
As if these numbers weren't disappointing enough, we get to measure the output impedance:
Even in "low z" mode the impedance is nearly 30 ohm. With high Z you have 80 ohm. You have a built-in equalizer here for just about any headphone that doesn't have a flat frequency response. And it is an equalizer you don't have a control over.
Channel imbalance was pretty good showing a constant level mismatch for most of the volume range:
Let's stop here as the advocates of tube amps are screaming as loud as they can: "it is all about the ear; of course tube amps don't measure well."
Listening Tests
I started my testing by playing a 1 khz tone and matching levels between my Topping DX3 Pro combo DAC and Amp and WA7. This proved to be useless as the volume level on the WA7 was content specific due to variations it caused due to its high output impedance. So I resorted to matching levels by ear.
First up was Sennheiser HD-650. On my bass heaving headphone test tracks, the bass on WA7 was more muted. It was also a bit more bloated compared to the tight bass from DX3 Pro. This was subtle. What was less subtle was the brighter signature of the WA7. Highs were exaggerated. The lisp in vocals became worse and high frequencies became grungier. Note that this effect was most audible during periods with strong bass frequencies.
On detail and resolution front, the WA7 lost it there too. The case was loud instruments that vibrated up and down rapidly in level. The detail in between the peaks was nicely rendered by the DX3 Pro but became muddy and lost in WA7. Note that this was not across the board. On content with infrequent transients, the issue was not there.
Switching to Hifiman HE-400i showed similar issues just the same although not always with the same tracks the same way.
These two headphones represent difficult loads for amplifiers since one is voltage hungry, and the other, current. So I thought I throw my AKG K92 headphone in there too. The K92 is extremely efficient compared to the others and is closed back to boot, letting one better examine detail.
The power output was impressive from both devices. I had to turn the levels way, way down for comfortable listening. Level matching became much more difficult here as even during a single track, the WA7 would be louder and less loud. Clearly the frequency response of the K92 was modified by the WA7. Working around that, I heard the same issue with lack of bass impact and resolution in WA7.
Since I had so much power available with K92 I decided to try a new experiment. Try not to laugh but I put in two foam ear plugs in my ear. No, these are not musician ones. They are normal noise reduction yellow foam plugs. What this did was essentially cut off everything from mid to high frequencies and heavily attenuated the bass. I was then able to turn the volume up substantially and could easily hear the bass becoming more distorted in WA7 than Topping DX3 Pro.
Correlating Measurements with Listening Tests
Can we explain the subjective results using the objective data? Sure we can. Take a look at at the FFT spectrum again:
The only thing we told the WA 7 to produce was that 1 kHz tone. What we got (ignoring the left side of the graph) was that tone plus a bunch of harmonics. By definition this 1 kHz tone just got "brighter" since we now added more frequencies going up to 20 kHz and above. Now, the brain masks some of the earlier ones such as some of the 2nd and 3rd harmonics. But the rest at some point can become audible. So one would expect harmonic distortion of sufficient level to make everything sound bright. Indeed if you turn up an amplifier beyond its output power to clipping, it does just that. This is what I was hearing in my testing.
Let's consider that in our music we have many tones. Each tone creates that spray of harmonic so together they add up and increase the high frequency content. But they do more than that: the harmonics now land on top of low level detail in music. That would obscure them and potentially mask them completely. Again, this is what I was hearing with detail getting lost between peaks.
Now the audible effects are nowhere as extreme as the measurements. Reason is that there is a lot of masking going on where the brain is throwing away lower level sounds -- whether they are distortion or musical detail. Selecting the right content, and playback level, plus focusing on specific attributes of music was necessary to hear these changes.
Most importantly, you need to have a reference. If I just listened to WA7 by itself, I would have no idea that these changes had occured. Your memory is simply not good enough to hear 10% more high frequency detail during some segments of some musical tracks. It was the instant AB switching with Topping DX3 Pro which allowed me to find and zoom in on them. All the reviewers who just listen to WA7 by itself have no prayer of observing any of this.
I should say that the fidelity I heard was 90% due to the content itself. As such, the WA7 can sound exceptional depending on how good the source content is. So don't think that measurements say it produces garbage. Again, perceptual masking saves us from hearing a lot of what the measurements show.
Conclusions
Yet again we measure a tube amplifier which produces copious amount of distortion. However, unlike some budget products, the WA7 is a quite amplifier and has major conveniences such as built-in DAC. It is also pretty to look at. For me as an audiophile, the WA7 takes a step backward. Last thing I want is the highs getting extra sizzle or the tight, impactful bass losing some of both characteristics.
As I have experienced time and time again in controlled testing, there is nothing euphonic or positive with such tube amplifiers. No soundstage is changed. No instrument is isolated more (it is actually worse with WA7). Detail is actually lost. The sound definitely is not warmer either. People read these attributes into such devices because they think they are there. A simple AB test with a $20 switcher is enough to convince you otherwise.
Lack of fidelity in measurements simply translates into lack of transparency and listening pleasure.
Now, if you are coming from a low power or no headphone amplifier, the WA7 will be a revelation. But remember, you can have your cake and eat it too by buying 100% transparent headphone amplifier with tons of power such as the $99 JDS Labs Atom. Or the $370 Massdrop THX AAA 789. From that vantage point, you will be going backward with Woo Audio WA7 in all respects but looks.
------------
As always, questions, comments, recommendations, etc. are welcome.
I will make this simple: you can never have enough money. So please consider donating 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).
Note: this is the generation 1 of the WA7. The generation 2 is supposed to have a better DAC (?).
I remember running into Woo Audio amp a few years ago and drooled over its industrial design. The WA7 series is not quite up there with the rest of the units but still benefits from its massive plastic transparent top to give it elegance and drool appeal:
Strangely, the power supply has a rotary control which does nothing. You push it in and it power cycles on and off. A multi-pin short cable attaches the two. Reading the reviews, there are complaints that no AC power cord comes with WA7tp. What the heck?
As you can imagine, these two boxes are heavy and stay where you put them. There are two headphone jacks: the 3.5mm has lower gain so I focused my testing on the 1/4 inch jack.
The tubes in the unit I tested are an upgrade to Russian made, 6C45Pi Electro Harmonix with nice looking gold pins. Woo sells a matched pair for USD $140.
Unusual in tube amp category, there is an ESS DAC in there with USB input. There is also a set of RCA analog in and a switch for high and low impedance output.
I watched a few video reviews for the WA7 and there were unanimously positive. One reviewer complained about the DAC not being good so in my testing, I went both ways: with the DAC input and analog.
Measurements
With combo DAC and headphone inputs, I always try to test the DAC portion first by setting the output impedance high as to not stress the amplifier and set the voltage to 2 volts. Here is what we get with WA7 combo:
Yuck. We have a heap of distortion and a mains peak which I could not get rid of by playing with grounding. The SINAD score (signal over noise and distortion) naturally lands the WA7 at the bottom of the pile:
Is the DAC implementation broken? Let's look at the same performance with analog input:
Nope. The digital input was actually better by 4 dB or so.
Using digital input again, dynamic range is actually decent for a device with so much distortion:
Frequency response was fairly rolled off at both ends:
We are down 3.4 dB at 20 kHz. We are OK at 20 Hz but from there on, we lose amplitude there.
Switching to high z (z standard for impedance), raises the output voltage substantially. Similar roll off exists.
Linearity was very poor indicating perhaps 16 bit implementation or just a lot of noise:
Between this and high frequency roll off previously, don't bother with any "high-res" content for the sake of bandwidth or resolution.
EDIT: was asked to run multitone tests in the thread and here is that:
It is by far the worst results I have seen since I started running this test on DACs. The tall spikes are the 32 tones the DAC is asked to play. It does that but then generates tons of distortion spikes in between tones which can make the music sound bright and reduce effective resolution of music (by masking low level music detail with distortion products).
The all important test for headphone amplifier is power versus distortion and noise. Here is that, with 300 ohm load:
I will give you the good news first: there is plenty of power, especially in high-z mode. It is the highest I have measured at nearly 1 watt in that setting!
Using the more sane low-z mode though, the amplifier starts become distortion dominant at just 1 milliwatt. It is downhill from there up to max of 1/4 watt at 0.6% distortion+noise. This is one noise and distortion factory compared to our reference Massdrop THX AAA 789 (in dashed blue) at 1/5th the price.
Here is the picture with 33 ohm load:
There is still plenty of power and the same soft rise in distortion. At peak power though, we are now in single percentage numbers with max output at 6%. Fortunately you will go deaf well below that level so not a concern there. The worry is that lowest distortion at noise is at just 0.2 milliwatt. Not 0.2 watt but 0.2 milliwatt! Needless to say, at any listening volume you are going to be experiencing distortion and noise with the WA7.
As if these numbers weren't disappointing enough, we get to measure the output impedance:
Even in "low z" mode the impedance is nearly 30 ohm. With high Z you have 80 ohm. You have a built-in equalizer here for just about any headphone that doesn't have a flat frequency response. And it is an equalizer you don't have a control over.
Channel imbalance was pretty good showing a constant level mismatch for most of the volume range:
Let's stop here as the advocates of tube amps are screaming as loud as they can: "it is all about the ear; of course tube amps don't measure well."
Listening Tests
I started my testing by playing a 1 khz tone and matching levels between my Topping DX3 Pro combo DAC and Amp and WA7. This proved to be useless as the volume level on the WA7 was content specific due to variations it caused due to its high output impedance. So I resorted to matching levels by ear.
First up was Sennheiser HD-650. On my bass heaving headphone test tracks, the bass on WA7 was more muted. It was also a bit more bloated compared to the tight bass from DX3 Pro. This was subtle. What was less subtle was the brighter signature of the WA7. Highs were exaggerated. The lisp in vocals became worse and high frequencies became grungier. Note that this effect was most audible during periods with strong bass frequencies.
On detail and resolution front, the WA7 lost it there too. The case was loud instruments that vibrated up and down rapidly in level. The detail in between the peaks was nicely rendered by the DX3 Pro but became muddy and lost in WA7. Note that this was not across the board. On content with infrequent transients, the issue was not there.
Switching to Hifiman HE-400i showed similar issues just the same although not always with the same tracks the same way.
These two headphones represent difficult loads for amplifiers since one is voltage hungry, and the other, current. So I thought I throw my AKG K92 headphone in there too. The K92 is extremely efficient compared to the others and is closed back to boot, letting one better examine detail.
The power output was impressive from both devices. I had to turn the levels way, way down for comfortable listening. Level matching became much more difficult here as even during a single track, the WA7 would be louder and less loud. Clearly the frequency response of the K92 was modified by the WA7. Working around that, I heard the same issue with lack of bass impact and resolution in WA7.
Since I had so much power available with K92 I decided to try a new experiment. Try not to laugh but I put in two foam ear plugs in my ear. No, these are not musician ones. They are normal noise reduction yellow foam plugs. What this did was essentially cut off everything from mid to high frequencies and heavily attenuated the bass. I was then able to turn the volume up substantially and could easily hear the bass becoming more distorted in WA7 than Topping DX3 Pro.
Correlating Measurements with Listening Tests
Can we explain the subjective results using the objective data? Sure we can. Take a look at at the FFT spectrum again:
The only thing we told the WA 7 to produce was that 1 kHz tone. What we got (ignoring the left side of the graph) was that tone plus a bunch of harmonics. By definition this 1 kHz tone just got "brighter" since we now added more frequencies going up to 20 kHz and above. Now, the brain masks some of the earlier ones such as some of the 2nd and 3rd harmonics. But the rest at some point can become audible. So one would expect harmonic distortion of sufficient level to make everything sound bright. Indeed if you turn up an amplifier beyond its output power to clipping, it does just that. This is what I was hearing in my testing.
Let's consider that in our music we have many tones. Each tone creates that spray of harmonic so together they add up and increase the high frequency content. But they do more than that: the harmonics now land on top of low level detail in music. That would obscure them and potentially mask them completely. Again, this is what I was hearing with detail getting lost between peaks.
Now the audible effects are nowhere as extreme as the measurements. Reason is that there is a lot of masking going on where the brain is throwing away lower level sounds -- whether they are distortion or musical detail. Selecting the right content, and playback level, plus focusing on specific attributes of music was necessary to hear these changes.
Most importantly, you need to have a reference. If I just listened to WA7 by itself, I would have no idea that these changes had occured. Your memory is simply not good enough to hear 10% more high frequency detail during some segments of some musical tracks. It was the instant AB switching with Topping DX3 Pro which allowed me to find and zoom in on them. All the reviewers who just listen to WA7 by itself have no prayer of observing any of this.
I should say that the fidelity I heard was 90% due to the content itself. As such, the WA7 can sound exceptional depending on how good the source content is. So don't think that measurements say it produces garbage. Again, perceptual masking saves us from hearing a lot of what the measurements show.
Conclusions
Yet again we measure a tube amplifier which produces copious amount of distortion. However, unlike some budget products, the WA7 is a quite amplifier and has major conveniences such as built-in DAC. It is also pretty to look at. For me as an audiophile, the WA7 takes a step backward. Last thing I want is the highs getting extra sizzle or the tight, impactful bass losing some of both characteristics.
As I have experienced time and time again in controlled testing, there is nothing euphonic or positive with such tube amplifiers. No soundstage is changed. No instrument is isolated more (it is actually worse with WA7). Detail is actually lost. The sound definitely is not warmer either. People read these attributes into such devices because they think they are there. A simple AB test with a $20 switcher is enough to convince you otherwise.
Lack of fidelity in measurements simply translates into lack of transparency and listening pleasure.
Now, if you are coming from a low power or no headphone amplifier, the WA7 will be a revelation. But remember, you can have your cake and eat it too by buying 100% transparent headphone amplifier with tons of power such as the $99 JDS Labs Atom. Or the $370 Massdrop THX AAA 789. From that vantage point, you will be going backward with Woo Audio WA7 in all respects but looks.
------------
As always, questions, comments, recommendations, etc. are welcome.
I will make this simple: you can never have enough money. So please consider donating 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: