This is a review and detailed measurements of the Airist R2R DAC. It is on kind loan from a member a while back who has been patiently waiting for its review. The Airist R2R costs US $350. Drop's website says 931 are sold so it seems like a popular DAC.
The Airist DAC employs the attractive design language of other products from massdrop/drop.com:
It is less tall than some of the other massdrop products making it even more attractive looking.
A single button selects the input format and that is all she told.
The back panel is as you expect:
In this price I like to see XLR balanced output which is not here. The reason is that balanced configuration doubles the cost of such DACs so it is naturally not included.
As the name indicates, the Airist uses a set of precision resistors to translate digital values to analog. This sounds good on paper but in reality presents many challenges in design which is the reason the industry moved away from them years ago.
A DAC is a simple device in that it takes digital samples and produces hopefully corresponding analog voltage values. It has no intelligence to know what it is playing. It is a simple "domain translator" from digital representation of audio samples to analog values. Such a device therefore can be simply characterized using measurements. So let's get into that.
DAC Audio Measurements
Feeding a 1 kHz tone and measuring the RCA outputs produced a slightly hotter signal at 2.5 volt which is fine. To be fair to it though as compared to DACs that produce the nominal 2 volts, I reduced the input level to get similar output:
We see a large spray of harmonics of our 1 kHz tone, easily travelling to 20 kHz. Unlike other products with such high distortion levels, the power of harmonics remains almost equal all the way to 20 kHz. That means that they step on detail in your music and obscure it with distortion products. SINAD which represents sum of distortion and noise as a result is very poor for a DAC, placing the Airist R2R DAC at the bottom of every DAC we have tested:
You can see its neighbors in the enlarged inset. I guess they can feel good that they beat the PS Audio DirectStream (DS) DAC.
The non-linearities in the DAC is also seen as intermodulation distortion among the 32 tone test to resemble music:
Your music has more than 32 tones so you can imagine a tall grass at the bottom of the output of this DAC hiding any detail that may be there in the source.
The DAC is reasonably quiet though:
So it is the distortion that is the problem, not noise level.
The biggest hit to its performance comes in the form of SMPTE intermodulation distortion test where we feed the dac a 60 Hz and 7 kHz tone. This shows non-linearities at both low and high frequencies at once:
Wow. We see a new curve we have not seen. For one thing the overall shape is flat which indicates high distortion level swamping noise at all input sample values. That is not the real killer though. The real killer is the clearly seen pattern which indicates distortion pattern that correlates (depends) on input digital samples. If you are going to have distortion and noise, you want it to be uncorrelated to what is being played so that the brain can tune it out. The constant hum from the fan in your computer is easier to ignore than if it kept going up and down based on what you did.
A measurement like this should have ran the alarm bells in the mind of the designer and a fix deployed. That it was not either measured or fixed, is unforgiving. Perhaps it is a cost cutting thing.
THD+N (distortion+noise) relative to frequency showed an ugly picture as well:
Where our hearing is most sensitive -- 3 to 4 kHz, distortion rises up to another 10 to 15 dB than our dashboard with 1 kHz indicates. So the news here is even worse than we have seen so far. Rise in ultrasonic distortion can be due to out of band signals unrelated to what we are playing. So let's look at that:
We see that our 10 kHz tone generates taller harmonic distortion spikes so this is clear non-linearity which increases with frequency.
That said, the reconstruction filter is weak:
I like to see 90 dB or more attenuation for ultrasonics.
Frequency response is dead flat and hence good:
So any tonal variation will be due to distortion products.
Linearity is surprisingly good:
This test ignores almost all noise and distortion so the ills of the DAC are hidden out of sight.
Jitter test shows random noise polluting our signal:
The graph is thickened and there is a skirt around the bottom of our main 12 kHz tone.
Conclusions
DAC chip companies have in a way commoditized the market. So I appreciate designers trying to differentiate by producing DACs built from scratch. If you are going to do this though, you better not compromise performance. Yes, it is trivial in audio to tell a story and get a bunch of people to believe and praise the "sound" of a DAC using improper listening tests. But really, an engineer should know better. They should know that distorting the input signal is not what the artist and talent intend when they distributed their music. There is no alternative universe here were bad things are good. You don't eat out of dirty dishes thinking it adds flavor to your feed. Why do that in the case of DAC?
Needless to say, the Airist R2R DAC should be in a category of audio DACs that are avoided. As such, I can't recommend it.
Now, if you want to have a conversation piece, take the guts out, put in the ones from the many $99 DACs that have superb performance, and have best of both worlds! You won't be crapping on your audio samples while having a unique audio DAC, if in looks only.
--------
As always, questions, comments, corrections, etc. are welcome.
Get this: the beheaded panther shown in this review wants to learn to ride a bike!!! Beside the fact that he can't see where he is going, I have told him that his legs don't move either but still he doesn't listen. I don't want him to lose more of his limbs in trying to ride a bike so am thinking of sending him to a psychiatrist to explain all of this to him. Alas, the ones I have priced want $350 per visit so need money for that. Appreciate you all donating money for this great cause using : https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
The Airist DAC employs the attractive design language of other products from massdrop/drop.com:
It is less tall than some of the other massdrop products making it even more attractive looking.
A single button selects the input format and that is all she told.
The back panel is as you expect:
In this price I like to see XLR balanced output which is not here. The reason is that balanced configuration doubles the cost of such DACs so it is naturally not included.
As the name indicates, the Airist uses a set of precision resistors to translate digital values to analog. This sounds good on paper but in reality presents many challenges in design which is the reason the industry moved away from them years ago.
A DAC is a simple device in that it takes digital samples and produces hopefully corresponding analog voltage values. It has no intelligence to know what it is playing. It is a simple "domain translator" from digital representation of audio samples to analog values. Such a device therefore can be simply characterized using measurements. So let's get into that.
DAC Audio Measurements
Feeding a 1 kHz tone and measuring the RCA outputs produced a slightly hotter signal at 2.5 volt which is fine. To be fair to it though as compared to DACs that produce the nominal 2 volts, I reduced the input level to get similar output:
We see a large spray of harmonics of our 1 kHz tone, easily travelling to 20 kHz. Unlike other products with such high distortion levels, the power of harmonics remains almost equal all the way to 20 kHz. That means that they step on detail in your music and obscure it with distortion products. SINAD which represents sum of distortion and noise as a result is very poor for a DAC, placing the Airist R2R DAC at the bottom of every DAC we have tested:
You can see its neighbors in the enlarged inset. I guess they can feel good that they beat the PS Audio DirectStream (DS) DAC.
The non-linearities in the DAC is also seen as intermodulation distortion among the 32 tone test to resemble music:
Your music has more than 32 tones so you can imagine a tall grass at the bottom of the output of this DAC hiding any detail that may be there in the source.
The DAC is reasonably quiet though:
So it is the distortion that is the problem, not noise level.
The biggest hit to its performance comes in the form of SMPTE intermodulation distortion test where we feed the dac a 60 Hz and 7 kHz tone. This shows non-linearities at both low and high frequencies at once:
Wow. We see a new curve we have not seen. For one thing the overall shape is flat which indicates high distortion level swamping noise at all input sample values. That is not the real killer though. The real killer is the clearly seen pattern which indicates distortion pattern that correlates (depends) on input digital samples. If you are going to have distortion and noise, you want it to be uncorrelated to what is being played so that the brain can tune it out. The constant hum from the fan in your computer is easier to ignore than if it kept going up and down based on what you did.
A measurement like this should have ran the alarm bells in the mind of the designer and a fix deployed. That it was not either measured or fixed, is unforgiving. Perhaps it is a cost cutting thing.
THD+N (distortion+noise) relative to frequency showed an ugly picture as well:
Where our hearing is most sensitive -- 3 to 4 kHz, distortion rises up to another 10 to 15 dB than our dashboard with 1 kHz indicates. So the news here is even worse than we have seen so far. Rise in ultrasonic distortion can be due to out of band signals unrelated to what we are playing. So let's look at that:
We see that our 10 kHz tone generates taller harmonic distortion spikes so this is clear non-linearity which increases with frequency.
That said, the reconstruction filter is weak:
I like to see 90 dB or more attenuation for ultrasonics.
Frequency response is dead flat and hence good:
So any tonal variation will be due to distortion products.
Linearity is surprisingly good:
This test ignores almost all noise and distortion so the ills of the DAC are hidden out of sight.
Jitter test shows random noise polluting our signal:
The graph is thickened and there is a skirt around the bottom of our main 12 kHz tone.
Conclusions
DAC chip companies have in a way commoditized the market. So I appreciate designers trying to differentiate by producing DACs built from scratch. If you are going to do this though, you better not compromise performance. Yes, it is trivial in audio to tell a story and get a bunch of people to believe and praise the "sound" of a DAC using improper listening tests. But really, an engineer should know better. They should know that distorting the input signal is not what the artist and talent intend when they distributed their music. There is no alternative universe here were bad things are good. You don't eat out of dirty dishes thinking it adds flavor to your feed. Why do that in the case of DAC?
Needless to say, the Airist R2R DAC should be in a category of audio DACs that are avoided. As such, I can't recommend it.
Now, if you want to have a conversation piece, take the guts out, put in the ones from the many $99 DACs that have superb performance, and have best of both worlds! You won't be crapping on your audio samples while having a unique audio DAC, if in looks only.
--------
As always, questions, comments, corrections, etc. are welcome.
Get this: the beheaded panther shown in this review wants to learn to ride a bike!!! Beside the fact that he can't see where he is going, I have told him that his legs don't move either but still he doesn't listen. I don't want him to lose more of his limbs in trying to ride a bike so am thinking of sending him to a psychiatrist to explain all of this to him. Alas, the ones I have priced want $350 per visit so need money for that. Appreciate you all donating money for this great cause using : https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
Last edited: