This is a review and detailed measurements of the Benchmark HPA4 headphone amplifier and preamplifier. The company was kind enough to loan me one for testing. The HPA4 retails for US $2999 so not cheap. It is a collaboration between Benchmark and THX to implement the "888" standard. Will it outperform the much cheaper THX designed Massdrop/Drop AAA 789? That is what we will attempt to determine later in this review.
The HPA4 is a departure from other Benchmark products with the addition of a color touchscreen:
The screen looks better in person (and in my picture) than what I have seen elsewhere. Still, it is a juxtaposition of industrial/professional look.
The display does not do a lot but hugely simplifies the configuration of the HPA4.
Volume control is a stepped attenuator implemented with a ton of relays with gold plated contacts that is stated to keep working long after I am dead and buried.
There are two headphone outputs but the XLR connection is the same as 1/4 inch with the exception of independent return wire for each channel instead of shared. Benchmark also states that the XLR connection makes a more secure, lower impedance connection. I tested both.
In addition to being a headphone amplifier, the HPA4 is also a pre-amplifier:
You have multiple inputs in different flavors. And of course RCA and XLR outputs. I have not looked but I assume the mono out can be used to drive a subwoofer.
Nice set of trigger ins and outs are provided so that you can power a bunch of gear at once from one source unit.
In operation, the HPA4 runs quite warm for a headphone amplifier/pre-amp. Nothing to be concerned about with respect to reliability but you may not want to leave it on all the time. Performance is instantly up to spec so there is no need for warm up anyway.
A switching power supply assures universal compatibility with whatever mains voltage is in your country.
Let's get into testing but a few notes. First, as with Benchmark AHB2, I was sent to the drawing board to improve the quality of my test fixture. The dummy load I had for testing at 33 Ohm has too many connectors/switches in the path so I had to build an optimized, dedicated one to eliminate sources of distortion there. I have also revised some of the test, putting emphasis on dB metrics for distortion versus percentage. The testing is als more extreme than some of my other reviews due to fidelity and capabilities of the HPA4.
Preamplifier Audio Measurements
Despite running an all-digital system I also have a Reel to Reel deck so a product like HPA4 fits very well for my needs. So let's start with a dashboard view, feeding the unit with balanced XLR input and measuring what comes out of pre-outs XLR:
Are your socks blown off as mine are? Second harmonic distortion is at an incredible -140 dB. That is insanely good. THD+N meter is running out of zeros to describe it as a percent! Going by SINAD version in dB, we have combination of noise and distortion that is at -120 dB. With our best case hearing dynamic range of 116 dB, we are in absolute transparency department. Don't even think blaming this preamplifier for anything bad you hear in your system. Blame is elsewhere.
Crosstalk performance is beyond imagination:
In 2 to 4 kHz where our hearing sensitivity is the best, crosstalk is around 140 dB!!! Even in the worst case situation of 20 kHz, what bleeds into the other channel is not only below threshold of hearing, but also below the measured THD+N. This means that the amount bleeding into the other channel cannot disturb it beyond its own excellent performance.
Perfection story translates into frequency response:
I might hook up the HPA4 to my boat sonar to look for underwater fish given this kind of ultrasonic response!
Setting one channel as reference, differential phase is as good as it gets just the same:
THD+N measured like I always do with high input impedance is superb:
Remember, this measurement includes noise and distortion well above audio band to 20 kHz (hence the reason it is lower than the SINAD). This shows how clean the HPA4 is above the audio band.
Upping the load to a punishing 600 ohm does stress a bit, bringing up 20 kHz THD+N to -95 dB. As a comparison to a budget product, I have included the Schiit Saga in its active mode in blue.
For intermodulation distortion, I don't have a reference so I decided to measure the Audio Precision APx555 analyzer as the reference (in blue):
Note that the APx555 is in loopback mode with an internal connection from input to output. The HPA4 involves multiple cables and active circuits in the HPA4. So this has some penalty in noise when the input signal/voltage is lower. But once we get to 1 volt, the IMD measurement floors at the same level as my analyzer. We cannot measure how good it is!
Headphone Audio Measurements
Let's start with our dashboard view using the 1/4 inch headphone jack (which is used in all but one of the tests):
Distortion is ever so slightly above pre-amp mode but it is still exceptionally low at -138 to -140 dB. This is by far class leading as the THX AAA 789 could "only" muster -130 dB:
SINAD is 119 dB which is much higher than THD so we know we are noise limited.
We can dig into contributions of the noise with the unit terminated with inputs but no signal and volume set to minimum:
On the left is the APx555 analyzer in loopback mode, measuring its own noise level. We see a baseline of 0.9 microvolts (millionth of a volt). Hooking up the Benchkar HPA4 ups that to 2 microvolts. With Massdrop THX, we have three different gain settings. Using the lowest one that is actually an attenuator, we get lower noise level of 1.6 microvolts. The attenuator in THX 789 is helping lower the noise in this extreme case.
The above then predicts that 50 millivolt SNR won't be as good as THX:
The full dynamic range though t 133 dB is absolutely superb and bests the THX AAA 789's measurements by whopping 10 dB.
For the following tests I re-measured the THX AAA 789 using my latest loads and settings to make sure everything is compared properly. First up is THD+N/SINAD versus power level at 300 ohm, emphasizing voltage delivery for high impedance headphones:
Most important thing is the power level. The HPA4 sails past the THX AAA 789, providing almost half a watt of output or three times more. The unit has extensive protection circuit and shuts down the output once clipping gets near hence the reason you don't see the hockey stick as you see with THX amp.
At lower levels, we benefit from lower distortion and noise. The gap is significant if you used THX in high gain mode. But if you take advantage of its three gain settings, you can close the gap.
Switching to my new low distortion 33 ohm load we get:
So similar story now with current delivery front and center. We have ample amount of power at 3.5 watts.
Switching the load to XLR connection allows the THX AAA 789 to pull ahead in power:
For that extra power though, distortion is increased with THX AAA 789 while it naturally remains low with the HPA4. Note that this load needs to be revised too as it doesn't use high precision resistors and hence the rising distortion in red channel.
Intermodulation distortion relative to output power delivers this:
More power with less IMD distortion. What else could you ask for?
The stepped attenuator in Benchmark HPA4 naturally leaves the THX AAA 789 in the dust with perfect channel balance:
Across incredible range of 110 dB of attenuation, you have perfect channel matching. There is a balance control anyway if you did need to adjust anything.
Frequency response is naturally flat up to the 200 kHz limit of the generator in APx555:
Headphone output impedance was comfortably low at 0.9 ohm:
Headphone Listening Test
As usual, I started my testing with the Sennheiser HD-650. You guys know that I like to test headphone amplifiers at high volumes. Even if it is for a second or two, I try to up the volume to max to see whether the headphone or the amplifier give out first. Not here. I was too scared to get even close to -10 dB let alone +15 dB range the HPA4 has! The sound had incredible authority with superb bass and fantastic detail. It is a revelation to have this much clean power on hand to drive a headphone.
Conclusions
Going into this test, based on measuring previous Benchmark products, I expected excellent performance. The benchmark HPA4 shattered even that high bar. It brings such a high performance to the table that leaves no doubt whatsoever about its transparency. No stone has been left unturned to optimize audio performance of this device. Even obscure measurements like ultrasonic interchannel phase shift is nailed.
And then we have tons of power without messing with balanced cables and such. Plug in any headphone and you are in heaven.
The preamp performance is even higher still, breaking and shattering any expectation one would have in a lab instrument let alone a consumer hi-fi.
Yes, the baseline noise is bit above the best we can get. If you have ultra sensitive IEMs and blast the volume when nothing is playing, you may hear some hiss. That is the only negative thing I can think of to mention. The solution for this would require and analog attenuator which is not in the cards in the architecture of the HPA4.
It goes without saying that the Benchmark HPA4 garners my strongest recommendation for a headphone amplifier and preamplifier. It doesn't get better than this folks.
------------
As always, questions, comments, recommendations, etc. are welcome.
One of my pink panthers heard about this thing called a "Keto Diet." He wants me to buy a bunch of books on it to learn more. I keep telling him that his diet or raw meats is already "keto" but he doesn't believe me. Being afraid of a hungry and angry panther, I thought I ask you all for money to buy him the darn books. So please donate using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
The HPA4 is a departure from other Benchmark products with the addition of a color touchscreen:
The screen looks better in person (and in my picture) than what I have seen elsewhere. Still, it is a juxtaposition of industrial/professional look.
The display does not do a lot but hugely simplifies the configuration of the HPA4.
Volume control is a stepped attenuator implemented with a ton of relays with gold plated contacts that is stated to keep working long after I am dead and buried.
There are two headphone outputs but the XLR connection is the same as 1/4 inch with the exception of independent return wire for each channel instead of shared. Benchmark also states that the XLR connection makes a more secure, lower impedance connection. I tested both.
In addition to being a headphone amplifier, the HPA4 is also a pre-amplifier:
You have multiple inputs in different flavors. And of course RCA and XLR outputs. I have not looked but I assume the mono out can be used to drive a subwoofer.
Nice set of trigger ins and outs are provided so that you can power a bunch of gear at once from one source unit.
In operation, the HPA4 runs quite warm for a headphone amplifier/pre-amp. Nothing to be concerned about with respect to reliability but you may not want to leave it on all the time. Performance is instantly up to spec so there is no need for warm up anyway.
A switching power supply assures universal compatibility with whatever mains voltage is in your country.
Let's get into testing but a few notes. First, as with Benchmark AHB2, I was sent to the drawing board to improve the quality of my test fixture. The dummy load I had for testing at 33 Ohm has too many connectors/switches in the path so I had to build an optimized, dedicated one to eliminate sources of distortion there. I have also revised some of the test, putting emphasis on dB metrics for distortion versus percentage. The testing is als more extreme than some of my other reviews due to fidelity and capabilities of the HPA4.
Preamplifier Audio Measurements
Despite running an all-digital system I also have a Reel to Reel deck so a product like HPA4 fits very well for my needs. So let's start with a dashboard view, feeding the unit with balanced XLR input and measuring what comes out of pre-outs XLR:
Are your socks blown off as mine are? Second harmonic distortion is at an incredible -140 dB. That is insanely good. THD+N meter is running out of zeros to describe it as a percent! Going by SINAD version in dB, we have combination of noise and distortion that is at -120 dB. With our best case hearing dynamic range of 116 dB, we are in absolute transparency department. Don't even think blaming this preamplifier for anything bad you hear in your system. Blame is elsewhere.
Crosstalk performance is beyond imagination:
In 2 to 4 kHz where our hearing sensitivity is the best, crosstalk is around 140 dB!!! Even in the worst case situation of 20 kHz, what bleeds into the other channel is not only below threshold of hearing, but also below the measured THD+N. This means that the amount bleeding into the other channel cannot disturb it beyond its own excellent performance.
Perfection story translates into frequency response:
I might hook up the HPA4 to my boat sonar to look for underwater fish given this kind of ultrasonic response!
Setting one channel as reference, differential phase is as good as it gets just the same:
THD+N measured like I always do with high input impedance is superb:
Remember, this measurement includes noise and distortion well above audio band to 20 kHz (hence the reason it is lower than the SINAD). This shows how clean the HPA4 is above the audio band.
Upping the load to a punishing 600 ohm does stress a bit, bringing up 20 kHz THD+N to -95 dB. As a comparison to a budget product, I have included the Schiit Saga in its active mode in blue.
For intermodulation distortion, I don't have a reference so I decided to measure the Audio Precision APx555 analyzer as the reference (in blue):
Note that the APx555 is in loopback mode with an internal connection from input to output. The HPA4 involves multiple cables and active circuits in the HPA4. So this has some penalty in noise when the input signal/voltage is lower. But once we get to 1 volt, the IMD measurement floors at the same level as my analyzer. We cannot measure how good it is!
Headphone Audio Measurements
Let's start with our dashboard view using the 1/4 inch headphone jack (which is used in all but one of the tests):
Distortion is ever so slightly above pre-amp mode but it is still exceptionally low at -138 to -140 dB. This is by far class leading as the THX AAA 789 could "only" muster -130 dB:
SINAD is 119 dB which is much higher than THD so we know we are noise limited.
We can dig into contributions of the noise with the unit terminated with inputs but no signal and volume set to minimum:
On the left is the APx555 analyzer in loopback mode, measuring its own noise level. We see a baseline of 0.9 microvolts (millionth of a volt). Hooking up the Benchkar HPA4 ups that to 2 microvolts. With Massdrop THX, we have three different gain settings. Using the lowest one that is actually an attenuator, we get lower noise level of 1.6 microvolts. The attenuator in THX 789 is helping lower the noise in this extreme case.
The above then predicts that 50 millivolt SNR won't be as good as THX:
The full dynamic range though t 133 dB is absolutely superb and bests the THX AAA 789's measurements by whopping 10 dB.
For the following tests I re-measured the THX AAA 789 using my latest loads and settings to make sure everything is compared properly. First up is THD+N/SINAD versus power level at 300 ohm, emphasizing voltage delivery for high impedance headphones:
Most important thing is the power level. The HPA4 sails past the THX AAA 789, providing almost half a watt of output or three times more. The unit has extensive protection circuit and shuts down the output once clipping gets near hence the reason you don't see the hockey stick as you see with THX amp.
At lower levels, we benefit from lower distortion and noise. The gap is significant if you used THX in high gain mode. But if you take advantage of its three gain settings, you can close the gap.
Switching to my new low distortion 33 ohm load we get:
So similar story now with current delivery front and center. We have ample amount of power at 3.5 watts.
Switching the load to XLR connection allows the THX AAA 789 to pull ahead in power:
For that extra power though, distortion is increased with THX AAA 789 while it naturally remains low with the HPA4. Note that this load needs to be revised too as it doesn't use high precision resistors and hence the rising distortion in red channel.
Intermodulation distortion relative to output power delivers this:
More power with less IMD distortion. What else could you ask for?
The stepped attenuator in Benchmark HPA4 naturally leaves the THX AAA 789 in the dust with perfect channel balance:
Across incredible range of 110 dB of attenuation, you have perfect channel matching. There is a balance control anyway if you did need to adjust anything.
Frequency response is naturally flat up to the 200 kHz limit of the generator in APx555:
Headphone output impedance was comfortably low at 0.9 ohm:
Headphone Listening Test
As usual, I started my testing with the Sennheiser HD-650. You guys know that I like to test headphone amplifiers at high volumes. Even if it is for a second or two, I try to up the volume to max to see whether the headphone or the amplifier give out first. Not here. I was too scared to get even close to -10 dB let alone +15 dB range the HPA4 has! The sound had incredible authority with superb bass and fantastic detail. It is a revelation to have this much clean power on hand to drive a headphone.
Conclusions
Going into this test, based on measuring previous Benchmark products, I expected excellent performance. The benchmark HPA4 shattered even that high bar. It brings such a high performance to the table that leaves no doubt whatsoever about its transparency. No stone has been left unturned to optimize audio performance of this device. Even obscure measurements like ultrasonic interchannel phase shift is nailed.
And then we have tons of power without messing with balanced cables and such. Plug in any headphone and you are in heaven.
The preamp performance is even higher still, breaking and shattering any expectation one would have in a lab instrument let alone a consumer hi-fi.
Yes, the baseline noise is bit above the best we can get. If you have ultra sensitive IEMs and blast the volume when nothing is playing, you may hear some hiss. That is the only negative thing I can think of to mention. The solution for this would require and analog attenuator which is not in the cards in the architecture of the HPA4.
It goes without saying that the Benchmark HPA4 garners my strongest recommendation for a headphone amplifier and preamplifier. It doesn't get better than this folks.
------------
As always, questions, comments, recommendations, etc. are welcome.
One of my pink panthers heard about this thing called a "Keto Diet." He wants me to buy a bunch of books on it to learn more. I keep telling him that his diet or raw meats is already "keto" but he doesn't believe me. Being afraid of a hungry and angry panther, I thought I ask you all for money to buy him the darn books. So please donate using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
