Reviews by Gippy


500+ Head-Fier
The best glass optical cable on the market.
Pros: Everything including the glass fiber is produced in-house
No-nonsense plugs make a solid connection
Cons: Snake oil if you think a $10 cable is all you need
It's still an optical cable, so care must be taken when handling it
For those who think a $10 plastic optical cable is good enough, cool. I'm making absolutely no claims that this cable improves sound quality. However, as this is an optical cable, it confers specific advantages over USB cables which I'll detail further below.

So why pay USD $135 for a 6-foot optical cable? What you're paying for is supporting a small, specialized company that produces a quality, handmade product. What makes Lifatec unique is that they specialize in supplying medical-grade optic fiber lines for endoscopes. Their optical cable business is a bonus, but it means that they know exactly what they're doing instead of outsourcing the cables to an OEM, then slapping their brand label on them. This means they can also make custom length cables. I specifically wanted a cable that was 4 feet long and they were happy to make it.

When compared against other boutique cables, Lifatec's cable doesn't look fancy at all, but it actually out-specs them at a fraction of the price. Lifatec's cable is made of 470 glass fibers. DH Labs charges $265 for their 6.5-foot Glass Master optical cable, but it has 300 glass fibers. AudioQuest charges $300 for their 5-foot Vodka optical cable, but it's just a pretty rebranded Asahi Kasei 217 plastic fiber cable.

The plug is well-built and snaps into the optical jack more solidly than any of my cheap optical cables. The plug is made of plastic, but that's because a gold-plated plug on a non-electrical cable is pointless. The cable itself is actually lighter and more flexible than my cheap optical cables because the sheathing isn't some thick PVC. I haven't attempted a stress test, and won't do so, but I wouldn't use this cable (or any optical cable) in an area where it may be stepped on. There's no protection other than the thin sheathing, as optical cables are immune to electromagnetic interference.

The Lifatec cable will carry a 24-bit 192khz PCM signal with no issues. USB can carry a 32-bit 768khz PCM signal, as well as more exotic signals such as DSD and MQA. But what looks great on paper doesn't necessarily translate to real results. I had a $700 Topping D90 DAC (review here) that had several issues with handling USB audio. First, there was an electrical ground loop, and I could clearly hear the noise even when there was no audio being played. Second, USB audio tends to turn on and off, leading to popping artifacts. Switching from USB to optical solved all of these issues, and so for me, optical is the DAC input of choice. A $10 plastic optical cable may have done the same. However, after reading reports that some optical cables couldn't handle 24-bit 192khz PCM signals, I decided to purchase the Lifatec cable for peace of mind.

Some have claimed that this is the best optical cable at any price, and that they've heard sonic improvements over other optical cables. I make no such claims, and neither does Lifatec, which is a breath of fresh air compared to other companies. It gets the job done, and even though it out-specs the competition, it's not monstrously expensive compared to boutique cables that cost over $1000. By purchasing this, you're getting a quality cable, handmade by a small company in the USA, instead of a mass-produced product made in China and sold at large margins. And that's good enough for me.
Last edited:
Thanks for reviewing a TOSLINK patch with S/T quality. Lucent Technologies specified connectors incompatible with TOSLINK for the industrial S/T standard, which uses optical glass fiber rather than the optical plastics of most TOSLINK connections.
fortunate son
fortunate son
The Lifatec is a well made tight fitting cable. It is the most analytical optical cable I have used.
I have the Lifatec optic cable. It sounds notable better than the one I had in its place.


500+ Head-Fier
Pros: Phenomenal performance at lowest (0dB) gain, especially with XLR4 out
Excellent solid state mode
Best implementation of iFi's XBass and 3D modes
Compact design compared to other full-featured desktop amps
Cons: 3.5mm outs have automatic IEMatch and don't provide the best sound quality
9dB and 18dB gain modes have significantly more noise
Tube modes are gimmicky
Runs hot
iFi's Micro iDSD Black Label (shortened to the MiDSD BL for the rest of the review) is perhaps iFi's most popular product. It's a portable combo DAC/amp unit that packs a comprehensive feature set, along with a ridiculous amount of power, at a price ($600 USD) that doesn't break the bank. What's very notable is its analog tweak features, XBass and 3D+. I own the MiDSD BL, and used it as my everyday desktop DAC/amp, but there were a few things I found frustrating about it. It didn't provide enough power on eco (lowest) gain for my headphones. If I raised it to normal gain, there would be hissing. And using the IEMatch feature to eliminate the hissing resulted in a less airy sound.

The only upgraded model from iFi that has the analog tweak features, and has more power, is the Pro iCAN, but it's triple the price. As I loved the 3D+ tweak on the MiDSD BL, upgrading to the Pro iCAN for me was a no-brainer. Those who aren't as enthralled with the analog tweak features may wonder if the Pro iCAN is worth it for them.

The unit itself is a very nice size, as it's reasonably compact compared to the rackmount widths of beefier amps. It's as wide as the 8.5" length of a letter-size paper sheet. The front panel looks very symmetrical in its design. The iFi logo on the top left will change color depending on its current status. There's no way to disable this LED, but it's small enough to not be too obtrusive.

This is a Class A amp, which means it runs hot. It won't burn your hand when you immediately touch it, but it's very warm, and you wouldn't want to put your palm on the chassis for more than half a minute. If you're coming from the MiDSD BL like I did, this might be a bit of a shock to you.

For inputs, there's a dual XLR3 in for the left and right channels, as well as 3 RCA ins. This is a bit of overkill, as for my purposes, I've never needed more than two inputs at the same time. But it's nice that the option is there. For this review, I mainly used the XLR3 input fed from a Topping D90 DAC, which at the time of this review was Topping's flagship DAC ($700 USD) using the AKM AK4499 chip. Other reviews have used iFi's own Pro iDSD ($2500 USD). I think $2500 for a DAC is a colossal waste of money, especially when the DAC has superfluous parts such as its own headphone amp and tubes. The Pro iDSD also uses the same DAC chip (TI PCM1793) as the MiDSD BL, and doesn't have the analog tweaks that the MiDSD BL has, so I couldn't justify it at all. Headphone amps affect the sound far more than the DAC, so by going with the Topping D90, I saved a lot of money.

As for outputs, there's an XLR4 headphone out, dual XLR3 headphone outs that double as 6.35mm outs, a single-ended 3.5mm out, and a balanced 3.5mm out. The Pro iCAN's vintage is 2016, so no balanced 2.5mm or 4.4mm outs. But you could plug up to a whopping 5 headphones into this unit. At the back is a dual XLR3 out and 1 RCA out, in which the iCAN may act as a preamp. Plugging headphones in at the front doesn't disable the rear outputs.

For this review, I mainly used the XLR4 headphone out with the Grado GS3000e, which has 40 ohm impedance and 97dB/mW sensitivity. (Grado's marketing states 32 ohm and 99.8dB/mW but measurements at RAA state otherwise.) At the time of this review, it was Grado's flagship wooden headphone ($1800 USD). Using XLR4 out over 6.35mm out provides a considerable advantage in that it doubles the power output without affecting the noise floor. This is very important as we'll see later.

The 3.5mm outs are disappointing. They automatically have iFi's IEMatch feature applied to them. Unlike the MiDSD BL, in which you had to determine the IEMatch strength yourself, the Pro iCAN detects the headphone's impedance and applies the appropriate IEMatch level. I don't like IEMatch and think it's a flawed solution to a problem that shouldn't exist in the first place. Users have reported that while it does remove hissing, it also reduces the overall volume, decreases the dynamic range, and makes things sound less airy. I agreed with these findings, and always had IEMatch off on the MiDSD BL, even though without it, I could hear a very faint amount of hissing. Instead of using IEMatch, it's better to just design an amp for ultra-sensitive headphones and IEMs that works. For moderately sensitive 3.5mm headphones, a 6.35mm adapter should be used in order to avoid IEMatch.

GAIN (0dB, 9dB, 18dB)
0dB gain mode was fantastic. Using the XLR4 out, 0dB gain mode outputs as much power as the MiDSD BL on normal gain, but without all the noise and distortion! It was plenty loud with my GS3000e at just 10 o'clock on the dial. ASR measurements state that 0dB gain mode is very clean, and excels with low to moderate impedance headphones. Most importantly, my GS3000e was finally dead silent when there's no output, even with the volume knob turned all the way to the maximum 5 o'clock position.

9dB gain mode introduced a very small amount of floor hiss on my GS3000e, about as much as I got on the MiDSD BL on normal power. Some users have reported that 9dB gain mode sounds more neutral than 0dB gain mode, but with no way to really do a fast comparison, I couldn't determine whether what I heard was placebo. Going to my Grado SR225i, 9dB was ideal for it, as it was a less efficient headphone (32 ohm, 93.5dB/mW) than the GS3000e (40 ohm, 97dB/mW), and was using the 6.35mm out, which provides half the power compared to XLR4.

18dB gain mode wasn't tested. I didn't have any headphones that required that much power. Attempting to use it on the GS3000e resulted in very audible hissing. ASR measurements state that distortion and noise are considerably higher, so if you have headphones that actually need this, you really should look at another amp.

Unique to the Pro iCAN is its three different amplification modes. The solid state and tube modes run on separate circuits, so this isn't some halfhearted hybrid design. It's literally two amplifiers in one. Note that the temperature of the unit stays hot regardless of whether you use solid state or tube mode.

Solid state mode is what I use most of the time. It provides the best clarity and dynamics, and I use a software EQ to remove the harshness from my GS3000e anyway.

The tube modes are more of a gimmick, and I wouldn't really miss them if they were gone from the next iteration of this unit. Tubes have a limited lifespan, and the tubes in the Pro iCAN aren't easily replaceable. Some day, I'd like to resell this unit, so I'd rather minimize my tube usage if possible. Tube mode doesn't sound too much different than solid state mode on most recordings. Maybe it removes a bit of harshness, but I couldn't barely tell a difference, if at all. I wouldn't recommend using tube mode over solid state mode.

Tube+ mode, on the other hand, noticeably changes the sound signature, as the increased harmonic distortion really takes the edge out of any harsh treble. This means there's some detail and dynamics loss, but if you're not feeling analytical and want a more relaxing sound, this mode will work for you.

Here we go. This is what I consider to be the primary selling point of the Pro iCAN: getting the "best" version of iFi's analog tweaks.

XBass comes in 3 flavors: 10hz/20hz/40hz. Note that these are poorly named. "10hz" starts with a +9.5dB boost @ 10hz and slopes down towards 0 @ 200hz. "20hz" has a +10.5dB boost @ 20hz and slopes down towards 0 @ 500hz. "40hz" has a +10.5dB boost @ 40hz and slopes down towards 0 @ 1khz. The "20hz" boost is about the same as the XBass+ boost on the MiDSD BL, though I typically never used it when I had the MiDSD BL because I thought there was too much lower-mids bleed. "40hz" was even worse, and I switched out of it after about 30 seconds of listening. However, there was a plus: the "10hz" boost is much more subtle, and I greatly enjoyed this on my GS3000e.

The 3D mode is interesting. For those who haven't used iFi's 3D, I need to debunk a misconception. It is not just crossfeed. Simple crossfeed reduces the soundstage width. What iFi's 3D does is mess around with the phase at various frequencies, sending some sound from one channel to the other, but this is done in such a way that it increases the soundstage width, not reduce it. It also preserves the bass, unlike a simple "out of phase" filter. Therefore, iFi's 3D tweak is superior than simple crossfeed, and it's something that I have turned on all the time for increased immersion. Different headphones will vary in the amount of benefit that iFi's 3D provides, but it really shines on my GS3000e.

Note that the Pro iCAN was launched before the MiDSD BL, about 8 months earlier, in the first half of 2016, while the MiDSD BL was launched in the beginning of 2017. The MiDSD has an updated 3D mode that's called "3D+". 3D on the Pro iCAN comes in 3 flavors: 30°, 60°, and 90°. 30° and 60° provide a lesser effect than the MiDSD BL, and so they're not worth using if you've been used to the MiDSD BL.

90° provides about as much width as 3D+. However, 3D+ introduced some extra airiness and brightness by increasing the amount of perceived treble. This is all done with phase voodoo, as measurements have shown that the frequency response doesn't change. 90° on the Pro iCAN doesn't do this. Some may feel the MiDSD BL might exhibit a "stronger" effect due to the treble lift.
When I used the MiDSD BL, I had to reduce the treble via software EQ to offset this. With the Pro iCAN, I adjusted my software EQ so that it was back to having the appropriate amount of brightness. Overall, I'd give a slight edge to the Pro iCAN because it keeps the heft of the mids, but I could totally see someone preferring the 3D+ of the MiDSD BL.

So to compare, for the MiDSD BL, I had XBass+ off and 3D+ on. On the Pro iCAN, I have XBass at "10hz" and 3D at 90°. I feel this is a slight improvement over the MiDSD BL's tweaked sound.

The Pro iCAN is a considerable upgrade, especially if you already enjoy the MiDSD BL and frequently use the XBass and 3D analog tweaks. It measures significantly better, provides more clean power, and is definitely worth triple the asking price ($1800 USD vs. $600 USD) if you can take advantage of its strengths. However, it's not a universal recommendation, as I feel it's best used specifically with balanced headphones with moderate sensitivity and impedance.

Those with sensitive IEMs should look elsewhere, as the forced IEMatch is a letdown. The marketing states that there's a monster amount of power on tap, and there is, but it's not clean. Those with low-sensitivity headphones have cleaner amp options at this price range, as moving to 9dB or 18dB gain results in worse performance. And finally, if you're a purist and think the XBass and 3D knobs are sacrilege, and won't take advantage of the preamp function, then don't get this amp. There are cheaper amps that will provide what you need.
Last edited:
Update: I'm now using 9dB gain mode exclusively. There is some faint hissing if I turn up the volume to very high levels, but I feel 9dB gives the sound some more dynamic range, a bit of extra energy, over 0dB gain mode. The hiss isn't audible at my normal listening levels. Here's hoping that the next version of this is improved!


500+ Head-Fier
Expensive, but at least it measures well.
Pros: Lower capacitance/inductance/resistance than most other cables
Reliable Neutrik XLR connectors
Cons: Price (not recommended if your total system value is <$1500)
Don't expect a sonic improvement with this
This is an XLR cable that sells for CAD $180 (USD $145) for a 3ft pair. However, you can save CAD $26 (USD $21) if you shorten the length to 2ft, which is what you should choose if this is being used as an interconnect from a DAC to an amp, and the two components are stacked or next to each other. 2ft is the length I bought.

I'm firmly in the camp that you shouldn't spend a fortune on cables, and that cables should cost less than 10% of your system. But as the combined retail value of my DAC, amp, and headphones is over $5000, I figured I could try a "better" XLR cable than the $15 Monoprice pair I was using. Let's get this out of the way now: I couldn't hear a difference between these cables and the Monoprice cables. Review over, right? Well, not exactly.

Take Five Audio posts measurements of all of their cables they sell. The reason to consider the NEMOI-3220 is that it measures better than the industry-standard Mogami cables when it comes to capacitance, inductance, and resistance (lower is better). This is thanks to the NEMOI-3220's unique rectangular copper conductor geometry. Take Five Audio sells more expensive cables than these, but the NEMOI-3220 actually measures better than them. It probably also measures better than other boutique cables that cost way more. Does that translate into better sound? At such short lengths, absolutely not. But compared against the Monoprice cable, the NEMOI-3220 is much thicker due to its additional shielding, and the Neutrik XLR connectors are more durable than the suspicious low-grade connectors on the Monoprice. The thickness does make it less flexible than other cables, but it will still bend enough to act as a suitable DAC to amp interconnect.

Minor quibbles: If you buy a pair, the cables will be color-coded, so if you don't want this, you must make that a special request. Upon visual inspection, one of the XLR cables had a small dimple on the exterior covering, but this hasn't affected performance. Still, for this much money, I expected a visually flawless cable, and so I'll deduct half a star for that.

Paying this much for a short XLR interconnect might seem a bit excessive, but at least the measurements back it up, and I wanted to support a local Canadian business during the difficult time of the COVID-19 pandemic. At the very least, this seems like a logical stopping point before the deep-end insanity of cables that cost thousands of dollars. To put it into perspective, a pair of Take Five Audio NEMOI-3220 XLR cables is still cheaper than some of the competition's cheapest XLR cables: AudioQuest's Red River ($175), Nordost's White Lightning ($200), or Siltech's Explorer 180ix ($725).
Last edited:
  • Like
Reactions: WK446


500+ Head-Fier
The Grado for those who don't like Grado.
Pros: Stock F pads present a heavy mid-treble dip (which may attract Grado haters!)
Aggressively priced lower than their least expensive standard wooden model, the RS2e
Stronger mid-bass impact than many other Grado models
Cons: Stock F pads present a heavy mid-treble dip (which may turn off Grado lovers!)
Typical love-it-or-hate-it Grado build quality and quirks such as the non-detachable cable

  • Weighs 190g without cable. As a comparison... GH4: 180g; GS2000e: 270g; SR325x: 280g; GS3000e: 330g; PS1000e: 530g
  • Wooden cups seem smaller than the RS2e. 80% of it is hemp, while the last 20% facing the ear is maple.
  • The engraving on the medallions was sharp.
  • 8-conductor cable had a few creases out of the box.
  • Headband is good. Other brands may look nicer, but they're made of plastic or non-malleable metal. Grado headbands may be finely adjusted through bending, which is great because heads aren't entirely symmetrical. When someone complains about this, they just don't get it.
  • Visible glue marks on the plastic edge of the driver, where it's attached to the wooden cup. This is different than other wooden models, which have a small gap between the edge and the wood, and no visible glue.

At the shop where I demoed it, I thought the Hemp sounded absolutely awful, especially when compared against Grado's other models. I almost didn't buy it at all. However, one of the best features of Grado headphones is that there are a variety of pads, made by both Grado and aftermarket brands. I bought the Hemp hoping that pad rolling would improve the sound. The rest of the review details my pad rolling findings.


With 7 different pads at my disposal, all with their own characteristics, it's time to listen away! I'll list 3 criteria: bass, treble, and full, on a scale of -5 to +5. 0 is ideal for my taste, so the closer to 0, the better, but note that 0 isn't necessarily perfect. Shifting from 0 leans toward too little or too much. Full is my subjective opinion on how thin or how veiled the overall sound is. All tests were done with no EQ and no ASP from my Pro iCAN. (We'll cover EQ results later.)

In the picture, the left column shows the Geekria L, "LL", and G, while the right column shows the Grado F, S, L, and G.

Grado F (stock)
Bass: +3
Treble: -4
Full: +5

Yuck. This is so veiled! It amazes me that people prefer this sound signature. I'm reminded of the Audio-Technica M50x, but I think even the M50x was brighter than this. This is the sound signature for people who typically hate Grado. To me, the SR60e sounds better than this.

Grado S
Bass: +2
Treble: -2
Full: +3

Even though these pads cover the driver, it's somehow less veiled than the Grado F. There's still a bit of a significant veil, though. However, to me, the Grado S had a more agreeable sound. Treble is still reduced, though it doesn't feel like there's a canyon of missing treble.

Grado L
Bass: -1
Treble: -1
Full: +1

Yikes, the foam is so coarse and rough. But with these, the Hemp actually sounds like a Grado. The drivers are still tuned a bit dark, though, and it's no longer a bassy headphone. It sounds like a less resolving RS2e or GH4. Not bad, but they're about to be beaten hard by...

Geekria L
Bass: 0
Treble: +1
Full: 0

WOW! This was an absolute shocker. Not only is the Geekria L cheaper than the Grado L, it sounds better, and has softer foam! This is the only Geekria pad with two foam densities. But it's not an exact knockoff: the diameter of the Grado L pad is 75mm, while the diameter of the Geekria L is 80mm. It somehow makes a huge difference. This is the best pad for the Hemp unless you are super treble sensitive, and if you own one, you should go on Amazon now and order a pair. Now I'm curious as to how the Geekria L performs on other Grados!

Geekria "LL"
Bass: +2
Treble: -1
Full: +2

These are interesting in that they have no official equivalent. They have almost the same thickness as the L pads, but are one density, and the tapering is steep, so everything sounds closer. The bass is quite prominent and leaks into the mids, but because the driver is exposed and a little further away, it's not as veiled as the Grado F or S. I would say that if you actually enjoy the Grado F sound signature, try these, as I think the slight bass loss is worth the extra clarity. And it's more comfortable, too!

Grado G
Bass: -2
Treble: +3
Full: -1

These are significantly more expensive than the other pads. Well, they are the most comfortable pads of the bunch. Putting these on gives a vast soundstage, but introduces the much-maligned 5k-6k treble spike. This is the EQers pad, as using this pad to EQ will allow a wider range of tweaking than with the other pads.

Geekria G
Bass: +3
Treble: -2
Full: +4

Unlike the Geekria L, which is a superior version of the Grado L, the Geekria G is rather disappointing. It's only one foam density, and the tapering isn't as steep. The Grado G has 0.3mm of foam from the driver until it starts tapering, while the Geekria G starts at 0.8mm, forming a tunnel. This causes more bass to be forced through, causing this to be just as bassy as the Grado F. Without the soundstage of the Grado G, this feels more like a super-sized Geekria LL. Note that when I briefly owned the Beautiful Audio G pads, they were constructed in a similar manner with a prominent "tunnel", and they sounded similar to this.

Ranking: Geekria L > Grado L > Geekria "LL" > Grado G > Grado S > Geekria G > Grado F

VS GH4 (MSRP: USD 550)

EDIT (Revised comparison 2022/10/28):
I now own the GH4, and bought it over the RS2x/2e/1x/1e. The GH4 with Grado L pads clearly beat the Hemp in terms of clarity. The GH4 is the ultimate refinement of the traditional Grado house sound, but won't pierce your ears with treble like the non-EQ Statement models, and I bought the GH4 over every other x-series model, including the flagship GS3000x. The GH4 is now discontinued and difficult to find, but I'd take it over the Hemp every time, and it doesn't cost that much more. You can always muffle the brighter GH4 sound with different pads, but you can't make the Hemp as clear as the GH4 without EQ. EDIT October 2023: I've now sold the Hemp and kept the GH4.

VS GS3000e (MSRP: USD 1795)

This doesn't seem like a fair comparison given that the GS3000e is more than quadruple the cost, but let's do it anyway. I've always maintained that the GS3000e has Grado's most technical driver that's ripe for EQ, as it has actual sub-bass. As using EQ is best with the Grado G, I didn't attempt EQ with any other pad, as other pads trade some comfort for a non-EQ sound tweak. After a few days of tweaking, I could EQ the Hemp to sound shockingly close to the GS3000e. The GS3000e still had significantly better bass extension and visceral impact, and flexes its resolving capability with many instruments at once, but if I had to put a percentage on it, I think my EQ'd Hemp is about 70% as good as my EQ'd GS3000e, and is better than the stock GS3000e. For less than one-quarter the price, that's incredible.

  • BONUS - GS3000e with Geekria L, no EQ: This is really nice! It's a better match for the GS3000e than the Grado G if you don't intend to EQ, and is noticeably clearer than the Hemp with Geekria L. This all but confirms that the Grado G adds the 5k-6k treble spike to every model, no matter what the tuning is, so it's inherently a flawed pad that must be corrected with EQ. Whoever designed the Geekria L should get a raise, because now I suspect it's better than the Grado L on every single model.

The GS3000e remains my daily driver. But I now understand the hype for the Hemp and why there was so much praise for it. It was squarely targeted at those who wouldn't have otherwise bought Grado. At stock, I think the Hemp sounds terrible, which is a dissenting opinion compared to others who have reviewed it. But with a bit of pad rolling, you can still transform this to sound Grado-ish, but without the aggressive treble. Hopefully this gives you a bit of insight, as most reviewers don't entertain the idea of pad rolling at all. What would've been a scathing review turned out much differently.

Good value, but not a giant killer. Experiment with pads to suit your personal taste.
Last edited:
Has anyone taken the puck off the hemp's grille? I am looking for photos of it.
A bit of a late comment, but your review is great for the Grado Hemp! Currently the Hemps are my daily driver. I got the Geekria L pad and it's got a sound that reminds me of my HD600. Great with mids and crispy treble, but never too much. The Geekria L pad is like a combination of the F and G pad imo. It's comfortable like a G pad, but still lets your ear be close to the driver, and there's nothing in front of the driver like an S pad or Yaxi pad or the yellow senny 414 pads. I listen to a lot of rock, metal, jazz, etc so the Grado Hemp has been a lot of fun to use as a daily driver. I didn't expect to love it as much as I did. But man, they've got some personality and character. Just a quirky little guy
Is the Hemp without the nex 'X' cable but with a white stitched headband also a Hemp V2?


500+ Head-Fier
For Topping D90 in DACs
Pros: Optical input works
Cons: Catastrophically broken USB implementation
Cannot turn off display
The Topping D90 was the first commercial DAC using AKM's latest and greatest DAC chip, the AK4499. On paper, this DAC chip boasts a THD+N of -124dB, which is lower and better than AKM's previous flagship, the AK4497 at -116dB. Many people believe that DAC technology has been more or less figured out at this point, and that a $100-200 DAC is good enough for most purposes. Therefore, at $700, the Topping D90 is held to a higher standard and must perform flawlessly for it to be a good buy.

Upon release, certain sites measured the D90 and gave it the highest praise. This is exactly why I bought the Topping D90. I thought it would be the last DAC I would ever need to buy. Unfortunately, the Topping D90 stumbled right out of the gate for me.

The first input I tried was USB. Immediately, there were two issues:

  • USB noise interference
  • Popping sound a few seconds after audio playback stops
There was a faint, but very audible noise coming from my headphones when no music was playing. To make sure it was the D90 and not my amp or cables, I plugged in several other USB DACs into my amp, as well as trying multiple cables. (None of the cables cost more than $15.) No other DACs exhibited the noise. In addition, after a few seconds of no playback, I heard an audible pop. Apparently, in Windows 7, the driver goes to sleep after a few seconds. This is documented at:

This is ridiculous. I have plenty of older DACs that don't exhibit this behavior. It doesn't matter if you set Windows settings to always never have the USB device or port sleep. The driver is ordering the DAC to go to sleep, thus the popping.

Fortunately, my PC tower has optical out. The D90's optical input works perfectly, though you're limited to up to 24-bit/192khz instead of up to 32-bit/768khz. I was happy with using optical input exclusively.

For this review, which came 3 months after purchase, I wanted to see if the issues mentioned above were due to Windows 7's USB audio implementation. So I obtained a Windows 10 test laptop. The issues became worse when I plugged the D90 into the laptop. The audio I heard crackled heavily for 1 minute. And then suddenly, the laptop no longer recognized the D90. I tried plugging it back into my PC tower. Nothing. The D90's USB input had completely died. Fortunately, optical input still worked. Did the USB input break because I had the optical cable plugged in with a music signal at the same time? Who knows, but it still shouldn't have broken like that.

One additional point of contention is that the display can't be turned off while the unit is operating. You can change how bright it is, but you can't turn it off completely. This is a major distraction when watching movies in a darkened room. You'll need to cover up the display with some electrical tape. Easy fix, but on a $700 unit, this shouldn't even be an issue.


With the D90's USB input giving me issues from the very beginning, then suddenly failing on me when writing this review, I can say with certainty that you are playing a slot machine when buying this device. It does have a 1-year warranty, but the exorbitant shipping cost for warranty service from North America to China has discouraged me from sending it in. At least I can still use the unit's optical input, and it does work well enough there, so I'll give it an extra half star for that.

Not recommended at all.
Last edited:
I've used the D90 MQA for a month now with Windows 10 and have not encountered any of these USB issues, any background noise, or any popping sounds. Did you test with a different USB cable? Have you considered exchanging the unit in case yours is faulty? Have you downloaded the latest drivers and firmware from Topping's website?

Edit: The catastrophic failure you describe would seem to confirm the hypothesis that you just got unlucky with a faulty unit. Noise and popping sounds on the USB may have been symptoms of this too.
Hi headwig,

Please read the link in the review for more details as to how I tried unsuccessfully to correct the USB issues. I tried everything you suggested. I am not exchanging the unit, as mail processing is still significantly delayed during COVID-19, and express shipping the unit to China would cost over $100.

Correction: I stated in an earlier reply that the iFi MiDSD BL had special drivers that disabled sleep mode. This is incorrect; I downloaded special firmware that disabled sleep mode. To my knowledge, the D90 non-MQA does not have new firmware.
John Massaria
John Massaria
Drivers from Topping web site make a huge difference as well.. (
I know the USB on Topping doesn't use the power line in any USB wire- not sure if that the wire you are using is defective or unit- but having a broken unit doesn't mean ALL D90s deserve that rating- would have been nice if Topping sent a new D90 for review and forget the defective unit altogether- talk to your APOS/Topping dealer maybe- send Topping a sample link of your review- I would today! WHY not? -their service is first rate for things like this is my understanding.

My D90 is dead quiet on all USB wires I tried- I will say I prefer the all copper USB by Kimber for sound - I hope you get a new unit- you are missing out -