The form factor is approximately that of a Chord Mojo, the weight is nicely like half as much, the battery is easily better – honestly difficult to make worse then Mojo on that… – and there’s quite some additional perimetral features too, all for less than half of Mojo’s price. How will this compare sound-quality wise?
With this question I approached an ifi nano iDSD BL USB DAC-AMP, and this article is about how it went for me.
nano iDSD BL is a USB DAC-AMP, and more precisely USB is its sole input. No SPDIF digital input option, nor analogue inputs or any kind by that matter (aka: the device can’t be used as a mere amp).
Like all USB DAC-AMPs it can be plugged onto just about any USB-capable source such as a PC a Laptop a phone or a tablet, and it will be “seen” as an audio card. Full Windows 10 support requires a driver, which is freely downloadable from ifi’s site.
Nano iDSD BL’s USB 2.0 digital input connector is quite uncommon: it’s actually a USB A male installed in a recess of the chassis’ back panel. Due to that, the cable to connect to the nano iDSD BL needs to have a Female (!) USB A termination on the device end, in lieu of the usual Male one.
A few USB OTG cables with different terminations are bundled within the package. (Ah by the way… if you don’t know what OTG technically means, read here. You might discover why that otherwise good cable of yours refuses to work with your phone…)
- USB-A Female to USB3.0-A Male, 1m long, usually good to connect to a PC
- USB-A Female to USB-C Male, 15cm long, good to connect to a PC or a phone or other transports
- USB-A Female to USB-B Female, 15cm long
- USB-A Female to USB-B Female adapter (same as above, but no short cable in the middle)
No Apple cable is supplied – Lightning or 30-pin – so that’ll have to be purchased separately if needed. Neither is a micro-USB adapter (or cable) supplied, useful to hook onto non last-gen phones and tablets.
Boring cabling apart, something of paramount importance is behind the USB input port: nano iDSD BL incorporates most of the features offered by ifi iPurifier3, the company’s standalone USB “cleanser” which takes care of reclocking, rebalancing and regenerating the USB signal on the fly.
Especially when connecting to noisy hosts like a PC or a Laptop the sound quality improvement is totally apparent and, at least in my experience, the sole reason not to employ an iPurifier3, a galvanic separator, and a cleaner power supply on the input gates of a good DAC is their relatively high cost – which indeed in the case of a budget DAC can easily exceed its price… even a few times over.
According to nano iDSD BL literature and some answers I got from their tech people, the device includes the same REBalancer as the original iPurifier, together with some additional USB buffering circuitry (market-named “Zero Jitter”) which pursues the same ends of the REClocking part inside iPurifier3. Considering an iPurifier3 is separately sold for € 130+, including many of its functionalities inside the nano iDSD BL (€ 230-ish) is a great value proposition.
Ifi released 3 incremental versions of its iPurifier device. Here are their differences, for those who may wonder
- Original iPurifier: REBalance only, and passive circuitry
- iPurifier2: REBalance, REClock, REGen with an active circuitry
- iPurifier3: same as iPurifier2 but with upgraded components, yielding somewhat even better effect
The same technology is by the way included into some standalone active filtering ifi devices called nano / micro iUSB3.0 and nano iGalvanic. Nano iUSB3.0 is indeed a centerpiece of my desktop stack and I covered it in this other article which I invite you to read for a better description of “what it does”.
On the output side, nano iDSD BL has 3 ports: two 3.5mm females are on the front panel, one labelled “Direct” the other “iEMatch”, and the third one – on the back panel – is a 3.5mm Line Out.
The font panel “Direct” port unassumingly presents itself like just any single-ended analogue output port, and indeed it does work as such when you plug a 3.5mm single-ended terminated IEM or Headphone cable on it.
Direct output specs are quite interesting:
- Output impedance is nicely lower than 1 Ω
- Supported load impedances range from 15 Ω to a whopping 600 Ω – an uncommonly extended range on this price bracket, especially on the high end.
- Output power is not bad: 20mW (> 3.5V) @ 600 Ω load, 285mW (> 2.9V) @ 30 Ω load and 200mW (> 1.7V) @ 15 Ω Load.
- The declared system dynamic range is > 109dB (@3V) and THD+N is listed as <0.005% (-86dB).
While both latest values are not particularly impressive, they are definitely in-line with the product price bracket and it’s also worth noting that thanks to the above-mentioned built-in “purification” features nano iDSD BL will do its job on an “apriori less noisy” digital signal. This made me expect better results than what printed numbers say and as I’ll report later I was kinda right.
Two very important additional things are now to be noted about output.
First: the iEMatch port.
In its standalone incarnation iEMatch is a device to be plugged in between an amp’s headphone port and a IEM or Headphone cable, and vulgarly said it does 3 things:
- It “tricks” the amp into sensing a predetermined (average) load impedance of 16Ω, regardless of the IEM/Headphone’s real (average) one.
- On the opposite end it also “tricks” the IEM/Headphone into sensing a predetermined amp output impedance, regardless of the amp’s real one. The user can flip a switch and choose between 2.5Ω or 1Ω.
- It attenuates – think about it as if it “sinked” – the amp’s output by a predetermined amount: -12dB when output impedance is set to 2.5Ω, and -24dB at 1Ω
Such features are helpful on three counts:
One: By “raising the volume” the amp increases the “audibility” of the signal (the music) only, but the device “base noise” (a.k.a. “noise floor”) stays unchanged. Correspondingly, at low volume levels the device noise will be more audible as the music will not be “loud enough on top of it”.
So I should always turn the volume as high as possible to “kill base noise”, right?
Sadly, hearing music too loud is not only uncomfortable, but even dangerous for our hearing. Furthermore, “high sensitivity” IEMs get very loud very soon as we raise the amp’s volume.
Long story short: very often we are forced to actually “keep the amp volume way down” unless we want to hurt our ears, which is the opposite of what would be ideal to counter the system’s noise floor.
That’s a first spot where an attenuator helps.
iEMatch adds a sort of “tax burden” on the shoulders of the amp, prior to reaching the (possibly oversensitive) IEM. All other factors unchanged, this requires us to “turn the amp volume up some more” (even “way more”) to obtain the same loudness out of the IEM, and this will “automatically” help reduce noise floor audibility.
iEMatch is not the sole attenuator on the market of course but it’s probably the smartest. Most others obtain the purpose by simply adding a resistor in series with the output line – which may and often does induce unwanted skewage on the IEM/Headphone’s response. iEMatch does this with some more sophysticated circuitry which gets to the point with no or very minor modification on the output sound. And in my experience it really does.
Two: The vast majority of budget DAC devices are equipped with digital volume control. I won’t go into a quite technical explanation (check here for a good one), simply put a digital volume control offers full digital resolution output only at its end-scale position, and reduces digital resolution (and sound quality with it) as volume is progressively reduced.
In other words: here’s another case where we’d get better results by having our source device work at or near full-volume, but we normally don’t as it would be too loud for our ears.
And again, a (good) attenuator plugged on the DAC output forces the user to “raise the (digital) volume” more, thus reducing the resolution loss.
Three: Building amps properly capable to drive very low impedance loads is not easy for a number of very technical reasons that I won’t discuss here.
Sadly, quite a few brilliant IEM models are on the market carrying very low impedances, so the problem of finding a competent quality source for them is not a pointless exercise.
iEMatch helps many amps bias extremely low (<<16Ω) impedance IEMs by “letting them amps believe” those IEMs carry a 16Ω average impedance instead. The amp needs to be powerful enough to compensate for the severe (up to -24dB) power sinking involved, but when that condition is met the IEM will be correctly amped, and the difference in its sound output compared to when they are plugged onto another amp just unfit for low impedances is nothing less than huge.
For how it practically went for me on such a case read my article about my experience with BGVP VG4.
Inside nano iDSD BL ifi put a modified iEMatch circuit, offering non user-selectable -16dB attenuation and 4Ω output impedance. Is it as effective as the standalone version? Let’s see:
- As for reducing noise floor (hiss) audibility on extra sensitive IEMs the benefit is entirely there: -16dB is quite bearable attenuation vs nano iDSD BL’s max power so yes it’s well calibrated, it works big time. Indeed, I just recently used it to tame hiss from possibly the “hissiest” IEM I ever auditioned: TRN BA8 – which I wrote about here.
- As for maximising resolution connected with digital-domain volume control : no, you don’t get that from nano iDSD BL’s iEMatch port… for the simple reason that nano iDSD BL already has analogue volume control (a feature normally implemented on higher tier models). iEMatch can’t “fix” what is not broke in the first place 🙂
- As, finally, for impedance matching… well, I have my doubts here. 4Ω output impedance is… if you ask me not low at all when it comes to managing extra-low (<16Ω) impedance IEMs, and anyhow it’s more than 4 times higher than the Direct port’s own impedance, declared at <1Ω. Penon Sphere (6 Ω) does in fact sound more open, un-veiled and simply “better” on the Direct port vs. on the iEMatch port.
Synthetically: nano iDSD BL’s “iEMatch output port” is nice to have, although just for reducing / removing hiss from too-sensitive IEMs.
Second: S-Balanced wiring.
I presume you already understand what “balanced” is all about. If not, get a primer here.
Very simply put: a “balanced” design in a source device offers in theory noise reduction all along the entire line (analogue reconstruction, amping, internal and external transfers, up to the speakers/drivers). Less noise means DAC chips producing more accurate analogue sound, AMP offering better sound dynamics and much more.
Wow, so is balanced always to be preferred to single ended?
Not necessarily. Cost is a factor as always: having it all double… costs twice as much. Even more significantly: doubling all internal components doubles… noise too! So in short it’s not easy as it may seem.
In my factual experience: all budget / mid-tier source devices (DACs, AMPs, DAC-AMPs, DAPs) I came across implementing both single and balanced-ended internal paths – with the possible sole exception of Lotoo Paw 6000, now that I think about it – result in balanced-ended quality significantly better vs their single ended option. Conversely, those few higher-tier sources I checked and/or own offer single-ended outputs only, which happen to offer much better output quality than lower-tier balanced-ended siblings.
Exploiting a balanced source (DAC, AMP and/or DAP) requires IEM/Headphones to have “balanced cabling”, and correspondingly “balance plugs” (see here), which is no big problem of course but only if the IEM/Headphone offers modular cabling, allowing the user to swap cables according to sources. And even then, well, you often still need to buy an extra cable.
Many non-entry-level budget-tier balanced-scheme source devices offer both headphone output options, via two separate ports: one for balance-ended cables, the other for single-ended cables.
Ifi adopted a smart in-between option called “S-Balanced” (short for “Single-ended compatible Balanced”). Refer to their own whitepaper for a nice technical description. It is included in ifi Pro iCAN, xCAN, xDSD and nano iDSD BL.
As a consequence, instead of the usual dual separated output ports on the chassis, a cabling scheme is put in place behind the 3.5mm phone port on nano iDSD BL :
- When plugging 3.5mm TRS plugs – aka the ordinary 3.5 male connectors found at the end of 99.9% budget fixed-cable IEMs, and modular single-ended cables alike – the port delivers “normal” single-ended output. All single ended drivers on the market will seemlessly work in there. In addition to that, thanks to how internal cabling is designed, they will also get 50% reduced crosstalk – for free.
- When plugging 3.5mm TRRS plugs, aka “Hifiman 3.5mm standard” (see here) – the port delivers full “balanced-ended” output to balanced-cabled drivers, resulting in quite apparently cleaner and more dynamic sound.
3.5mm TRRS termination is very uncommon on today’s balanced IEMs and Headphones, so I needed to procure myself an adapter to exploit that (and you won’t be lucky enough to already have one in your drawer either, I’m afraid).
This is nice as it delivers full balanced-ended quality, and even improves single-ended quality a little bit, while keeping full backwards compatibility, all without requiring further faceplate space for an extra female connector.
Add that such dual-standard “trick” is applied both behind the Direct and the iEMatch port, too !
On the flip side, I find it odd that no 3.5 TRRS adapter is included inside nano iDSD BL’s box. Ok maybe I shouldn’t expect one to be bundled for free, but why none is available as an orderable SKU# from ifi ?
Nano iDSD BL supports a wide range of digital input formats and moreover resolutions: DSD up to 256, PCM up to 384KHz and – drumroll here – MQA up to 192KHz.
I’m not at all interested into MQA so I’m not going to assess that – and even if I did I would have zero comparative experience to rely on.
On the back panel a small switch also allows the user to choose between two filters labelled “Listen” and “Measure”. The Listen option enables a Minimum Phase bezier filter, while the Measure option switches to a Linear Phase Transient-Aligned filter.
The topic may become too technical but let me try to simplify: a Minimum Phase filter makes sound “behave” more closely to our human auditory system – which is incapable of perceiving vibrations before an impulse, and tends to like when those following it over time are smoother – and is therefore by many called “more musical”. A Linear Phase filter yields a little bit edgier notes, which is indeed preferred by a population of listeners, but most of all comes handy when submitting the device to sampling and measuring, hence its given label name (“Measure”).
One more very important note is deserved about available firmware versions and their differences.
When I acquired it, my nano iDSD BL unit carried the latest available fw, version 5.3c. I looked into possible firmware variations and I found something quite interesting, as follows:
|F/W version||Key notes|
|5.2 “Limoncello”||DSD512 (Windows), DSD256 (Mac) support|
768kHz (on capable machines)
No MQA support
|5.3||Full MQA support|
DSD256 (Windows), DSD128 (Mac) support
|5.3c||Same as 5.3 plus:|
GTO filter, which upsamples USB audio
As you can read on ifi’s PDF paper linked above, Gibb’s Transient Optimised (GTO) filter is supposed to be an upgrade to the previous Minimum Phase Filter. There’s much more to it, read the paper 🙂
Long story short again: by downgrading from 5.3c to 5.3, thus going back to the “original” Minimum and Linear Phase filters and their upsampling algorithms I perceive a distinct sound output improvement! May be a matter of tastes of course, or maybe related to the GTO upsampling being less refined (yet) than its predecessors. Be as it may, to me it sounds better, and I settled to 5.3.
Lastly, the form factor is not “ultrasmall” nor “ultrathin” but it stays very easily transportable, and pocketable – at least in terms of coat pockets. With a little intention it can be “paired” with another device, also exploiting the 2 rubber bands found in the box. Weight is also quite light (139g) and the 1200mAh battery offers up to 10h of theoretical life, which I could test down to 7-8hrs max which is good in its class.
How does it sound…
After all these structural descriptions it’s finally time to go back to the prologue and assess how this light (also quite money-wise) device performs in terms of sound output.
…as a DAC-AMP ?
Much like in virtually all other cases I encountered, true-balanced output is better than single ended on nano iDSD BL too. Once the 3.5 trrs adapter riddle gets sorted, using nano iDSD BL’s true-balanced features is a strong recommendation: soundstage, imaging and most of all dynamics get significantly better.
Even on its balanced Direct output nano iDSD BL’s general tonality is warm, and timbre is dark-ish. Bass is well bodied in positive, yet relatively slow in negative, this predominantly resulting in some bleeding into the mids. Trebles lack some sparkle, not a masterpiece but better than the bass. Range extension is by-laterally, deifinitely on par with devices on this price bracket at least as far as my experience goes. Soundstage and imaging are on the average mark for the price.
…as a DAC, with another AMP ?
Nano iDSD BL’s Line Out port offers surprising better quality.
Plugging the amazing little amp that I use as my “hyperportable transparency reference” (iBasso T3) in, nano iDSD BL’s sound presentation changes dramatically: “darkness” goes away and the general timbre becomes definitely neutral, tonality keeps a modest, possibly welcome warmth, treble suddenly becomes airy and unoffensively sparkly. Clarity goes up 2 notches, soundstage gets airier, separation gets much better too. By the way: T3 is single-ended only!
So putting it simply: nano iDSD BL internal amp does not seem to offer justice to the quality of its dac, which in facts seems capable to kick much above its weight.
…(unfairly) compared to the Mojo ?
I started the day asking myself if this device could hold a candle to Mojo sound-wise though. How about that? Simply put: as a standalone unit the answer is “not by a mile”, while as a DAC to be complemented by a decent (or even good, why not) external amp the score changes quite a bit.
Compared to nano iDSD BL’s Direct full-balanced output Mojo’s output wins hands down an all counts: bilateral extension, bass and treble control, clarity, soundstage, imaging. It simply partakes to a higher class, full stop.
Escaping from nano iDSD BL’s internal amp via the LO port, and adopting an even inexpensive amp as the above mentioned iBasso T3, the gap reduces big time. Mojo still wins by definition, extension and its outstanding (unique in its bracket, possibly) capacity to manage background voices with incredible clarity, but the timbres and tonalities become at least comparable, in the same ballpark so to say.
…or vs to other “more in-line” alternatives ?
Ok nano iDSD BL is not a Mojo. Where does it stand then ?
Let’s run another head to head comparison: Fiio BTR5 DAC/AMP.
The two devices are apriori not really equivalent in terms of intended use, and features: BTR5 is indeed marketed as a BT DAC-AMP for IEMs mainly, with some complimentary USB connectivity but that’s all, nano iDSD BL as an easily portable USB device supporting MQA, higher DSD and PCM resolutions, and high impedance cans. Still, BTR5 gained vast market appreciation in terms of high-sound-quality-for-its-price, and being its price roughly 40% less than nano iDSD BL’s I’m stimulated to compare the two, using BTR5 as a USB device in this case of course.
Compared to nano iDSD BL, BTR5 bass is less bodied (but also less bleeding), mids and highmids come up much less controlled, grainy, and raising volume makes them edge quite quickly. Stage on BTR5 is evidently narrower, imaging is more congested, instruments come accross less defined and separated. BTR5’s dynamics, while not bad per se, are also a notch below nano iDSD BL’s.
Such comparison refers to both devices’ balanced outputs by the way, using a pair of TIN T4 as IEMs.
Let me try another comparison I have at easy hand: my ol’ Fiio X3 mk-III.
I find it interesting as a comparison as I’ll be using X3 as a standalone device, not connected to my PC and therefore apriori unaffected by USB noise. As X3’s balanced output is – as an exception to what commonly happens – not really better then its single ended one, I’ll run this comparison on both devices’ single ended channels for a change. I’ll use a pair of final E1000 as supremely neutral drivers.
X3 comes out as a further bit warmer (nano iDSD BL’s SE already being such), and its trebles are even less extended – which on the up side makes it nigh-impossible to make X3 go edgy let alone screamy. X3’s soundstage is also a bit less extended, imaging is on par. Simply put: the two devices’ single ended phone out are definitely comparable in terms of overall quality.
Now let’s compare the two devices’ Line Outs – always with the help of my iBasso T3.
X3’s tonality stays almost unmodified, trebles become just a little bit edgier but it’s a nuance; soundstage, imaging and separation get better.
On the other hand, as previously noted, nano iDSD BL gets much better when its LO is exploited: bass is cleaner and faster, bleeding is very modest, treble still unextended but much airier, detailed and engaging, soundstage and separation get 2 notches up.
Alas!… ifi nano iDSD BL does not sound on par with Mojo, costing 2.5X more. Is it really a problem? Of course not.
Its phone output quality, especially on the full-balanced side, is in line with its price bracket, and offers the significant extra advantage of the built-in iEMatch circuitry proving decisive to cope with extrasensitive IEMs hiss, paired with direct support – and enough muscle power – for 600 Ohm headphone on the opposite end.
Its DAC – taken alone – is more than good, I’d call it outstanding actually. Its reconstruction quality is not so easy to find at this price in a semi-pocketable device. Those – like me – who want to pull the max out of nano iDSD BL in terms of sound quality will pair it with a portable amp, and will get a very significant device for a quite affordable overall price.
|Outstanding DAC quality for the price||External AMP recommended for best sound quality output|
|Full balanced output support||Warm tonality|
|Built-in USB regen and reclock working features||Uncommon 3.5TRRS adapter required for full balanced exploitation|
|Hiss-taming iEMatch features|
|Support for high impedance headphones|
|Compact and lightweight, nice form factor compromise|