[earfonia]

This post is the extended part of the review of iFi micro iDSD Black-Label posted here:
http://www.head-fi.org/products/ifi-audio-micro-idsd/reviews/17934
 
Review Sections:
Introduction, Summary, Pros, and Cons
Sound Quality
Features and Measurement
 
 
I wanted to post all of the sections of this review there in the review section, but was stuck with the maximum content limit of the review posting section. So I posted my complete review here.
 
 

 

[earfonia]

Many thanks to iFi for the tour program, to let us have some experience with the new iFi micro iDSD Black-Label!

 



 

iFi micro iDSD Black-Label product web page:
http://ifi-audio.com/portfolio-view/micro-idsd-bl/

Manual:
http://ifi-audio.com/wp-content/uploads/data/manual/miDSDBL_manual.pdf
 

The iFi micro iDSD Black-Label is the improved version of the previous iFi micro iDSD. iFi has shared to us in detail, many of their design considerations during the development of the micro iDSD. Lot’s to learn from the post, therefore I think it is worth to post the link to the early discussion here:

http://www.head-fi.org/t/711217/idsd-micro-black-label-tour-details-page-147-release-info-page-153

I bought the iFi micro iDSD pre-ordered from Stereo Singapore in September 2014. Since then it has been one of my favorite portable DAC. I like the line output sound quality especially when paired with iFi micro iCan, but the headphone output of iFi micro iDSD requires some matching to sound best. My biggest complaint so far from the iFi micro iDSD is the quality of the iEMatch switch that often glitchy and causes loss of the right channel or severe channel imbalance. The volume pot of my iFi micro iDSD also has audible channel imbalance below 9:30’ position. Together with the glitchy iEMatch switch, it makes me difficult to use it for sensitive IEMs. I’m glad to say that I found the channel imbalance of the review unit of the iFi micro iDSD Black-Label has been greatly reduced, and practically I didn’t have any channel imbalance issue even at low volume setting. I hope this will be the case for all iFi micro iDSD Black-Label units. I also hope that the iEMatch switch durability has been improved on the Black-Label version.
 

 

 

Some of the improvements in the Black-Label version are some of the electronic components, power sections, clock system, and some other improvement on both digital and analog circuit sections, including the implementation of custom Op-Amp. There is no changes in the technical specifications and features from the previous iFi micro iDSD, so feature wise both the iFi micro iDSD and the Black-Label version are similar. The improvement is more on the sound quality. One might ask when there is an improvement in the sound quality, why it is not shown in the specification? The simple answer is, the measured specifications don't cover all aspects of the sound quality. Basic specifications such as FR, THD, and SNR are only a few aspects of the audio quality and quite often are not advertised in detail. THD for example, usually only advertised as average THD, but manufacturer usually doesn't give further detail like what is the distortion profile across the audio band, which type of distortion that is more dominant, etc. Therefore, usually, it is close to impossible to judge the sound quality of a DAC or Amplifier only by looking at the advertised specifications.

In summary, iFi micro iDSD Black-Label is an excellent sounding, feature rich DAC + headphone amplifier. It does require some knowledge to get the most out of it. Sound quality wise, it is on the neutral side with no obvious coloration. For those who are looking for warm, intimate, mellow type of sound signature, better look elsewhere. Transparency, clarity, speed, and detail retrieval are still the main characteristics of iFi micro iDSD Black-Label sound signature, similar to the iFi micro iDSD. And iFi has improved it further in a more musical manner on the Black-Label version. Besides some technical improvement from the previous iFi micro iDSD, the sound quality improvement that I observed on the Black-Label are transparency, dynamic, and instrument separation. The Black-Label is more transparent and realistic sounding than the already transparent sounding iFi micro iDSD. Not a night and day differences, but noticeable. And I’m glad to say that the increase in transparency and detail retrieval doesn’t make the iFi micro iDSD Black-Label sounding more analytical than the iFi micro iDSD. Subjectively, iFi micro iDSD Black-Label is actually sounding more musical to me. Even though not by much, I do prefer the iFi micro iDSD Black-Label sound quality than the iFi micro iDSD.
 

 

Pros:

1. Feature rich with high performance to price ratio.
2. Neutral sound quality with superb transparency, speed, and detail retrieval.
3. Good multi-platforms compatibility with various operating systems.
4. Isolated USB and analog ground with excellent USB EMI noise rejection.
5. Various digital and analog filters to suit listening preference.
6. A wide range of gain and headphone output power settings to suit various loads, from sensitive IEMs to demanding headphones.
7. Useful and good sounding analog bass boost and stereo enhancement analog circuit.
8. Good battery life.

Cons:

1. 1-2 seconds of silence at the beginning of playback (from a stop). This short period of silence causes the first 1-2 seconds of the song gets muted at the start. This can be quite annoying for some songs that start immediately at the 1st second. This is the only most annoying flaw of iFi micro iDSD Black-Label so far, but I believe it can be fixed by firmware update if iFi is willing to fix it, or probably by releasing a special driver only for PCM playback. I notice that the silence period is slightly longer on the iFi micro iDSD Black-Label compared to the iFi micro iDSD. Due to the short review time, I’ve only tested it with foobar v1.3.12 (WASAPI and DSD ASIO). Probably there is a way to shorten the silence from the setting, but I didn’t have enough time to play around with the setting or checked this symptom using other media player applications.​ This short period of silence at the beginning of playback is could be due to ‘pop’ issue described here:

          http://ifi-audio.com/audio_blog/pop-goes-dsd-why-does-this-happen/

1. 1 LED indicator to indicate many operating conditions. It is not user-friendly to expect a user to memorize so many color codes from a single LED indicator.
2. Volume level indicator is hard to see.

Suggestions for improvements:

1. To shorten the start play silence.
2. A more user-friendly LED indicator. Suggested 3 LEDs indicator as described at the end part of this review.
3. White or silver volume level indicator for better visibility.
4. Better design rubber feet with a stronger attachment to the metal case. It is preferable to have better rubber feet that have been fixed to the metal case from the factory.

 

 

[earfonia]

Sound Quality

Sound quality observations were done using my regular test tracks as shown at the end of this review. As for headphones and IEMs, I mostly used the following during this review:
 
Headphones:
Audio-Technica ATH-R70x
Audio-Technica ATH-MSR7
Beyerdynamic T1
HiFiMan HE-6
Massdrop HD6xx
Sennheiser HD800
Philips Fidelio X1
STAX SR-L300 + SRM-252S
 
In-Ear Monitors:
AK T8iE Mk2
Brainwavz B200
DUNU DN-2000
 

 
 

Headphone Output Sound Signature
Transparent with good detail and dynamic is probably the simplest way to describe iFi micro iDSD Black-Label sound signature. Generally, it sounds quite neutral with no obvious coloration. The iFi micro iDSD Black-Label is not a warm and mellow sounding type of DAC that tends to ‘beautify’ recording flaws. It is a bit on the dry and analytical side, but iFi has done it in a nice and musical way. It is still lean on the analytical side but it doesn’t sound thin. iFi micro iDSD Black-Label has excellent stereo imaging, spacious and holographic with good depth. The headphone output is powerful with lightning fast transient, always giving the impression that it can drive any IEMs and headphones with ease. iFi micro iDSD Black-Label might not be for those looking for smooth warm and polite sounding DAC, but I imagine that the Black-Label could easily be the sound engineer favorite portable DAC.

With the mentioned headphones and IEMs above, I prefer to match the iFi micro iDSD Black-Label with the less analytical sounding ones. Though pairing the iFi micro iDSD Black-Label with HD800 and T1 give and impressive transparent and holographic sonic presentation, but overall still rather too bright for my preference. The iFi micro iDSD Black-Label despite the small size also surprisingly able to drive the HiFiMan HE-6 quite well, but the pair also a bit too bright for me.
 

 
 
So the headphones and IEMs that I consider pairs well with iFi micro iDSD Black-Label are:

Audio-Technica ATH-R70x
Philips Fidelio X1
STAX SR-L300 + SRM-252S (Connected to Line Output)

AK T8iE Mk2
Brainwavz B200

Most surprising is how iFi micro iDSD Black-Label improves the sound quality of the new Brainwavz B200, dual BA drivers IEM. B200 usually sounds polite with soft treble with my Onkyo DP-X1, not so much excitement. But when driven from iFi micro iDSD Black-Label, the treble suddenly shines and sparkling nicely. B200 sounds more lively and exciting with iFi micro iDSD Black-Label. Quite a significant improvement. The Audio-Technica ATH-R70x and STAX SR-L300 + SRM-252S (Connected to Line Output) are also wonderful pairs with the iFi micro iDSD Black-Label.
 

 


Comparison to iFi micro iDSD Headphone Output
At the same volume level, the Black-Label sounds more powerful with greater dynamic and sense of driving power. Bass sounds slightly thicker, tighter, punchier, and has a better texture. I feel both bass and midrange texture and micro dynamic seems to be improved on the Black-Label, giving a slightly better perception of depth, transparency, and instruments separation. Treble is more or less the same, but on some recordings with sibilance, the sibilant sounds a tad more prominent on the older micro iDSD, and a tad less sharp on the Black-Label. Just a tad, basically the difference is quite small. The level of treble and treble sparkle are about the same, but with slightly different character. The sparkling character of the treble of iFi micro iDSD Black-Label is somehow sounding a tad more natural to my ears. In summary, the Black-Label sounds more transparent, bolder, and more energetic than the previous micro iDSD. The difference is audible but not a night and day kind of differences. What I mean is, that if we already have the micro iDSD, I think it is not necessary to sell it to get the Black-Label. But if I have to choose, I would definitely choose the Black-Label over the silver micro iDSD.
 

iFi micro iDSD Black-Label Line Output + iFi micro iCan
I remember that in past, ever mentioned in the forum that some suggested to iFi to tweak the headphone amplifier of the micro iDSD to be closer to the sound signature of the micro iCan. So is the headphone amplifier of the iFi micro iDSD Black-Label now sounds close to the micro iCan? Well not quite yet. The headphone output of iFi micro iDSD Black-Label sounds dryer than the iCan. In my opinion, the iFi micro iDSD Black-Label line output connected to micro iCan still sounds better. They do share some similarity, like the level of transparency, detail retrieval, and dynamic are probably about the same, but the micro iCan sounds slightly smoother and warmer that makes the micro iCan more friendly for analytical headphones like HD800 and T1. The micro iCan has slightly longer decay than the Black-Label headphone amplifier that makes it sounds less dry and more pleasing to my ears. I’m still hoping that one day I could have a new generation of micro iDSD with the headphone out sound quality that is similar to the micro iCan sound quality. So I don’t have to bring two units to enjoy the sound quality of the combination of micro iDSD + micro iCan. In the past, I’ve compared the line output sound quality of my micro iDSD to bigger and more expensive desktop DACs, and micro iDSD line output has been proven to exceed its price bracket. iFi micro iDSD Black-Label line output doesn’t disappoint and even improved it further on the transparency, detail, and instrument separation. Very impressive line output sound quality from such a small portable DAC. IMHO, iFi micro iDSD Black-Label is worth it even just for the DAC section alone.
 

 

 
 

Chord Mojo (Headphone Output Comparisons)
Listening to classical DSD tracks, Super Artists on Super Audio sampler vol.5 from Channel Classics Records, when using the analogy of medium and large concert hall, Chord Mojo sounds like we are listening to the concert in a medium size hall, with a tad better micro detail and impact. Listening to Chord Mojo is like sitting closer to the musical performance, more intimate presentation with a tad clearer micro detail and slightly better sense of micro-dynamic. iFi micro iDSD Black-Label, on the other hand, provides a more spacious sensation, like listening in a larger hall. Less intimate with a larger sense of space. iFi micro iDSD BL is also perceived as a tad smoother sounding than Mojo. The difference is not day and night, but quite easy to distinguish. Both performs admirably in their own ways. I do need more time for better comparison between Mojo and iFi micro iDSD Black-Label, but the most distinguishable difference is in the presentation, between the more intimate presentation of Mojo and the more holographic presentation of iFi micro iDSD Black-Label. Honestly, I can’t really tell which one is better. I guess it is not for better or worst but more about personal preference.

 

[earfonia]

Features and Measurement

Disclaimer:
Measurement in this review was conducted for the purpose of general observation of the review unit, and not meant for comparison with the official specification from iFi. Test results may differ from the official specification due to many factors, mostly due to the different setup and measurement equipment. Therefore all test results posted in this review are not to be compared with another measurement result, and only useful for comparison between measurement in a similar setup.

RMAA measurement was used in some measurement as a convenient way to get frequency response graph. Although RMAA is a convenient way to check frequency response, RMAA is not a very reliable and accurate method to measure other parameters. So please read the measurement result with that consideration. HRT LineStreamer+ was used as the audio interface for all RMAA measurements.

The following is my standard settings on QuantAsylum QA401 (24-bit audio analyzer) for amplitude measurement:
FFT resolution: 32768 points
Averaging: 5
Sampling frequency: 48.0 kHz
Windowing: FlatTop
Weighting: None

SNR measurement uses similar setting as above but using different windowing and weighting setting:
Windowing: Hann
Weighting: A weighting

THD measurement setting is similar to amplitude measurement setting but with different windowing function:
Windowing: Hann

Unless otherwise specified, the settings above are currently my standard settings for amplitude measurements, SNR, and THD on QuantAsylum QA401.
 

 
 
 
Table of Features
 
 
Both the older version of iFi micro iDSD and the Black-Label version have similar features and specifications, therefore I listed only the Black-Label version in this table of features. 
Below is a table of features of iFi micro iDSD Black-Label in comparison with Chord Mojo. Voltage, current, and impedance indicated in this table are from my own measurement, and not from the official specification from the manufacturer. This review is based on firmware version 5.20.
 

Parameter	iFi micro iDSD Black-label​	Chord Mojo​
DAC	Dual-Core Burr-Brown (2-DAC Chip)​	Chord Custom FPGA DAC​
PCM	PCM 768/ 705.6/ 384/ 352.8/ 192/ 176.4/ 96/ 88.2/ 48/ 44.1kHz​	PCM 768/ 705.6/ 384/ 352.8/ 192/ 176.4/ 96/ 88.2/ 48/ 44.1kHz​
DSD	up to DSD 512​	up to DSD 256​
Multi-platform compatibility	Yes​	Yes​
USB Input	USB 2.0 type A “OTG” Socket (with iPurifier® technology built-in)​	Micro-B USB​
SPDIF Coaxial Input	RCA - Up to 192kHz PCM​	3.5mm jack - Up to 768kHz PCM​
SPDIF Optical Input	Up to 192kHz PCM​	Up to 192kHz PCM​
SPDIF Output	RCA Coaxial - Up to 192kHz PCM​	-​
USB to SPDIF Conversion	Yes - Up to 192kHz PCM​	-​
Selectable Filter	Yes - 3 options for each PCM and DSD​	-​
Analog Line Input	Yes - 3.5mm socket​	-​
Analog Line Output	Yes - Dedicated RCA​	Integrated with headphone output​
Line Output Level	Direct: 2V Fixed PreAmp - Eco: 0 - 2.18 V Variable - Normal: 0 - 5.66 V Variable - Turbo: 0 - 6.43 V​	0V - 4.79V Variable​
Headphone Output	1x 6.5mm socket​	2x 3.5mm socket​
Adjustable HO Gain	Yes - 9 combinations​	-​
Maximum HO Voltage - measured @ 600 ohms load	9.71 Vrms​	4.79 Vrms​
Maximum HO Current - measured @ 15 ohms load	306 mA​	199 mA​
HO Output Impedance	IEMatch Off: 0.34 ohms IEMatch High Sensitivity: 4.1 ohms IEMatch Ultra Sensitivity: 0.95 ohms​	0.44 ohms​
HO SNR @ 50 mV @ 33 ohms (for very sensitive IEM)	Eco - Ultra Sens. : 87.3 dB Normal - Ultra Sens. : 87.0 dB Turbo - Ultra Sens. : 83.0 dB​	82.9 dB​
Volume Control	Analog Potentiometer​	Digital​
Extra Features	XBass Plus, 3D Matrix Plus, Polarity Switch, & USB Power Bank (5V, 1.5A)​	-​
Weight	310g​	180g​
Dimension	177mm (l) x 67mm (w) x 28mm(h)​	82mm (l) x 60mm (w) x 22mm (h)​

 
 

USB Input with iFi iPurifier® technology
Here is some info from the iFi iPurifier® webpage:
http://ifi-audio.com/portfolio-view/accessory-ipurifier/
From the explanation above, the function of iFi iPurifier® with PurePurification® technology is to remove unwanted EMI from USB audio + power. Therefore I prepared a test to test it, in comparison with Chord Mojo that doesn’t advertise any special feature on its USB input. One thing I know from the measurement is, the USB ground of iFi micro iDSD BL is isolated from the analog output ground. Having some experience with ground loop and EMI on USB cable, isolation of the USB ground and the analog circuit ground is an important feature for any USB audio interface. Many of my USB DACs don’t isolate USB ground and the analog circuit ground, and sometimes I do hear EMI noise especially when there is cordless phone near the setup. Chord Mojo for example, it doesn’t isolate USB ground and the analog circuit ground, as shown in the picture below.

  
 

The EMI test setup is as the following:

 

The following video is the test, showing the effect of strong electromagnetic interference (EMI) on USB cable, and the effect of iFi iPurifier® technology isolating the USB ground and the analog circuit ground to stop EMI noise from transmitting from the digital domain to the analog domain. I did the test in 2 setups, while the DAC is idle and when the DAC is playing a silent track. The results are pretty close between the 2 setups.

Note: The overall noise has been double amplified to make it more audible on this video. In reality, the EMI noise is much softer, but audible on sensitive In-Ear Monitor.
 

 
It is always a good idea to use a short USB cable for USB DAC 

 
 

Multi-Platform Compatibility (OTG Feature)
I’ve only tested the iFi micro iDSD Black-Label using Windows PC and Android devices. It works well with my old Android smartphone, Samsung Galaxy S4 with USB Audio Player PRO (UAPP), also fully compatible with my Android based DAP, the Onkyo DP-X1.
 

 

An interesting note when using the iFi micro iDSD Black-Label with a smartphone is, that it will try to recharge while in Sleep Mode. That means when it is connected to a smartphone, and being idle for some time and goes into sleep mode, it might withdraw power from the smartphone to charge its battery. Meaning it will drain the smartphone battery. iFi has a special firmware to disable the sleep mode to prevent it draining the smartphone battery. If we often use it with mobile devices, it is a good idea to use that version of firmware. Here is the link to the firmware page:

http://ifi-audio.com/micro-idsd-ifi-xmos-firmware/
 


SPDIF Input and Output
The coaxial SPDIF switches automatically between input and output function. When iFi micro iDSD Black-Label is being used as USB DAC, the coaxial SPDIF functions as SPDIF output. Without USB audio signal applied, the SPDIF coaxial SPDIF input. The optical interface of the SPDIF only functions as input.

When USB input is used, iFi micro iDSD Black-Label functions as USB to SPDIF Converter. And the cool thing about this feature, it works even with the unit is powered off. The USB DAC and SPDIF output work using the power from the computer, therefore it doesn’t require for the unit to be turned on. The SPDIF works for all PCM sampling rates Up to 192kHz.


 
 

Digital and Analog Filters
The feature of selectable filters though probably a bit too technical for the majorities, it is definitely an interesting feature for audiophile enthusiasts. It is rather too lengthy to discuss about digital and analog filters in this review, so I will just show some spectrum analysis and the pre and post ringing of the filters to give a better idea of the characteristic of each filter. I used 750 Hz square wave (0.5 dBFS), measured from the RCA line output directly connected to the audio analyzer or oscilloscope. The line output switch was set to Direct. The reason for using square wave is because I could observe both harmonic frequencies and the pre-post ringing characteristic of the filters at the same time. I measured both iFi micro iDSD and iFi micro iDSD Black-Label. Both have a pretty similar response, therefore I will only post the result of the iFi micro iDSD Black-Label. Chord Mojo also measured for comparison. From the sound quality perspective, Minimum Phase is what I use most, and Bit Perfect next. Here is a post from iFi to find out more about the filters:

http://www.head-fi.org/t/711217/idsd-micro-black-label-tour-details-page-147-release-info-page-153/1050#post_10662147
 

PCM Playback - FFT Measurement

iFi micro iDSD Black-Label - Bit Perfect - 750 Hz Square Wave - 24bit / 96kHz PCM:

 
iFi micro iDSD Black-Label - Minimum Phase - 750 Hz Square Wave - 24bit / 96kHz PCM:

 
iFi micro iDSD Black-Label - Standard - 750 Hz Square Wave - 24bit / 96kHz PCM:

 
Chord Mojo- 750 Hz Square Wave - 24bit / 96kHz PCM:

 

The FFT analysis of the Chord Mojo output is quite interesting with the present of the even-integer harmonic frequencies, as the ideal square wave contains only components of odd-integer harmonic frequencies. The level of the even-integer harmonic frequencies from Chord Mojo output is very low, but still, it is an interesting observation as we can see that the output of iFi micro iDSD Black-Label is relatively free from the even-integer harmonic frequencies. Besides that, the roll-off frequency of Chord Mojo filter for PCM playback looks very sharp, much sharper than the micro iDSD standard filter. Both have a very different approach to the implementation of the low-pass filter.
 

PCM Playback - Pre and Post Ringing Characteristic
 
iFi micro iDSD Black-Label - Bit Perfect - 750 Hz Square Wave - 24bit / 96kHz PCM:

 

 
iFi micro iDSD Black-Label - Minimum Phase - 750 Hz Square Wave - 24bit / 96kHz PCM:

 

 
iFi micro iDSD Black-Label - Standard- 750 Hz Square Wave - 24bit / 96kHz PCM:

 

 
Chord Mojo- 750 Hz Square Wave - 24bit / 96kHz PCM:

 

 

Bit Perfect filter that essentially no low-pass filter, as expected, showing a good shape of Square Wave with only a tiny post ringing. The rest showing low to moderate pre and post ringing due to the high frequencies filtering. Chord Mojo output has the longest pre and post ringing, probably due to the sharp rolls-off filter.
 

RMAA Comparison Between the Digital Filters
Here is the summary of the RMAA measurement results between the 3 digital filters. Most interesting is the frequency response graph, to see the different steepness of the high-frequency roll-off.
 

Test​	ifi micro iDSD BL Bit Perfect Nor-H HO 2V​	ifi micro iDSD BL Min Phase Nor-H HO 2V​	ifi micro iDSD BL Standard Nor-H HO 2V​
Frequency response (from 40 Hz to 15 kHz), dB:​	+0.58, +0.21​	+0.56, +0.53​	+0.00, -0.03​
Noise level, dB (A):​	-91.1​	-91.3​	-91.4​
Dynamic range, dB (A):​	90.9​	91.2​	91.2​
THD, %:​	0.0041​	0.0043​	0.0042​
IMD + Noise, %:​	0.029​	0.029​	0.029​
Stereo crosstalk, dB:​	-83.1​	-84.0​	-83.9​

 

 
 

DSD Playback - FFT Measurement

I forgot to capture the DSD playback FFT measurement from the iFi micro iDSD Black-Label before I returned the unit, therefore the following it the FFT measurement of the iFi micro iDSD, which is similar to the Black-Label version.

iFi micro iDSD - Ultra-wide (Bit Perfect) - 750 Hz Square Wave - DSD128:

 
iFi micro iDSD - Extended (Minimum Phase) - 750 Hz Square Wave - DSD128:

 
iFi micro iDSD - Standard (Standard)- 750 Hz Square Wave - DSD128:

 
Chord Mojo- 750 Hz Square Wave - DSD128:

 

DSD Playback - Pre and Post Ringing Characteristic

iFi micro iDSD Black-Label - Ultra-wide (Bit Perfect) - 750 Hz Square Wave - DSD128:

 
iFi micro iDSD Black-Label - Extended (Minimum Phase) - 750 Hz Square Wave - DSD128:

 
iFi micro iDSD Black-Label - Standard (Standard)- 750 Hz Square Wave - DSD128:

 
Chord Mojo- 750 Hz Square Wave - DSD128:

 
 
While the PCM filters significantly affect the pre and post ringing of the Square wave, the DSD filters affect more of the output level. This is an important note to be taken when comparing the DSD analog filters using an external amplifier, as the output level is different at different filter setting. Chord Mojo Square wave output looks slightly better on DSD playback in comparison to the PCM playback, with less pre and post ringing.
 


Volume Channel Imbalance
One most important aspect for me from the analog section is the channel balance. No matter how good the sound quality is, if the left and right channels are severely imbalance, all are practically useless. The following measurement is the comparison of the level of channel imbalance of the headphone output of my iFi micro iDSD and the review unit of iFi micro iDSD Black-Label. Power mode was set to Normal on both.


 

 
 
The review unit of iFi micro iDSD Black-Label has very good balance between left and right channel. Mostly less than 0.5 dB across the range, only peaking at around 0.66 dB on a certain range. While on my iFi micro iDSD, I need to switch and tap the glitchy iEMatch switch a few times to get the Right channel sounds balance with the Left channel. Even after getting the right channel sounds right, the channel imbalance on my iFi micro iDSD is still not very good. I hope all iFi micro iDSD Black-Label will have similar performance as the review unit I tested.
 
 

Maximum Volume Setting
Another note for the volume control is the safe maximum volume setting at different Power mode. On Normal and Turbo power mode, 0 dBFS signals (maximum level in the digital format) will clip before the volume knob turned to the maximum position. For example, at Turbo mode and iEMatch sets to Off, loud recording may start to clip at volume knob position around 13:00’ o’clock. Modern recordings usually make full use of the available headroom to achieve maximum loudness. That means the loud parts of the recording may be close to 0 dBFS. Therefore it is recommended not to set the volume beyond the clipping point to avoid distortion. Only when the recording is very soft and needs more volume, then it can be considered to use higher volume setting.

Below is the maximum volume position at various power mode settings, before the THD reach 1%. Or in other words, before the 0 dBFS clips. All measurement were done using QuantAsylum QA401 audio analyzer and OWON VDS3102 oscilloscope, monitoring the THD level and the shape of the sine wave at the same time. 600 ohms load was used.


 

The following screenshots are an example of maximum volume observation with THD monitoring, and SNR measurement with ‘A weighting’ applied and different windowing setting. Notice the significant different of SNR value on the two different settings. QA401 has a built-in 20 dB attenuator. When applied, the SNR measurement has to be compensated as it will cause lower reading of SNR. In my setup, the compensation is measured around 3.7 dB higher. The ‘A weighting’ will also improve SNR reading by around 3 dB. Lastly, the different windowing will also cause different SNR reading. Therefore in many cases, we cannot simply compare SNR value from different measurement setup. SNR measurement is not as simple as voltage measurement. In this case, for example, SNR reading from 2 different setups can be around 13 dB different.


 

 

Here is a table of maximum output voltage measured at <1% THD of 1kHz tone, at 600 ohms load, and the SNR measurement at maximum voltage (dynamic range).

	Max HO Vout @ 600Ω (V)​	Max HO DR @ 600Ω (dBA)​
	L (US)​	M (HS)​	H (Off)​	L (US)​	M (HS)​	H (Off)​
Eco	0.137​	0.583​	2.120​	90.5​	91.2​	92.0​
Normal	0.260​	1.160​	4.900​	90.8​	91.2​	93.5​
Turbo	0.300​	1.480​	9.710​	90.0​	90.3​	93.2​

 
L: Ultra Sensitivity
M: High Sensitivity
H: Off
 
 

Headphone and Line Output SNR
When looking at SNR measurement, I take 2 simple practical approaches:
Headphone output: SNR at intended listening value is expected around 85 dB or better.
Line output: SNR @ 2Vrms is expected to be around 96 dB or better, and around 90 dB or better @ 1Vrms if the line output is variable.

For headphone output, the threshold of our human hearing sensitivity on hiss noise is around 85 dB SNR. Meaning, when the hissing noise is 85 dB lower than the music signal, we generally won’t be able to hear the noise, and the headphone output can be considered quiet when there is no music is playing, no audible hiss noise on the IEMs or headphones. This is more relevant with sensitive IEMs. My standard listening level on general sensitive IEMs, like my 1964 Ears V3, is around 50 mV (or even less). Therefore I expect the SNR at 50 mV is to be around 85 dB or better. SNR between 80 dB to 85 dB is practically still pretty quiet and acceptable. Most probably only a very mild audible hiss noise can be heard in a quiet environment. Less than 80 dB, hiss noise will start to be more audible and some people that sensitive to hiss noise might start to complain. But please take note, 50 mV is my standard measurement for sensitive IEMs. Most general IEMs are not very sensitive, therefore even 75 dB to 80 dB SNR at 50 mV can still be considered quiet for many IEMs. So, the rule of thumb, 80 dB SNR or greater is very good, and 85 dB or greater is excellent.

For normal headphones, my standard SNR measurement is at around 250 mV, with the same 80 dB and 85 dB target SNR for very good and excellent category.

As for line out, I expect higher SNR because it is measured at a higher voltage, 1V and 2V. And when we look at Redbook CD dynamic range, which is 96 dB, it can be a practical approach for line output SNR performance to be able to meet Redbook CD maximum dynamic range.

Somehow my SNR measurement result is significantly lower than iFi specification for the micro iDSD Black-Label. The significant differences are most probably due to the different quality of the measurement equipment. Practically, from my short experience with the iFi micro iDSD Black-Label, I didn’t have any issue with noise, both on headphone output and line output. With the right setting, the headphone output is practically quiet even on my most sensitive IEM.
 

	50mV HO SNR @ 33Ω (dBA)​	250mV HO SNR @ 33Ω (dBA)​	1V HO SNR @ 600Ω (dBA)​
	L (US)​	M (HS)​	H (Off)​	L (US)​	M (HS)​	H (Off)​	L (US)​	M (HS)​	H (Off)​
iDSD BL - Eco	87.3​	83.1​	74.6​	-​	90.0​	86.8​	-​	-​	90.0​
iDSD BL - Normal	87.0​	82.6​	73.8​	90.6​	89.9​	86.1​	-​	90.6​	89.8​
iDSD BL - Turbo	83.0​	74.8​	63.8​	89.8​	86.6​	77.6​	-​	90.4​	86.5​
Chord Mojo	82.9​	88.8​	89.3​

Line Output Noise Floor: 105.1 dBV

 
 
Line Output SNR, THD, and THD+SNR at 1V (Preamp setting): 88.3 dBA
Line Output SNR, THD, and THD+SNR at 2V (Direct setting): 89.4 dBA

 

RMAA measurement of Line Output (Standard - Direct setting):
 

Frequency response (from 40 Hz to 15 kHz), dB	+0.01, -0.01​	Excellent​
Noise level, dB (A)	-96.6​	Excellent​
Dynamic range, dB (A)	96.6​	Excellent​
THD, %	0.0092​	Very good​
THD + Noise, dB (A)	-78.0​	Average​
IMD + Noise, %	0.011​	Very good​
Stereo crosstalk, dB	-88.4​	Excellent​
IMD at 10 kHz, %	0.011​	Very good​
General performance	​	Very good​

 
You can see that there are discrepancies between QA401 measurement and RMAA measurement using HRT LineStreamer+. This is just to show that we should not take all measurement result as absolute. Read it with a grain of salt.
 
 

Measured Headphone Output Impedance
The headphone output impedance was done comparing the peak output voltage between no load and 33 ohms load. The output impedance varies with different iEMatch switch position.

IEMatch Off: 0.34 ohms
IEMatch High Sensitivity: 4.1 ohms
IEMatch Ultra Sensitivity: 0.95 ohms
 


Headphone Amplifier Output Power
Practical observation of the power mode setting:
Eco and Normal mode have pretty close THD and SNR performance and are recommended for most IEMs and headphones.
Turbo mode generally has slightly higher noise and THD, and should only be used when the extra higher power is required. Otherwise, it is recommended to stay on Eco and Normal power mode. Working temperature at Turbo mode can be quite high, can reach up to around 51.5 degree Celsius during my test.


 

The maximum output power depends on the power mode settings and can be calculated based on the maximum output current and voltage.

Headphone Output Maximum Current Output at 15 ohms load:
Turbo mode - iEMatch Off: 306 mA  4.59V
Normal mode - iEMatch Off: 265 mA  3.97V
Eco mode - iEMatch Off: 139 mA  2.09V

Headphone Output Maximum Voltage Output at 600 ohms load:
Turbo mode - iEMatch Off: 9.71 V
Normal mode - iEMatch Off: 4.90 V
Eco mode - iEMatch Off: 2.12 V

Calculated Maximum Output Power at 600 ohms load:
Turbo mode - iEMatch Off: 157.1 mW
Normal mode - iEMatch Off: 40.2 mW
Eco mode - iEMatch Off: 7.5 mW

Calculated Maximum Output Power at 32 ohms load:
Turbo mode - iEMatch Off: 2946 mW
Normal mode - iEMatch Off: 750 mW
Eco mode - iEMatch Off: 66 mW


To observe the different level of gain between various setting, I set the power mode to Eco and the iEMatch switch set to Ultra Sensitivity, headphone output volume set to 10 mV. Then I set iEMatch switch and power mode to a different setting combination and measured the voltage output. Volume setting was not changed and was kept at the initial position. Here is the result of the measured output voltage:
​

	HO Gain @ 600Ω (mV)​
	L (US)​	M (HS)​	H (Off)​
Eco	10.0​	42.6​	157.0​
Normal	29.5​	125.0​	465.0​
Turbo	137.0​	582.0​	2,130.0​

 
 
 
XBass and 3D Matrix Plus
The XBass works very well with HD6xx. When testing HD6xx my preference is to use the XBass most of the time. It gives a natural bass boost that is just nice, not too much, and doesn't sound artificial. Midrange stays clear with XBass, no bass bleed to midrange. iDSD BL seems to be an excellent match for HD6xx.

The following is the RMAA measurement of the headphone output, comparing without and with XBass and 3D Matrix Plus.
 

Test	iFi micro iDSD BL Nor HO 600ohms​	iFi micro iDSD BL Nor HO 600ohms XBass​	iFi micro iDSD BL Nor HO 600ohms 3D​	iFi micro iDSD BL Nor HO 600ohms XBass & 3D​
Frequency response (from 40 Hz to 15 kHz), dB:​	+0.01, -0.03​	+6.02, -0.56​	+0.06, +0.00​	+5.90, -0.67​
Noise level, dB (A):​	-78.4​	-84.6​	-81.7​	-83.3​
Dynamic range, dB (A):​	78.0​	84.2​	81.4​	83.1​
THD, %:​	0.0049​	0.0045​	0.0049​	0.0035​
IMD + Noise, %:​	0.031​	0.022​	0.031​	0.018​
Stereo crosstalk, dB:​	-82.7​	-89.7​	-18.7​	-20.9​

 
 


 
 
A few interesting things to observe is the THD level is pretty much similar with and without the XBass and 3D Matrix Plus. iFi managed to enhance the bass and spatial perception without adding harmonic distortion, which is pretty impressive. As expected, stereo crosstalk is greatly affected when 3D Matrix Plus is activated.
 


Battery life and charging time.
I didn’t have enough time to do an extensive observation of the battery life. Managed only to do a single continuous playback of Pink Noise at 250 mV at 16 ohms, at Normal power setting, and the battery last to almost 8 hours. This is pretty much like driving headphone continuously at Normal power mode. At Eco power mode the battery is expected to last longer.

Below is the charging curve of iFi micro iDSD Black-Label from battery level at empty state to 100%. It requires around 6:16’ hours to fully charge the battery.
 

 

 

One interesting question would be if there is any difference in sound quality between battery powered and USB powered mode. Unfortunately, I didn’t have enough time to properly observe the difference. I only managed to do a short test, and I couldn’t hear the difference between battery powered and USB powered mode.
 
 
 
LED Indicator
Lastly is the LED indicator. I really have no idea why a simple device such as micro iCan has 3 LED indicators, and such a complex device like the micro iDSD Black-Label has only 1 LED indicator.
From the manual:
 
Power ON - LED Color:
Magenta: DSD512 22.5/24.5MHz
Blue: DSD256 11.2/12.2MHz
Cyan: DSD128/DSD64 2.8/3.1/5.6/6.2MHz
White: DXD705/768kHz
Yellow: 176/192kHz DXD352/384kHz
Green: 44/48/88/96kHz
Green(Flashing): Awaiting USB Connection
Red: Battery Low
None: Battery Empty
 
Power OFF - LED Color:
Blue*: Charging (OFF when fully-charged)
 
Are the ifi micro iDSD Black-Label users expected to memorize all of those color codes? I’m sorry to say that in my opinion that is the most ridiculously complicated LED indicator I’ve ever seen. I couldn’t find any pattern to memorize them easily. There is no indicator for the current battery level before the capacity reached to the low level. I have no idea if my WASAPI drive works since all the common PCM rates grouped into one color. Basically, it is not a user-friendly indicator.
I would suggest for a complex and feature rich device such as micro iDSD Black-Label to use at least 3 LED indicators, based on easy to remember RGB pattern as the following:
 

 
LED 1: Power indicator
Off: When the LED 1 is off when the system is switched ON, it indicates battery flat or protection circuit is activated and shutting down the system temporarily.
Red blinking: Battery powered, Battery Low, battery level <10%
Red: Battery powered, battery level 10% - <40%
Yellow: Battery powered, battery level 40% - <70%
Green: Battery powered, battery level 70% - 100%
Blue: USB powered or fully charged
Blue blinking: Battery charging in progress

LED 2: PCM sampling rate or DSD format
Off: No digital signal is present, analog input is selected
Red: 44.1 kHz PCM (Redbook)
Green: 48 kHz PCM
Blue: DSD

LED 3: PCM and DSD multiplier
Off: 1x
Red: 2x
Green: 4x
Blue: 8x
White: 16x

If the 1st power LED spaced a bit further from the 2nd and 3rd LEDs for easy identification, it would be easier to identify the 1st LED from any orientation. I hope iFi would use a more user-friendly, and useful LED indicator in their next version of micro iDSD.




iFi micro iDSD Black-Label is probably the most unique and feature rich DAC+Amp combo in its class. The Black-Label version is a proof of iFi main priority in their design philosophy, which is sound quality. The Black-Label version has similar features to the older version of micro iDSD, and all the effort and improvement is only to achieve one goal, better sound quality. And I think iFi has achieved it. Kudos to iFi!
 
 

 

 

 

 
 



Equipment used in this review:

Headphones:
Audio-Technica ATH-R70x
Audio-Technica ATH-MSR7
Beyerdynamic T1
HiFiMan HE-6
Massdrop HD6xx
Sennheiser HD800
Philips Fidelio X1
STAX SR-L300 + SRM-252S
 
In-Ear Monitors:
1964 Audio V3 (universal)
AK T8iE Mk2
Brainwavz B200
DUNU DN-2000
 
DAC and Amplifiers:
Chord Mojo
iFi micro iDSD
iFi micro iCan
Audio-Technica AT-HA22Tube
 
Measurement Equipment:
QuantAsylum QA401 - 24-bit Audio Analyzer
Owon VDS3102 - 100 MHz Digital Storage Oscilloscope
Brymen BM829s - Digital Multimeter
HRT LineStreamer+ - Analog to Digital Converter
ZKE EBD-USB+ - USB Power Meter
 
Computer & Player:
DIY Desktop PC: Gigabyte GA-H77-D3H-MVP motherboard, Intel i7-3770, 16 GB RAM, Windows 7 Home Premium 64 bit SP1.
foobar2000 v1.3.12



Some recordings used in this review:
 

 

[thatonenoob]

@earfonia, very simply, an exhaustive and impressive review.  Thanks for sharing.

 

[gr8soundz]

Great review. Most detailed I've seen yet on the iDSD.
 
Had a Micro for 2 years and your impressions of the Black Label have me thinking about getting another one.
 

 

[Cotnijoe]

This write up is a BEAST. Awesome job putting it all together!

 

[Triodemode]

Always look forward to reading your reviews earfonia.  Your beautifully lucid, detailed and concise review mirrors exactly what I am hearing when comparing the original micro iDSD to the Black Label.  Huge bonus is the comparison to the Cord Mojo, which until now has never been clearly conveyed, in which my musical preference is more in line with what the Black Label offers. I mostly listen to the DAC only using  line out driving an Audio Design labs Svetlana with NOS tubes, and am completely satisfied with the results.

 

[earfonia]

Thanks guys! iFi micro iDSD Black-Label deserves the effort 

 

[beancakes]

Very good review ! Appreciate the hard work put in to share with members . I am a newbie. Will there also review for iTube too ? My music taste is Chinese, Cantonese, english pop songs and class 95.Another worry is good music source to bring out the best in the dac. What would the headphones to pair with(not too expensive)? Thanks.

 

[yufeng23]

Kudos to you for writing this review, especially the comparisons with the Chord Mojo. I'm still contemplating whether to trade my Chord Mojo with the iFi micro iDSD BL. What's your take on this exchange? 

 

[iclickjohn]

Dang! I'm so impressed with that review.
You mentioned simbilance "from" the recording... And I always thought it was from the headphones. I hear it on some tracks but not others and usually not DSD but allways blamed my HE-560s. I did compare those moments with different settings on the iDSD, with no difference.

 

[earfonia]

Very good review ! Appreciate the hard work put in to share with members . I am a newbie. Will there also review for iTube too ? My music taste is Chinese, Cantonese, english pop songs and class 95.Another worry is good music source to bring out the best in the dac. What would the headphones to pair with(not too expensive)? Thanks.

 
I would love to review iTube if I have it. Unfortunately I don't have it. Borrowed from a friend long time ago, and quite like it.
For those pop songs, I like the ATH-R70x.
 
 

  Kudos to you for writing this review, especially the comparisons with the Chord Mojo. I'm still contemplating whether to trade my Chord Mojo with the iFi micro iDSD BL. What's your take on this exchange? 

 
I'm keeping my Mojo, great size for daily on the go DAC. Love the sound signature too, although rather different than iDSD BL. If you enjoy it, keep it. 

 
 

Dang! I'm so impressed with that review.
You mentioned simbilance "from" the recording... And I always thought it was from the headphones. I hear it on some tracks but not others and usually not DSD but allways blamed my HE-560s. I did compare those moments with different settings on the iDSD, with no difference.

 
Those settings, especially the filters, is not very audible. Need a very good recording / source to hear the difference. It's subtle.
While the power mode, Turbo you can hear that it is less refine, a tad harsher than the Normal and Eco mode.

 

[MLGrado]

Thank you for the channel imbalance measurements.  
 
 
It is very clear that there is an imbalance near the low end of the attenuation range of the silver.  That isn't news... but the measurements also show that if you followed iFi's advice and configured the controls to operate out of that range, everything is fine and dandy, which is what I have been saying to anyone with that issue that will listen.  I could understand some extremely sensitive gear still presenting a problem, but for the majority of listeners, (remember there is yet an even lower power mode that what is shown in the measurements) follow the instructions in the manual and there isn't a problem.  BUT, that said, it is very good to see that iFi has gone and addressed the issue so there shouldn't be a problem for anyone in any config.  NICE.  
 
 
SNR does seem low... I measured my Silver back when I was trying to pin down an issue a couple years ago, and in Standard and Minimum phase the SNR was at the limits of my measuring equipment at -98db.  Bit Perfect was much, much lower at around -74, but that is only at redbook sampling rates.  Bit Perfect goes up to the high nineties at 96khz and above sampling rates.  Weird?  Not so much when you understand why, but that's better left alone and/or for another day.  I am curious to see if Thorsten was able to improve the SNR of the bit-perfect filter in the Black Label.  Actually, that is the first thing I am going to check when I get my review sample  Then again, I was measuring the RCA outputs... were you measuring via that path or via the actual headphone output?  
 
 
How did you do your DSD measurements?  Did you convert a PCM test signal to DSD?  If I were to guess, that is what I would say... the ringing in a 'pure' DSD signal, if such a thing even exists, shouldn't be like that.   And that is a cautionary tale for anyone who coverts PCM to DSD expecting an improvement.  The biggest, or one of the biggest factors in quality is going to be the oversampling/reconstruction filter used before the SD modulator.  For instance, JRiver's filter rings like crazy.  In contrast, the ASIO proxy filter from Foobar/Soundforge has very little ringing and sounds amazing.  And the loss of volume when using the most extended filters in the BurrBrown DAC is very notable via ear alone.  We could go into how the DSD filter works in the BB chip, but again, that's for another day (my review??? probably
 
 
Also, one can CLEARLY see the aperture artifact in effect via the Bit-Perfect filter, which attests to the fact that despite the Sigma-Delta nature of this DAC (well, part of it) the top 6 bits of the converted signal are not sigma-delta, and the bit-perfect filter as attested to by the measurements is giving real, bit-perfect playback as advertised for that part of the signal.  
 
 
 
WONDERFUL REVIEW.  The original iDSD Micro was an impressive piece.  Some minor flaws, most were either corrected in software or could be worked around.  Sounds like the Black Label is a wonderful refinement, just as its name suggests.  

 

[earfonia]

  Thank you for the channel imbalance measurements.  
 
 
It is very clear that there is an imbalance near the low end of the attenuation range of the silver.  That isn't news... but the measurements also show that if you followed iFi's advice and configured the controls to operate out of that range, everything is fine and dandy, which is what I have been saying to anyone with that issue that will listen.  I could understand some extremely sensitive gear still presenting a problem, but for the majority of listeners, (remember there is yet an even lower power mode that what is shown in the measurements) follow the instructions in the manual and there isn't a problem.  BUT, that said, it is very good to see that iFi has gone and addressed the issue so there shouldn't be a problem for anyone in any config.  NICE. 

 
Right, no issue with recommended volume setting as long as the iEMatch switch function well. Again my problem is more with the glitchy iEMatch switch. Difficult to get good balance, have to tap and switch many time. Time to repair my micro iDSD.
 
 

 
...were you measuring via that path or via the actual headphone output?  
 
 
How did you do your DSD measurements?  Did you convert a PCM test signal to DSD?  If I were to guess, that is what I would say... the ringing in a 'pure' DSD signal, if such a thing even exists, shouldn't be like that.   And that is a cautionary tale for anyone who coverts PCM to DSD expecting an improvement.  The biggest, or one of the biggest factors in quality is going to be the oversampling/reconstruction filter used before the SD modulator.  For instance, JRiver's filter rings like crazy.  In contrast, the ASIO proxy filter from Foobar/Soundforge has very little ringing and sounds amazing.  And the loss of volume when using the most extended filters in the BurrBrown DAC is very notable via ear alone.  We could go into how the DSD filter works in the BB chip, but again, that's for another day (my review??? probably

 
When I indicated HO, it means Headphone Output. Line Output indicated as line output. I measured both HO and LO. Line Output measurement is direct connection from RCA to the BNC input of QA401.
 
Yes, excellent observation! I did convert the PCM square wave signal to DSD using Korg AudioGate 2.3.3. I never compared different converter output for square wave, so your input is very valuable to me. Please share if you have done comparisons on DSD converters. Or have you wrote it somewhere? Or yes, probably in your review Thanks!