- Joined
- Apr 28, 2007
- Posts
- 5,494
- Likes
- 1,085
As promised, here's my comparison of the following DACs:
From left to right:
CEntrance DACport LX ($299), iBasso DB2 Boomslang 2 ($299), and JDS Labs Objective DAC ($149)
Here are the chains that were compared in a blind a/b/c comparison I performed with a friend's assistance:
96 kHz / 24-bit WAV -> USB -> iBasso DB2 (balanced out) -> iBasso PB2 with LME49990 + HA5002 (balanced out) -> Toxic Cables Silver Poison -> Audeze LCD-2 Rev.1
vs.
96 kHz / 24-bit WAV -> USB -> JDS Labs ODAC (unbalanced out) -> iBasso PB2 with LME49990 + HA5002 (balanced out) -> Toxic Cables Silver Poison -> Audeze LCD-2 Rev.1
vs.
96 kHz / 24-bit WAV -> USB -> Centrance DACport LX (unbalanced out) -> iBasso PB2 with LME49990 + HA5002 (balanced out) -> Toxic Cables Silver Poison -> Audeze LCD-2 Rev.1
On the (lack of) need for Balanced Output vs. Single-Ended Output from a DAC that's driving a Balanced Amp:
You may be disappointed to see that use of the iBasso DB2 (single-ended out) was not included in the blind study, but I had separately concluded, both before the blind comparison with my friend, and later, in an admittedly brief blind comparison performed with my wife's assistance, that I can hear no difference between single-ended output and balanced output once the volume is adjusted at the amp such that a white noise WAV file measures 85.0 dB SPL at the LCD-2 ear pads. Seriously, I cannot distinguish single-ended output from the DB2 vs. balanced output from the DB2, when the volume is matched to 85.0 dB.
I considered that this finding might be unique to the DB2 or unique to the rest of my equipment chain, or my ears, but it's nevertheless consistent with Cypher Labs showing no concern (at least that I'm aware of) for releasing a balanced output version of their Algorhythm Solo (CLAS), and it's consistent with Ken Ball (the owner of ALO) having once said the following in a 6 Moons interview:
Quoting Ken Ball, of ALO:
I'm aware of several Head-Fi members who own the CLAS and a balanced amp, such as the iBasso PB2, the RSA SR-71B, or the ALO Rx-MK3-B, who have decided to bypass the highly praised DAC in the CLAS, inserting an iBasso DB2 (via Coaxial S/PDIF connection), just to achieve balanced input to their balanced amps. Spending money is fun, but I don't get it, personally.
Head-Fi member ianmedium, for one, is content to drive his SR-71B with the CLAS, directly, with single-ended input, despite his vigorous agreement that balanced output from his SR-71B to his LCD-2 rev. 1 is a must. And I agree! In my limited experience of comparing singled-ended vs. balanced input to my PB2 that runs balanced out to my LCD-2, I can't hear any difference, once the volume is matched via an SPL meter, but the difference between single-ended and balanced output from the amp is huge. (Note that I use an adapter pig-tail when attaching my balanced Silver Poison cables to the single-ended output of the PB2.)
Until I have the opportunity to audition another DAC that offers both single-ended and balanced output, I have to conclude there's no need to balance the entire chain, from source to transducers. The only potential benefit I can see for running balanced out of a DAC is to secure noise reduction over long cable runs, as might be the case with a home audio system, where the DAC cannot sit within a few centimeters of the amp's inputs.
Running balanced out of an amp is entirely different, where amps like the PB2, the SR-71B or the ALO Rx-MK3-B can use their push-pull quad designs to tightly control the motion of a transducer's mass. I liken this to a pair of swinging doors at the entrance to an old-West saloon, as seen in many Western movies. Instead of only one man trying to control the motion of both doors, accelerating them to get them moving, then trying to dampen their motion as necessary, a quad amp places a man on both sides of each door. One man pushes on one door while his partner operates precisely 180 degrees out of phase to pull simultaneously. The same thing is happening at the other door (at the other channel), overcoming the inertia of the moving parts with twice the power and much improved control. So consider this: A DAC doesn't have to worry about controlling the motion of any moving parts that have mass and inertia. Only an amp faces that challenge. Balanced is the way to go with amps, for power and control. I just can't imagine a corollary with DACs. I can't hear one either, at least not with the DB2.
In any case, even those who are using DB2's between their CLAS and balanced amp, should be happy to hear my opinion that the PB2 does not sound better single-ended than balanced.
But I have to wonder if they have used a SPL meter and white noise to match volumes in a blind study...
On matching white noise volumes to 85.0 dB SPL - an unexpected discovery:
We used an iPhone app SPL meter (by JL Audio) to synchronize the levels to 85.0 dB, as measured with a lapel mic that was sandwiched between the ear pads of my LCD-2, which was lying on the table as we played a white noise FLAC file (Track 56, as downloaded from Michael Knowles' page, then converted to FLAC.)
To achieve 85.0 dB SPL at the face-to-face ear cups of the LCD-2, we had to adjust the volume control as shown in the following table (on my iBasso PB2, running LME49990's in L/R and HA5002 buffers, balanced out via Toxic Cables' Silver Poison to the LCD-2):
JDS Labs Objective DAC ---------------------------- 5 and 1/3 dots = 85.0 dB
CEntrance DACport LX ------------------------------ 5 and 1/2 dots = 85.0 dB
iBasso DB2 (balanced) ------------------------------ 7 and 2/3 dots = 85.0 dB
iBasso DB2 (singled-ended) ------------------------- 8 and 1/2 dots = 85.0 dB
We performed these SPL measurements with volume matching twice - first, in advance of the blind study, and then, right afterwards, to reassure ourselves that the volume settings we had used for the blind study (as shown in the table, above) all produced a SPL of 85.0 dB when playing the white noise file. Even if the JL Audio SPL app, in combination with the lapel mic we used, is inaccurate (too high or too low by a few dB), I can at least say that the measurements were consistent, and thus, the volume levels during the blind study were well-matched, whether they were actually all at 85.0 dB, or all at something other than 85.0 dB.
Obviously, the ODAC is somehow "louder" than the competition, but the LX is close on its heels, requiring 5 and 1/2 dots vs. only 5 and 1/3 dots on the PB2's volume control to achieve the same SPL at the LCD-2. The unexpected discovery, here, is that even though the manufacturer's specifications for all three DACs say that they can deliver an output of 2.0 Vrms (with the exception of the single-ended DB2 output spec'd at 1.0 Vrms), for some reason, the DB2 required a lot more amplification than either the ODAC or the LX to achieve the same SPL at the LCD-2. That's not good, as it eats into your amp's available headroom.
And for the record, when leaving the amp's volume control at any fixed position, my ears were unable to detect any difference in volume when switching between the ODAC and the LX, but the difference in volume when switching from the ODAC to the DB2 or from the LX to the DB2 was dramatic - a big reduction in audible volume could be heard when switching to the DB2, with an additional obvious drop in volume when switching from DB2 balanced to DB2 single-ended (but that's to be expected, given that the specifications indicate only 1.0 Vrms for the DB2 singled-ended).
I can only conclude that the iBasso DB2's Vrms was well short of meeting iBasso's published specs (certainly for balanced output, where it should have delivered the same SPL at the LCD-2 as the LX and ODAC, for any given volume setting at the amp.)
For what it's worth, by way of comparison, the DB2 balanced out is much "louder" into my PB2 than my Sansa Clip+, but NwAvGuy once measured the output of a Sansa Clip+ at only 0.5 Vrms - nothing close to 2.0 Vrms.
And in case anyone is wondering if the DB2's batteries weren't fully charged - they were - and I came up with the same 7 and 2/3 dots required at my PB2 volume control when the DB2 was plugged into the wall adapter with the charger turned on while playing the white noise file. Maybe I was shipped a defective DB2, but it gave every appearance of working as designed, otherwise.
How we conducted the blind comparison:
We repeated the steps I'm about to describe for each of the following six tracks - which I selected after a lot of listening to all three DACs, expressly because I'm very familiar with these tracks and felt they would give me the best chance of blindly distinguishing the three volume-matched DACs from each other:
Now, some might say the blind study was rigged, given that I did not select these tracks at random, but if it's true that any two of these DACs sound identical when volume matched, I'd have no ability to discern one from the other, no matter which tracks I played, right?
One track at a time, my friend used a random number generator (in Excel) to determine the sequence in which each of the three DACs would be presented to me. Thus, for some tracks, I heard the ODAC first, then the PB2, then the LX, for example, where for another track, I might have heard the PB2 first, then the ODAC, and then the LX. You get the idea. The sequence was randomly determined for each track.
For each combination of track and DAC, I would go to another room and close the door - well clear of any ability to hear him re-cabling DACs. Then, when he was ready, having queued the track and cabled the randomly selected DAC, he would knock on the wall to summon me. (This was great fun, by the way!)
I would back into the room with my eyes closed and feel my way along the wall to back into a chair next to the table where I could pick up the LCD-2 and put them on.
He would play the track until I signaled that I'd heard enough, at which point, I would leave the room with my eyes closed, to repeat the cycle.
Only after hearing all three randomly-sequenced and volume-matched DACs for a given track, would I attempt to identify the sequence in which I had heard each DAC. If you can imagine, this is a fairly daunting task, given that I'm not just switching back and forth at will between two DACs while the track plays. Instead, I listened to a minute or two of a given track, only once for each randomly sequenced DAC.
Under these conditions, aural memory comes into play, but I can tell you that I found it surprisingly easy to distinguish the three DACs by their sonic traits.
I correctly identified the CEntrance DACport LX with four out of six tracks.
I correctly identified the iBasso DB2 (balanced) with five out of six tracks.
I correctly identified the JDS Labs Objective DAC with four out of six tracks.
That's 13 out of 18 combinations correctly identified - for a score of 72.2%.
That could certainly happen by chance, but please consider this: For one of the six tracks (Bel Canto - Walking Wheel), I got all three of them wrong. For all five of the remaining tracks, I managed to correctly identify the iBasso PB2, five out of five attempts. I have no way of proving this, but I told my friend at the time, that I was shocked at having incorrectly identified all three DACs when listening to (Bel Canto - Walking Wheel) because the PB2 frankly sticks out like a sore thumb in sonic signature compared to the LX and ODAC, which are far more difficult to distinguish from each other. I had communicated at the time, that I had second-guessed my original identifications, wrestling over a difference between the LX and ODAC, and that I genuinely believe that I had simply suffered a lapse in concentration in trying to remember which was which after hearing them all in sequence. I know this sounds "fishy," but it's not as if I decided to throw out the results for that track altogether. I'm just trying to say that if there's any one DAC of the three that I can nail every time, it's the DB2 - for its dark and smooth demeanor, relative to the ODAC and LX, which share a far livelier, brighter, and more analytical personality - but even they are not identical in signature.
If anyone is willing to humor me, we could toss out the track where I second-guessed myself, and you'd see the results change dramatically:
I correctly identified the CEntrance DACport LX with four out of five tracks.
I correctly identified the iBasso DB2 (balanced) with five out of five tracks.
I correctly identified the JDS Labs Objective DAC with four out of five tracks.
That's 13 out of 15 combinations correctly identified - for a score of 86.7%.
Either way, I can at least say that I was able to blindly distinguish all three DACs from each other, a majority of the time.
What was I hearing that allowed me to distinguish each volume-matched DAC from the other two?
I'll start with my favorite of the three DACs (yes, there's the punchline of this entire comparison) - the CEntrance DACport LX:
In short, the Centrance DACport LX provides the best sound of any source I've spent considerable time with to date (an admittedly short list, that includes the DB2, the ODAC, the Sony PCM-M10, the April Music DA100, the RWA iMod, and the Rockboxed Sansa Clip+).
The DACport LX delivers an articulate, natural rendering of just about everything I play with it - highly resolving with very realistic transients, creating a very believable ambiance. I'm convinced that it's impossible to have good imaging in the absence of all the micro-details that reside at low signal levels - the small echoes and reflections that define the space in which the voices and instruments reside. The DACport delivers these details in a way that's very believable, for lack of a better word. A pitch-black noise floor is also necessary to render voices and instruments with good separation and presence, and the DACport LX delivers in this regard, as well. It's analytical, yes, but never harsh or etched in the highs, not in the least bit fatiguing, yet not too smooth, either - a trait that I know some people like, especially those who are into tube gear, but which I don't. In that sense, I suppose you could say the DACport LX is somewhat unforgiving of poor recordings, but that's what I prefer. You have to listen to other DACs side-by-side with the DACport LX, to realize just how much it's NOT doing to the music! Lastly, it's wonderfully uncolored - very neutral and transparent in every way - while offering lots of control and detail in the bass. I think the LX would be appealing to anyone who loves Jan Meier's amps and DACs - where uninhibited transparency is the goal.
Edit (applied a few hours later): I neglected to add that with the DACport LX, I'm able to listen at much lower volume levels than with other sources, yet still feel as if I'm hearing everything I could want to hear, across the entire frequency spectrum - with no loss of detail or any other desirable trait. I'm fascinated by this, because with the iBasso DB2, for example, I feel compelled to turn up the volume to levels higher than 85.0 dB, before I can begin hearing everything a track has to offer. The DB2 just sounds its best to me at levels that could cause hearing damage. I think I'm onto something here, and having made this observation, from now on, I plan to use listening at low volume levels as a final test of overall quality in any given chain I find myself auditioning.
Next up, my second favorite of the three DACs - the JDS Labs Objective DAC:
To my ears, with the equipment chain described, the JDS Labs Objective DAC is difficult to distinguish from the CEntrance DACport LX, with most tracks, most of the time. There are a lot of little differences, but they are subtle. The two DACs have a lot in common, but where the DACport LX is never harsh or etched in the highs and never fatiguing, the highly resolving, otherwise very neutral and transparent Objective DAC throws little fits from time to time that are very annoying. This momentary aberrant behavior seems to occur whenever there's a particularly loud high frequency signal during an instant of high-complexity, where the DAC has a lot of work to do to sort everything out. It's as if the ODAC just gives up, momentarily, and renders those instantaneous high frequencies like two blocks of styrofoam being rubbed together - to produce a very short in duration, but very loud distortion of information that the LX and the PB2 have no difficulty rendering cleanly.
When I first started listening to the ODAC, several days before we did the blind comparison, I was very impressed with it's amazing detail and neutrality, and the great soundstage and imaging that I've come to believe is, in part, dependent on the ability to hear micro-details. But I started noticing that where it sounded absolutely perfect with some tracks, really complex signals, like those heard in Daft Punk - Tron Legacy (End Titles), seemed to just overwhelm it - not throughout the entire track, but just here and there. The thing is, once I heard it, I started listening for it - and then I started to hear it more readily in several different tracks. This very infrequent but nevertheless loud, short-duration screeching sound began to haunt me. This is a big FAIL in terms of transparency. Switching to the DACport LX or the DB2, with all else remaining the same, convinced me that this was not coming from the recordings, but rather, from the ODAC itself. Yuck. What a shame. It's like smoking a fine Cuban cigar that has tiny fire-crackers hidden throughout the tobacco. I don't smoke cigars, cigarettes, or anything else, but I do think I've come up with a good analogy! The ODAC is great until, very briefly, it's not, and then it's great again. How frustrating! Close, but no cigar, as they say!
Still, I want to make it clear that everything in my chain is very revealing and that the problem I experienced with the ODAC is not necessarily a show-stopper for everyone else. I think the ODAC would be a great match, as is, with less resolving and darker amps, like the Schiit Lyr, for example, where the Lyr's smoothness would go a long way to taking the edge off the ODAC's infrequent anomalies.
During the blind study, to distinguish the ODAC from the DACport LX, I just listened for the momentary outbursts of distortion in the high frequencies - but I can tell you that it was much more difficult to distinguish the DACport LX from the ODAC than distinguishing either of them from the PB2 - the ODAC and the DACport LX are that similar in sound signature. The DB2 is another critter, entirely.
Lastly, my least favorite of the three DACs - the iBasso DB2 Boomslang 2:
OK, first, I realize that saying something is "my least favorite," is not at all flattering, but please don't overlook that I'm making a subjective assessment and, as I've pointed out above, I am not a fan of dark and smooth. I'm so biased against dark and smooth sounding gear that I'm only mildly interested in ever auditioning any of Ray Samuel's amps, including the SR-71B - because of all the reviews I've read that describe the RSA "house sound" as being dark. Keep in mind, too, that I'm using the LCD-2, which is known for its shelved high frequencies. Some people like dark and smooth - including some very respected reviewers with a lot more experience than I'll probably ever have - Sklyab, for instance. I don't think I'd be misquoting Skylab in the least to say that he loves the sound of his LCD-2 in combination with his very expensive Leben tube amp, but HeadphoneAddict, another Head-Fi member I greatly respect, much prefers the brighter, livelier, sound of the LCD-2 in combination with his CEntrance DACmini CX. My tastes are definitely more aligned with HeadphoneAddict than with Skylab.
Note, too, that I'm a big fan of the iBasso PB2 Pelican. The DB2, however, just isn't my cup of tea, at least not when using the USB input. Unfortunately, I had to order, audition, and return a DB2, to learn this (taking advantage of their very generous 14-day no-questions-asked return policy), despite having searched dilligently for reviews that describe the DB2's traits, and asking DB2 owners for their opinion. I've read, but not personally experienced, that the DB2's coaxial S/PDIF input sounds "better" - but the comment I'm referencing offered no additional qualifications.
As is the case with a lot of gear, many people just don't possess alternates with which to make direct comparisons, but I do find it noteworthy that there is a conspicuous lack of DB2 reviews out there, at Head-Fi, or any other sites, for that matter. In general, it's hard to find anyone raving about the DB2's virtues. The strongest affirmation of the DB2 of which I'm aware is its use by several CLAS owners who are convinced that the DB2 sounds better, inserted into the chain between the CLAS and an SR-71B, for example, than using the DAC within the CLAS single-ended to the SR-71B. That's really hard to ignore - it pushed me into deciding that I needed to hear the DB2 for myself.
Here are my impressions of the iBasso DB2 (as heard with my ears, my gear): Relative to the ODAC and LX, the DB2 is just not as "lively," in my opinion. It sounds dynamically weak - as if it is compressing the inherent dynamic range of the recording. It's also dark and and very slightly lower-resolving than the other two DACs - but I admit this perception might just be due to a difference in brightness. It is very smooth and clean sounding - absolutely silky smooth - making my entire chain sound much like what I would expect from a good tube amp - which is not a problem, at all, for many people who prefer that sound. Again, it's just not my cup of tea. The PB2 might work well with an equipment chain that is otherwise too bright or analytical for the user's tastes - maybe to sweeten the HD800, for example - but for my tastes, it doesn't work well with a transparent amp and the LCD-2 Rev.1, especially with the LCD-2's shelved highs.
I have to reiterate that after adjusting for the difference in volume to achieve 85.0 dB, I cannot detect any difference between using the DB2 single-ended out to the PB2 vs. balanced-out to the PB2. In both cases, as with testing the other DACs, I'm using the PB2 balanced out to my LCD-2 via Toxic Cables' Silver Poison - which are known for being bright relative to the stock Audeze ADZ-5 cables, but they're not bright enough to compensate the darkness of the DB2.
Other than my personal dislike for the DB2's smooth, dark sound, I'm really bothered by its lack of dynamics - it just does't have as much punch or slam as the two other DACs. Again, hearing them side by side makes this obvious.
Other considerations that should affect a purchase decision:
The CEntrance DACport LX has a peculiar shape, with input and output at opposite ends (unappealing), and gets quite warm during play (due to its Class A design). But it's built like a tank on the outside, and I'm convinced that it's the most sophisticated design of the three DACs I've reviewed - a lot for the money, even at $299.
The JDS Labs Objective DAC weighs very little, is very small, appears to be indestructible in build quality, has both of its jacks at one end, instead of at opposite ends of the case, and, amazingly, it sells for only $149.00. I find all of those factors to be very appealing!
The iBasso DB2 offers coaxial and TOSLINK S/PDIF inputs in addition to USB - making it far more flexible for use as a desktop DAC. It's designed with two, internal lithium-ion batteries, divorcing it from any reliance on obtaining power from the PC's USB port. This means that it's likely a much better candidate than the other two DACs for use with an iPad, where the USB port might not provide enough current to operate a DAC. The iBasso DB2 includes far more accessories than the other DACs, as well - an optical cable, spare screws for the case, a second pair of Low Pass Filter op-amps (which made no difference that I could detect), a soft pouch, a budget 3.5mm TRS interconnect cable, and a couple of other tidbits. Given the greater feature list of the DB2, it too, should be considered a great value at $299.
Lastly, I should probably mention that I didn't have any need to install drivers (or troubleshoot in any way) installation of these three DACs,for use with my Windows 7 laptop. In fact, having installed all three of them, I didn't suffer any conflicts at all when unplugging one DAC from the USB port and switching to another. I did not attempt to mount more than one USB DAC at a time.
The End
Update: Corrected an occurrence of "DB2" to read "PB2" instead.
From left to right:
CEntrance DACport LX ($299), iBasso DB2 Boomslang 2 ($299), and JDS Labs Objective DAC ($149)
Here are the chains that were compared in a blind a/b/c comparison I performed with a friend's assistance:
96 kHz / 24-bit WAV -> USB -> iBasso DB2 (balanced out) -> iBasso PB2 with LME49990 + HA5002 (balanced out) -> Toxic Cables Silver Poison -> Audeze LCD-2 Rev.1
vs.
96 kHz / 24-bit WAV -> USB -> JDS Labs ODAC (unbalanced out) -> iBasso PB2 with LME49990 + HA5002 (balanced out) -> Toxic Cables Silver Poison -> Audeze LCD-2 Rev.1
vs.
96 kHz / 24-bit WAV -> USB -> Centrance DACport LX (unbalanced out) -> iBasso PB2 with LME49990 + HA5002 (balanced out) -> Toxic Cables Silver Poison -> Audeze LCD-2 Rev.1
On the (lack of) need for Balanced Output vs. Single-Ended Output from a DAC that's driving a Balanced Amp:
You may be disappointed to see that use of the iBasso DB2 (single-ended out) was not included in the blind study, but I had separately concluded, both before the blind comparison with my friend, and later, in an admittedly brief blind comparison performed with my wife's assistance, that I can hear no difference between single-ended output and balanced output once the volume is adjusted at the amp such that a white noise WAV file measures 85.0 dB SPL at the LCD-2 ear pads. Seriously, I cannot distinguish single-ended output from the DB2 vs. balanced output from the DB2, when the volume is matched to 85.0 dB.
I considered that this finding might be unique to the DB2 or unique to the rest of my equipment chain, or my ears, but it's nevertheless consistent with Cypher Labs showing no concern (at least that I'm aware of) for releasing a balanced output version of their Algorhythm Solo (CLAS), and it's consistent with Ken Ball (the owner of ALO) having once said the following in a 6 Moons interview:
Quoting Ken Ball, of ALO:
I'm aware of several Head-Fi members who own the CLAS and a balanced amp, such as the iBasso PB2, the RSA SR-71B, or the ALO Rx-MK3-B, who have decided to bypass the highly praised DAC in the CLAS, inserting an iBasso DB2 (via Coaxial S/PDIF connection), just to achieve balanced input to their balanced amps. Spending money is fun, but I don't get it, personally.
Head-Fi member ianmedium, for one, is content to drive his SR-71B with the CLAS, directly, with single-ended input, despite his vigorous agreement that balanced output from his SR-71B to his LCD-2 rev. 1 is a must. And I agree! In my limited experience of comparing singled-ended vs. balanced input to my PB2 that runs balanced out to my LCD-2, I can't hear any difference, once the volume is matched via an SPL meter, but the difference between single-ended and balanced output from the amp is huge. (Note that I use an adapter pig-tail when attaching my balanced Silver Poison cables to the single-ended output of the PB2.)
Until I have the opportunity to audition another DAC that offers both single-ended and balanced output, I have to conclude there's no need to balance the entire chain, from source to transducers. The only potential benefit I can see for running balanced out of a DAC is to secure noise reduction over long cable runs, as might be the case with a home audio system, where the DAC cannot sit within a few centimeters of the amp's inputs.
Running balanced out of an amp is entirely different, where amps like the PB2, the SR-71B or the ALO Rx-MK3-B can use their push-pull quad designs to tightly control the motion of a transducer's mass. I liken this to a pair of swinging doors at the entrance to an old-West saloon, as seen in many Western movies. Instead of only one man trying to control the motion of both doors, accelerating them to get them moving, then trying to dampen their motion as necessary, a quad amp places a man on both sides of each door. One man pushes on one door while his partner operates precisely 180 degrees out of phase to pull simultaneously. The same thing is happening at the other door (at the other channel), overcoming the inertia of the moving parts with twice the power and much improved control. So consider this: A DAC doesn't have to worry about controlling the motion of any moving parts that have mass and inertia. Only an amp faces that challenge. Balanced is the way to go with amps, for power and control. I just can't imagine a corollary with DACs. I can't hear one either, at least not with the DB2.
In any case, even those who are using DB2's between their CLAS and balanced amp, should be happy to hear my opinion that the PB2 does not sound better single-ended than balanced.
On matching white noise volumes to 85.0 dB SPL - an unexpected discovery:
We used an iPhone app SPL meter (by JL Audio) to synchronize the levels to 85.0 dB, as measured with a lapel mic that was sandwiched between the ear pads of my LCD-2, which was lying on the table as we played a white noise FLAC file (Track 56, as downloaded from Michael Knowles' page, then converted to FLAC.)
To achieve 85.0 dB SPL at the face-to-face ear cups of the LCD-2, we had to adjust the volume control as shown in the following table (on my iBasso PB2, running LME49990's in L/R and HA5002 buffers, balanced out via Toxic Cables' Silver Poison to the LCD-2):
JDS Labs Objective DAC ---------------------------- 5 and 1/3 dots = 85.0 dB
CEntrance DACport LX ------------------------------ 5 and 1/2 dots = 85.0 dB
iBasso DB2 (balanced) ------------------------------ 7 and 2/3 dots = 85.0 dB
iBasso DB2 (singled-ended) ------------------------- 8 and 1/2 dots = 85.0 dB
We performed these SPL measurements with volume matching twice - first, in advance of the blind study, and then, right afterwards, to reassure ourselves that the volume settings we had used for the blind study (as shown in the table, above) all produced a SPL of 85.0 dB when playing the white noise file. Even if the JL Audio SPL app, in combination with the lapel mic we used, is inaccurate (too high or too low by a few dB), I can at least say that the measurements were consistent, and thus, the volume levels during the blind study were well-matched, whether they were actually all at 85.0 dB, or all at something other than 85.0 dB.
Obviously, the ODAC is somehow "louder" than the competition, but the LX is close on its heels, requiring 5 and 1/2 dots vs. only 5 and 1/3 dots on the PB2's volume control to achieve the same SPL at the LCD-2. The unexpected discovery, here, is that even though the manufacturer's specifications for all three DACs say that they can deliver an output of 2.0 Vrms (with the exception of the single-ended DB2 output spec'd at 1.0 Vrms), for some reason, the DB2 required a lot more amplification than either the ODAC or the LX to achieve the same SPL at the LCD-2. That's not good, as it eats into your amp's available headroom.
And for the record, when leaving the amp's volume control at any fixed position, my ears were unable to detect any difference in volume when switching between the ODAC and the LX, but the difference in volume when switching from the ODAC to the DB2 or from the LX to the DB2 was dramatic - a big reduction in audible volume could be heard when switching to the DB2, with an additional obvious drop in volume when switching from DB2 balanced to DB2 single-ended (but that's to be expected, given that the specifications indicate only 1.0 Vrms for the DB2 singled-ended).
I can only conclude that the iBasso DB2's Vrms was well short of meeting iBasso's published specs (certainly for balanced output, where it should have delivered the same SPL at the LCD-2 as the LX and ODAC, for any given volume setting at the amp.)
For what it's worth, by way of comparison, the DB2 balanced out is much "louder" into my PB2 than my Sansa Clip+, but NwAvGuy once measured the output of a Sansa Clip+ at only 0.5 Vrms - nothing close to 2.0 Vrms.
And in case anyone is wondering if the DB2's batteries weren't fully charged - they were - and I came up with the same 7 and 2/3 dots required at my PB2 volume control when the DB2 was plugged into the wall adapter with the charger turned on while playing the white noise file. Maybe I was shipped a defective DB2, but it gave every appearance of working as designed, otherwise.
How we conducted the blind comparison:
We repeated the steps I'm about to describe for each of the following six tracks - which I selected after a lot of listening to all three DACs, expressly because I'm very familiar with these tracks and felt they would give me the best chance of blindly distinguishing the three volume-matched DACs from each other:
| Team Sleep - The Passportal |
| Bel Canto - Walking Wheel (Live) |
| Daft Punk - Tron Legacy (End Titles) |
| Hugo Montenegro - Boo-Qui-Woo-Qui |
| Shapeshifters - Lola's Theme (Extended Mix) |
| Eagles - Hotel California |
Now, some might say the blind study was rigged, given that I did not select these tracks at random, but if it's true that any two of these DACs sound identical when volume matched, I'd have no ability to discern one from the other, no matter which tracks I played, right?
One track at a time, my friend used a random number generator (in Excel) to determine the sequence in which each of the three DACs would be presented to me. Thus, for some tracks, I heard the ODAC first, then the PB2, then the LX, for example, where for another track, I might have heard the PB2 first, then the ODAC, and then the LX. You get the idea. The sequence was randomly determined for each track.
For each combination of track and DAC, I would go to another room and close the door - well clear of any ability to hear him re-cabling DACs. Then, when he was ready, having queued the track and cabled the randomly selected DAC, he would knock on the wall to summon me. (This was great fun, by the way!)
I would back into the room with my eyes closed and feel my way along the wall to back into a chair next to the table where I could pick up the LCD-2 and put them on.
He would play the track until I signaled that I'd heard enough, at which point, I would leave the room with my eyes closed, to repeat the cycle.
Only after hearing all three randomly-sequenced and volume-matched DACs for a given track, would I attempt to identify the sequence in which I had heard each DAC. If you can imagine, this is a fairly daunting task, given that I'm not just switching back and forth at will between two DACs while the track plays. Instead, I listened to a minute or two of a given track, only once for each randomly sequenced DAC.
Under these conditions, aural memory comes into play, but I can tell you that I found it surprisingly easy to distinguish the three DACs by their sonic traits.
I correctly identified the CEntrance DACport LX with four out of six tracks.
I correctly identified the iBasso DB2 (balanced) with five out of six tracks.
I correctly identified the JDS Labs Objective DAC with four out of six tracks.
That's 13 out of 18 combinations correctly identified - for a score of 72.2%.
That could certainly happen by chance, but please consider this: For one of the six tracks (Bel Canto - Walking Wheel), I got all three of them wrong. For all five of the remaining tracks, I managed to correctly identify the iBasso PB2, five out of five attempts. I have no way of proving this, but I told my friend at the time, that I was shocked at having incorrectly identified all three DACs when listening to (Bel Canto - Walking Wheel) because the PB2 frankly sticks out like a sore thumb in sonic signature compared to the LX and ODAC, which are far more difficult to distinguish from each other. I had communicated at the time, that I had second-guessed my original identifications, wrestling over a difference between the LX and ODAC, and that I genuinely believe that I had simply suffered a lapse in concentration in trying to remember which was which after hearing them all in sequence. I know this sounds "fishy," but it's not as if I decided to throw out the results for that track altogether. I'm just trying to say that if there's any one DAC of the three that I can nail every time, it's the DB2 - for its dark and smooth demeanor, relative to the ODAC and LX, which share a far livelier, brighter, and more analytical personality - but even they are not identical in signature.
If anyone is willing to humor me, we could toss out the track where I second-guessed myself, and you'd see the results change dramatically:
I correctly identified the CEntrance DACport LX with four out of five tracks.
I correctly identified the iBasso DB2 (balanced) with five out of five tracks.
I correctly identified the JDS Labs Objective DAC with four out of five tracks.
That's 13 out of 15 combinations correctly identified - for a score of 86.7%.
Either way, I can at least say that I was able to blindly distinguish all three DACs from each other, a majority of the time.
What was I hearing that allowed me to distinguish each volume-matched DAC from the other two?
I'll start with my favorite of the three DACs (yes, there's the punchline of this entire comparison) - the CEntrance DACport LX:
In short, the Centrance DACport LX provides the best sound of any source I've spent considerable time with to date (an admittedly short list, that includes the DB2, the ODAC, the Sony PCM-M10, the April Music DA100, the RWA iMod, and the Rockboxed Sansa Clip+).
The DACport LX delivers an articulate, natural rendering of just about everything I play with it - highly resolving with very realistic transients, creating a very believable ambiance. I'm convinced that it's impossible to have good imaging in the absence of all the micro-details that reside at low signal levels - the small echoes and reflections that define the space in which the voices and instruments reside. The DACport delivers these details in a way that's very believable, for lack of a better word. A pitch-black noise floor is also necessary to render voices and instruments with good separation and presence, and the DACport LX delivers in this regard, as well. It's analytical, yes, but never harsh or etched in the highs, not in the least bit fatiguing, yet not too smooth, either - a trait that I know some people like, especially those who are into tube gear, but which I don't. In that sense, I suppose you could say the DACport LX is somewhat unforgiving of poor recordings, but that's what I prefer. You have to listen to other DACs side-by-side with the DACport LX, to realize just how much it's NOT doing to the music! Lastly, it's wonderfully uncolored - very neutral and transparent in every way - while offering lots of control and detail in the bass. I think the LX would be appealing to anyone who loves Jan Meier's amps and DACs - where uninhibited transparency is the goal.
Edit (applied a few hours later): I neglected to add that with the DACport LX, I'm able to listen at much lower volume levels than with other sources, yet still feel as if I'm hearing everything I could want to hear, across the entire frequency spectrum - with no loss of detail or any other desirable trait. I'm fascinated by this, because with the iBasso DB2, for example, I feel compelled to turn up the volume to levels higher than 85.0 dB, before I can begin hearing everything a track has to offer. The DB2 just sounds its best to me at levels that could cause hearing damage. I think I'm onto something here, and having made this observation, from now on, I plan to use listening at low volume levels as a final test of overall quality in any given chain I find myself auditioning.
Next up, my second favorite of the three DACs - the JDS Labs Objective DAC:
To my ears, with the equipment chain described, the JDS Labs Objective DAC is difficult to distinguish from the CEntrance DACport LX, with most tracks, most of the time. There are a lot of little differences, but they are subtle. The two DACs have a lot in common, but where the DACport LX is never harsh or etched in the highs and never fatiguing, the highly resolving, otherwise very neutral and transparent Objective DAC throws little fits from time to time that are very annoying. This momentary aberrant behavior seems to occur whenever there's a particularly loud high frequency signal during an instant of high-complexity, where the DAC has a lot of work to do to sort everything out. It's as if the ODAC just gives up, momentarily, and renders those instantaneous high frequencies like two blocks of styrofoam being rubbed together - to produce a very short in duration, but very loud distortion of information that the LX and the PB2 have no difficulty rendering cleanly.
When I first started listening to the ODAC, several days before we did the blind comparison, I was very impressed with it's amazing detail and neutrality, and the great soundstage and imaging that I've come to believe is, in part, dependent on the ability to hear micro-details. But I started noticing that where it sounded absolutely perfect with some tracks, really complex signals, like those heard in Daft Punk - Tron Legacy (End Titles), seemed to just overwhelm it - not throughout the entire track, but just here and there. The thing is, once I heard it, I started listening for it - and then I started to hear it more readily in several different tracks. This very infrequent but nevertheless loud, short-duration screeching sound began to haunt me. This is a big FAIL in terms of transparency. Switching to the DACport LX or the DB2, with all else remaining the same, convinced me that this was not coming from the recordings, but rather, from the ODAC itself. Yuck. What a shame. It's like smoking a fine Cuban cigar that has tiny fire-crackers hidden throughout the tobacco. I don't smoke cigars, cigarettes, or anything else, but I do think I've come up with a good analogy! The ODAC is great until, very briefly, it's not, and then it's great again. How frustrating! Close, but no cigar, as they say!
Still, I want to make it clear that everything in my chain is very revealing and that the problem I experienced with the ODAC is not necessarily a show-stopper for everyone else. I think the ODAC would be a great match, as is, with less resolving and darker amps, like the Schiit Lyr, for example, where the Lyr's smoothness would go a long way to taking the edge off the ODAC's infrequent anomalies.
During the blind study, to distinguish the ODAC from the DACport LX, I just listened for the momentary outbursts of distortion in the high frequencies - but I can tell you that it was much more difficult to distinguish the DACport LX from the ODAC than distinguishing either of them from the PB2 - the ODAC and the DACport LX are that similar in sound signature. The DB2 is another critter, entirely.
Lastly, my least favorite of the three DACs - the iBasso DB2 Boomslang 2:
OK, first, I realize that saying something is "my least favorite," is not at all flattering, but please don't overlook that I'm making a subjective assessment and, as I've pointed out above, I am not a fan of dark and smooth. I'm so biased against dark and smooth sounding gear that I'm only mildly interested in ever auditioning any of Ray Samuel's amps, including the SR-71B - because of all the reviews I've read that describe the RSA "house sound" as being dark. Keep in mind, too, that I'm using the LCD-2, which is known for its shelved high frequencies. Some people like dark and smooth - including some very respected reviewers with a lot more experience than I'll probably ever have - Sklyab, for instance. I don't think I'd be misquoting Skylab in the least to say that he loves the sound of his LCD-2 in combination with his very expensive Leben tube amp, but HeadphoneAddict, another Head-Fi member I greatly respect, much prefers the brighter, livelier, sound of the LCD-2 in combination with his CEntrance DACmini CX. My tastes are definitely more aligned with HeadphoneAddict than with Skylab.
Note, too, that I'm a big fan of the iBasso PB2 Pelican. The DB2, however, just isn't my cup of tea, at least not when using the USB input. Unfortunately, I had to order, audition, and return a DB2, to learn this (taking advantage of their very generous 14-day no-questions-asked return policy), despite having searched dilligently for reviews that describe the DB2's traits, and asking DB2 owners for their opinion. I've read, but not personally experienced, that the DB2's coaxial S/PDIF input sounds "better" - but the comment I'm referencing offered no additional qualifications.
As is the case with a lot of gear, many people just don't possess alternates with which to make direct comparisons, but I do find it noteworthy that there is a conspicuous lack of DB2 reviews out there, at Head-Fi, or any other sites, for that matter. In general, it's hard to find anyone raving about the DB2's virtues. The strongest affirmation of the DB2 of which I'm aware is its use by several CLAS owners who are convinced that the DB2 sounds better, inserted into the chain between the CLAS and an SR-71B, for example, than using the DAC within the CLAS single-ended to the SR-71B. That's really hard to ignore - it pushed me into deciding that I needed to hear the DB2 for myself.
Here are my impressions of the iBasso DB2 (as heard with my ears, my gear): Relative to the ODAC and LX, the DB2 is just not as "lively," in my opinion. It sounds dynamically weak - as if it is compressing the inherent dynamic range of the recording. It's also dark and and very slightly lower-resolving than the other two DACs - but I admit this perception might just be due to a difference in brightness. It is very smooth and clean sounding - absolutely silky smooth - making my entire chain sound much like what I would expect from a good tube amp - which is not a problem, at all, for many people who prefer that sound. Again, it's just not my cup of tea. The PB2 might work well with an equipment chain that is otherwise too bright or analytical for the user's tastes - maybe to sweeten the HD800, for example - but for my tastes, it doesn't work well with a transparent amp and the LCD-2 Rev.1, especially with the LCD-2's shelved highs.
I have to reiterate that after adjusting for the difference in volume to achieve 85.0 dB, I cannot detect any difference between using the DB2 single-ended out to the PB2 vs. balanced-out to the PB2. In both cases, as with testing the other DACs, I'm using the PB2 balanced out to my LCD-2 via Toxic Cables' Silver Poison - which are known for being bright relative to the stock Audeze ADZ-5 cables, but they're not bright enough to compensate the darkness of the DB2.
Other than my personal dislike for the DB2's smooth, dark sound, I'm really bothered by its lack of dynamics - it just does't have as much punch or slam as the two other DACs. Again, hearing them side by side makes this obvious.
Other considerations that should affect a purchase decision:
The CEntrance DACport LX has a peculiar shape, with input and output at opposite ends (unappealing), and gets quite warm during play (due to its Class A design). But it's built like a tank on the outside, and I'm convinced that it's the most sophisticated design of the three DACs I've reviewed - a lot for the money, even at $299.
The JDS Labs Objective DAC weighs very little, is very small, appears to be indestructible in build quality, has both of its jacks at one end, instead of at opposite ends of the case, and, amazingly, it sells for only $149.00. I find all of those factors to be very appealing!
The iBasso DB2 offers coaxial and TOSLINK S/PDIF inputs in addition to USB - making it far more flexible for use as a desktop DAC. It's designed with two, internal lithium-ion batteries, divorcing it from any reliance on obtaining power from the PC's USB port. This means that it's likely a much better candidate than the other two DACs for use with an iPad, where the USB port might not provide enough current to operate a DAC. The iBasso DB2 includes far more accessories than the other DACs, as well - an optical cable, spare screws for the case, a second pair of Low Pass Filter op-amps (which made no difference that I could detect), a soft pouch, a budget 3.5mm TRS interconnect cable, and a couple of other tidbits. Given the greater feature list of the DB2, it too, should be considered a great value at $299.
Lastly, I should probably mention that I didn't have any need to install drivers (or troubleshoot in any way) installation of these three DACs,for use with my Windows 7 laptop. In fact, having installed all three of them, I didn't suffer any conflicts at all when unplugging one DAC from the USB port and switching to another. I did not attempt to mount more than one USB DAC at a time.
The End
Update: Corrected an occurrence of "DB2" to read "PB2" instead.
Also as I've told you through PM, I think you need to buy the L3 and the RX3. haha, I really want to try the PB-2 too!
Cheers big time!!! Checking these out for sure
