Gandasaputra
100+ Head-Fier
Oh wow, ClieOs is back. Come back here and get your IEMs review thread updated!!
DUNU DN-2000J -- More Than Evolution?
- Thread starter tomscy2000
- Start date
tomscy2000
Headphoneus Supremus
You'll have to ask Brooko about the comparison between the 2000J and the A83; I don't really jive with the A83 signature, so I'm going to believe the 2000J is unquestionably better.
Is the 2000J "worth the price" over the Titan 1? That's a relative question. The 2000J is definitely a clearly better earphone, but would you necessarily have a much better time enjoying music? That's a debatable answer. The 2000J is for people who are chasing the "the next level" of resolution and presentation; most people would be well-served with the Titan 1, however. I believe that if you have to ask this question, you're better off not knowing and going with the Titan 1. Save yourself a few bucks.
How 'strongly' does the 2000 "roll-off" for you? I've always felt the 2000 had more than adequate bass extension for what it was. It's a bit more mid-bassy than the 2000J, but only on account of it just not having as good of a woofer transducer as does the 2000J. My reference device is the Concero HP, which has 2.2 ohms OI (Resonessence publishes very accurate values for their specs, off an AP SYS2722). That's not exactly low, but not unreasonably high either. I get plenty of bass extension. I do feel there is some mid-bass kick, but most of the bass emphasis comes below 100 Hz, between 50-80 Hz.
If the impedance values <500 Hz are indeed ~8 ohms, it is a bit low, but I have reason to question your theory of high-pass effects on the dynamic transducer. If the impedance is indeed that low, then the dynamic transducer and the TWFK are wired in parallel. Any RC circuit connected to the TWFK is not going to exert a high-pass effect on the dynamic driver. To me, what's difficult is that an 8 ohm reactive load will demand a lot of current and be a strain to the amplifier, so I'd recommend good, low-distortion analog circuits with excellent buffers (e.g. LME49600, BUF634, AD8397, etc. --- I know I'm missing a few others, opamps aren't my forte).
With regard to hiss and sensitivity, yes, the 2000J is going to pick up hiss and strangely actually has a tendency to pickup and amplify DC offset current. That's why it's important to use good, quiet sources that have low published values for residual noise/hum (usually measured in uV, and I find that devices with <4 uV are pretty quiet, but for the 2000J, dead quiet will probably have to be <2 uV, because the Concero HP, at 3.5 uV, will report some transformer hum).
Thus, given all this, I don't expect the 2000J to report pretty THD+N values, because it'll be a strain of a load on the amplifier.
Sounds good!
Though I do enjoy DN-2000, but a more analytical or detailed version of DN-2000 is very welcomed!
How does the effect of high impedance output on DN-2000J ? Does it make DN-2000J sounds brighter or warmer?
tomscy2000
Headphoneus Supremus
I only tested for a few moments with a 90 ohm resistor so it wasn't a careful listen; it's not really any brighter, but definitely incurs a loss of warmth, due to some kind of lower midrange "dilution" (for lack of a better way to describe the effect). Everything sounds flat and bland, though I did notice peaks were definitely lessened. What I've found, from testing a few hybrids with the resistor adapter, is that the SPL drops more greatly with the dynamic driver than it does with the BA, so with a resistor, the sound of the BA (with changes to the FR because of variable impedance) is more greatly conveyed.
Never left really, I just spend most of my time in the portable gears forum. Sorry, no more big IEM review from me as I don't have the time to maintain big review thread anymore.
GR07 BE has more bass presence. Bass on 2KJ is more neutral in quantity, though it does reach just as deep as GR07 BE and perhaps hit with slightly more impact. It just doesn't quite have the same body.
Udi
Head-Fier
- Joined
- Mar 10, 2008
- Posts
- 64
- Likes
- 18
I notice the roll-off below 50Hz very noticeably on a source with an OI of 2.3 ohms, starting at about 60Hz. I'm not sure how you can rate an IEM with a poor response to bass frequencies below 50-80Hz as having "plenty of bass extension" but it's nice to see that our results are consistent. Early on in the 2000 thread, syvr also reported a similar falloff below 60Hz.
I can't comment on the source capacitance effects as it's definitely not my area of expertise (I thought I mentioned that - apologies if I did not), it was just something an EE suggested to me might be a secondary cause aside from the impedance ratio being likely less than ideal which was the main point mentioned. However given our relatively similar results I think it may just be a function of the impedance ratio and the impedance curve. The interesting thing about the fall-off on the 2000 is that no amount of EQ rectifies the SPL in those frequencies below the cutoff.
But have another look at the innerfidelity FR graph for the 2000 - there is no falloff visible, yet it has been verified by numerous users now. This makes me certain that it's a function of source and load matching. For reference, on a source with an OI of 0.31 ohms, the bass extension is noticeably improved (in fact, excellent), with reasonable SPL down to the 20-30Hz mark. EQ also has a noticeable effect around the 32Hz band, which is a polarising difference compared to the 2.3 ohm OI source. This is the only reason I've kept mine personally, in hopes that I can correct response on other sources. I don't think any of this is a coincidence, I am however curious whether the problem is a function of the impedance ratio or the significant non-linearity in the impedance curve (or how much of each factor is responsible).
It bothers me slightly that people rate IEMs with insufficient coverage of the bass spectrum of frequencies as "accurate", however on the bright side I believe in this case the fault is not with the IEM's actual frequency response, rather it's capacity to reproduce that range based on the output it is attached to. To put it bluntly, you're missing out on part of the frequency spectrum that these IEMs are capable of reproducing. I can't comment on whether the same thing is occurring on the 2000J but I'd really like to try some because I'm fairly sure it'll have a flatter impedance curve (good) as well as the lower overall impedance (not so good). This would help identify which factor most effects the problem.
As for a solution - I think for now the best bet is to stick with <0.5ohm OI sources on IEMs like the DN1000/2000/2000J unless you don't mind losing a portion of your frequency response curve. A surprising number of people seem to be completely unaware they've lost anything, perhaps a case of blissful ignorance. I think there's also potential in impedance adaptors improving this issue on higher OI sources, however a lot of the factory items use cheap resistors with low accuracy and potential noise - with further issues like mismatched series v. parallel resistance which throws out the ideal/factory damping ratios. I've ordered a pile of Vishay-Dales to build a few of my own designs to see if I can lower the cutoff frequency and bring back some lost tones without coloring the rest of the spectrum too much. From the calculations I've run so far I think 90 ohms is far too much, and a better solution will involve trying to minimise added resistance to achieve a balance between improved (but not perfect) frequency response and less coloration of the rest of the curve / minimised losses in clarity. I do think you make a valid point on how the resistance will mate with the peaks in the impedance curve, in which case the 1000 and 2000J may fare better than the 2000 with such an adaptor. I'm just assuming that the 2000J will have a flatter curve though, do share results if you measure it.
tomscy2000
Headphoneus Supremus
Well, what I meant is what you determine as a "roll-off" --- can you give me an indication of how many dB/decade (just an estimate) that you're talking about? And what would be your reference as to a proper bass response?
I'm not doubting you or questioning you --- just trying to get a better idea. I actually don't have that much experience with the original DN-2000, but from what I've heard from numerous people (like joker), the low bass of the DN-2000 is pretty good?
Udi
Head-Fier
- Joined
- Mar 10, 2008
- Posts
- 64
- Likes
- 18
At 2.3 ohm OI, I would say a ~5-6dB drop below 50Hz where it is most pronounced, with a smaller ~2-3dB drop in the 50-70Hz region where the falloff starts.
I initially thought the joker may have been wrong in his sentiments (on purchase) but after further testing, I think he is correct - as are you - in saying that the lower spectrum of the DN2000 is pretty good (reflected by the InnerFidelity measurements). However more specifically, it is the capability that is good, and my point is that this capability is not actually achieved when paired with all sources. I unfortunately don't have many sources to test with, but as I said with the 2.3 ohm OI there is a noticeable loss. Now, it doesn't sound bad, and for someone with no superior point of comparison it would probably be acceptable - particularly for anyone that wasn't looking for deep bass response. However with a 0.3 ohm OI source the response is noticeably better as I described previously, and in this scenario easily reflects joker's sentiments. To put it bluntly, it only reflects what you read in the dazzling reviews if you mate it to particular sources - and if you don't, the differences can be substantial.
Now that's not to say that a ~2 ohm OI source is a bad one, in fact, I think IEMs should really be designed to work well even with 3-4 ohm sources (as this seems to describe a lot of phones these days). Furthermore I suspect the 2000J may exaggerate this issue given its load impedance @ 1kHz is halved which is why I thought I'd mention it here. As I said though, I may be wrong on this count - and it likely depends on the impedance curve of the 2000J (aside from the single 8 ohm @ 1kHz value).
What would be ideal is if someone had a jig like Tyll Hertsens (perhaps you do?) and compared FR curves between various sources on the same IEM. I suspect on these modern IEMs with very low load impedance and multiple drivers there may be substantial differences.
I initially thought the joker may have been wrong in his sentiments (on purchase) but after further testing, I think he is correct - as are you - in saying that the lower spectrum of the DN2000 is pretty good (reflected by the InnerFidelity measurements). However more specifically, it is the capability that is good, and my point is that this capability is not actually achieved when paired with all sources. I unfortunately don't have many sources to test with, but as I said with the 2.3 ohm OI there is a noticeable loss. Now, it doesn't sound bad, and for someone with no superior point of comparison it would probably be acceptable - particularly for anyone that wasn't looking for deep bass response. However with a 0.3 ohm OI source the response is noticeably better as I described previously, and in this scenario easily reflects joker's sentiments. To put it bluntly, it only reflects what you read in the dazzling reviews if you mate it to particular sources - and if you don't, the differences can be substantial.
Now that's not to say that a ~2 ohm OI source is a bad one, in fact, I think IEMs should really be designed to work well even with 3-4 ohm sources (as this seems to describe a lot of phones these days). Furthermore I suspect the 2000J may exaggerate this issue given its load impedance @ 1kHz is halved which is why I thought I'd mention it here. As I said though, I may be wrong on this count - and it likely depends on the impedance curve of the 2000J (aside from the single 8 ohm @ 1kHz value).
What would be ideal is if someone had a jig like Tyll Hertsens (perhaps you do?) and compared FR curves between various sources on the same IEM. I suspect on these modern IEMs with very low load impedance and multiple drivers there may be substantial differences.
SoAmusing777
1000+ Head-Fier
- Joined
- Dec 1, 2012
- Posts
- 1,011
- Likes
- 26
Would really like to hear how this 2kJ fares against the Ortofon EQ-8.
Udi
Head-Fier
- Joined
- Mar 10, 2008
- Posts
- 64
- Likes
- 18
Same here, apologies if I came across as standoffish. I have noticed discrepancies between subjective reviews and actual performance with IEMs, and while initially I blamed subjective opinions, I think often it's attributable to non-standardized comparisons (particularly impedance ratios when it comes to IEMs). I've been researching this issue for the last month or so, and what I wrote above - particularly the last bit about "substantial differences" was based on my hand calculations and subjective A/B testing (between 0.3 and 2.3 ohm OI sources) but no measured data. However I just stumbled across this graph which performs the test in question and illustrates the point vividly - it compares the FR curve of a BA IEM (SuperFi 5) on varying OI sources. Looks like drops of up to ~3dB are experienced in this case.
The reason for my interest is that I'd like to design higher quality impedance adaptors, not just in terms of good componentry, but actually calculating the ideal circuit based on the source and load impedances used, maintaining or improving the stock damping factor (something many adaptors destroy), and factoring in the desired FR range improvement. I think current adaptors are quite crude and colour the frequency response too much to be viable, mostly due to not being tailored to specific applications. I've accepted that a good enough solution may not be possible and a different amp will be needed, but I'm going to have a solid go at making it work.
Of course there's little need for those who own a 0.5ohm OI source to mate with their 8 ohm IEM, but I presume many out there like their existing sources (myself included) and may not want the complication or cost of a dedicated amp (or the potential drawbacks over their existing amp - for those who might have a really clean 2 ohm OI source for example).
Hope this hasn't muddied up the thread - and just so it's clear this isn't a stab at the 2000J (for all I know, there may be a substantial dip at 1kHz on its impedance curve), just an issue I've noticed with modern low impedance IEMs which would be nice to understand and resolve.
Dobrescu George
Reviewer: AudiophileHeaven
I do not think that it is a fair comparation all the way, i had read the reviews for ortofon, because i was interested in getting it, and it seems like a different signature. Still a good iem though.
Users who are viewing this thread
