Having not had a chance to respond properly in the other thread (working all day), I'll post a response here.
My main issue with the Havi is simply that their reproduction is not as accurate or high-fidelity as some of its competitors. In fact, it's very far off of some entry-level sets.
Since talking about subjective listening preferences is an exercise in futility, I'll steer things in a more objectivity-oriented comparison.
ljokerl recently sent both a pair of VSonic VSD3S and Havi B3 Pro 1 to Tyll for measurements. Given ljokerl's knowledge, experience, and reputation, it's incredibly likely that these were both real, correctly-functioning units. Now measurements may tell the whole truth, but barring measurement error, they do not lie and the results are very telling of overall performance. I chose to compare the Havi to the VSD3S because they are likely going to be competing for the same market, and for no other reason. I personally own a pair of VSD3S that I bought to replace a pair of old GR07 I broke, and they've exceeded my expectations and perform incredibly similar to the GR07 at a fraction of the cost., but that's neither here nor there, and is entirely irrelevant.
Anywho, on to measurement analysis!
VSD3S:
http://www.innerfidelity.com/images/VSonicVSD35.pdf
Havi:
http://www.innerfidelity.com/images/HaviB3Pro1.pdf
Looking at these graphs, both of these headphones measure exactly as I would expect them to given my time listening to each of them.
Frequency Response
The VSonics show a very smooth response with a slight bass-tilt. This should come as a surprise to no one, because it mirrors how they sound exactly. There's a slight recess centered around 4k which does pull some of the harmonics out of vocals and can make them sound somewhat withdrawn, but is otherwise very inoffensive.
The Havi runs into some trouble here. It looks like a 3-band equalizer with each band turned up. The low end is far less linear then the VSD3S with a ~100 Hz-centered hump and rolled-off sub-bass. There's also a bit of a null at about 500-600 Hz. The transition into the midrange is not very smooth, but the mids themselves are not terribly non-linear. Perhaps slightly elevated, but it also adds some bite which can be desirable. The VSonics might sound somewhat soft when comparing directly to the Havi. The treble is highly elevated and quite uneven, but the elevation is more noteworthy than the peaks and valleys. Also of note is the rather severe channel mis-match. The left channel seems like it could be a pleasant, albeit still "flavorful" listen, but the right channel appears bright and harsh.
Isolation and Phase/Impedence curves are similar and not really noteworthy in a direct comparison.
30 Hz Square Wave
This square wave is essentially playing the entire frequency band down to 30 Hz simultaneously. It's used for gauging bass response within the time domain.
The VSonics exhibit a very clean, well-shaped square wave here. Ideally there would be a small overshoot with a perfectly horizontal top to the wave, but these perform well in this measurement. The slight hump and downward trend is a sign of more mid-bass emphasis, and less than ideal sub-bass response. However, the line remains above zero through the entire wave, meaning that the bass remains tight and extends low.
The Havi perform very poorly in this measurement. The large hump and severely downward trend of the graph points to a strong midrange and midbass emphasis over the lower frequencies. The wave hits zero about halfway through, pointing to some trouble in the low-end.
300 Hz Square Wave
This is just like the 30 Hz wave, but it only plays frequencies from 300 Hz and up. It's a zoomed-in look at the very beginning portion of the 30 Hz wave, and is used for evaluating performance in all other frequencies and in the time domain.
Once again, the VSonics perform excellently. There is a small overshoot, followed by a brief period of some very minor ringing, ending in a flat waveform top. The overshoot is slightly less than ideal, meaning that the high frequencies are somewhat lacking in presence, but it does reach the waveform top, so they are not going to sound dull. The slight null following the overshoot is very likely due to the upper-mid recession mirrored in the FR, and the ringing is very likely responsible for the hard treble and sibilance heard in listening. Overall, though, the wave is very clean and has very good shape, a clear sign of a well-balanced headphone where each frequency is well-integrated.
The performance of the Havi in this measurement is...well...it's horrid. The overshoot is of a considerable magnitude, and there is actually another smaller overshoot prior to the large one as well. This has a tendency to manifest as a smeared image where it is difficult to localize audio cues within the soundstage because the ear is receiving impulses at different times. The top of the waveform never settles into a clean line and has a strong downward trend, pointing to uneven frequency transitions and a thin tone.
THD
It's getting ridiculous at this point, but once again the VSonics perform very well here. Distortion is low and mostly uniform across the board. The 100 dB line rising above the 90 dB line in the lower frequencies is a sign that the bass tends to loosen up some when you crank the volume. 100 dB is quite loud though, and you probably shouldn't be listening that loudly, anyway. Overall low distortion in the low frequencies reinforces the fact that these headphones have tight, articulate bass response. The midrange is is very clean, so there's really nothing of note there. The slight rise in the upper frequencies is another likely culprit of the hard treble and sibilance that can be heard,
The Havi drop the ball here as well. For one, the 100 dB curve is above the 90 dB curve for almost the entire spectrum, which is a clear sign of a headphone that falls apart at higher volumes. The bass distortion is very, very high. Comebined with the poor performance of the 30 Hz wave, it's very clear that these are going to have loose, imprecise bass. There's a large increase in distortion that accompanies the FR null at about 500-600 Hz. I'm not sure what that's about. I'm wondering if it could be a measurement artifact, but given how they sound, I wouldn't be surprised if it isn't. The distortion does mellow out in the higher frequencies, which does help treble performance and, combined with the boosted treble, is likely why people find them to sound detailed.
Impulse Response
This is a bit redundant after looking at the square waves, but it's worth looking at.
Again, the Vsonic is very clean. A clean initial impulse followed by some modest ringing with a timely decay. There is some very low amplitude ringing that continues for quite some time, which is probably the same ringing that we see in the 300 Hz wave that I'm sure is responsible for the hard treble and sibilance.
The Havi seem very under-damped. The left channel seems pretty clean, if a bit slow, but the right channel has has some severe ringing after the initial impulse, though it does fully decay in a reasonable timeframe. Still, that ringing is going to cause issues.
And there you have it. Of these two headphones, I greatly prefer the VSonics, and I think the measurements explain quite clearly why. Some might prefer the sound of the Havi, but there is no denying the fact that the VSonics perform better. If I were to recommend one of the two, it would be the VSonics without question.
