For some reason, I was particularly eager for these headphones. The Moondrop team has been training in making planars for only the second year, but they seem to be doing it quite consciously. Therefore, my personal expectations were high, especially because Laputa has become the ‘second most expensive’ model, leaving Cosmo in the first place.

Whether Moondrop managed to make a decent model is what we’ll talk about today. Here they are, Moondrop Laputa, classic open-back planar magnetic headphones for $690.

Special thanks to Kirill who sent me these headphones – I am grateful to you to the utmost degree!

This seems to be the first online review of them ever.

What’s included

The headphones come in a hefty metal case.

The case is really metal, mostly aluminum, so it reliably protects the headphones during transportation. I’ll say that much, it looks respectable. It has plenty of that. There is a good-for-nothing foam bed wrapped in a black cloth inside the case. The cloth is not attached to the bed in any way and, therefore, tries to get out from under it all the time. When I took the headphones out of the case, the cloth straightened out, and now, when closing the case, it will stick out of the gap between the upper and lower lids. What an idiocy.

What do you get:

the headphones;
a balanced 4.4 mm cable with XLR and 6.3 mm adapters;
one pair of earpads;
some papers.

The model has not yet appeared in the English version of the manufacturer’s website (it’s here in the Chinese version), so here is a picture from the user manual:

All but the sound

Laputa are made on the basis of the same ‘standard’ Moondrop frame, but with some differences. The headband is carbon fiber now. It’s more flexible than that of PARA or Cosmo, although being structurally and geometrically identical to its predecessors.

The headband overlay is 3D printed from a rubber-like material. It’s sort of ‘double-layered’ and has a honeycomb structure.

The earpads are traditionally mounted on thin 10-centimeter metal plates that are magnetized to the earcups. The original earpads are completely identical in shape to PARA ones, but they don’t have perforations on their inner surface. Internal space is 6 cm in height, 4.1 cm in width, and 2.2 cm in depth.

The driver membrane is round, 10 centimeters in diameter.

The headphone earcups are made of solid pieces of aluminum, the desired shape of which was produced on a CNC machine. The earcups are completely acoustically open from the outside (Moondrop explicitly uses the term ‘unveiled’ in relation to this model), but they’re protected by convex metal grids.

The sockets of the connectors are directed at a slight angle forward.

The headphones are assembled well and feel reliable.

How does this design perform in terms of ergonomics? It’s good in some aspects and bad in some others.

As for the rubber overlay, it fits tightly around the head and perfectly distributes the weight of the headphones. It’s also perfectly ventilated, so the head doesn’t sweat under it. However, it collects dust very quickly, and it’s almost impossible to wipe it properly. It could be rinsed with water, but you’ll have to partially disassemble the headphones in order to remove it.

As for the headband, the use of carbon fibers has reduced its weight, but has also added flexibility to the headband to the extent that the headphones are literally inconvenient to use. When trying to put them on the table, they’re randomly deformed, as the headband doesn’t hold heavy earcups vertically. When worn on the head, the headphones feel like… a loose ear-flapped hat, with weights attached to the earflaps. The earcups swing on the head with any, literally any movement of the head. With a weight of 540 g, the vast majority of which is accounted for by the earcups, the headband should be more rigid.

Therefore, I can’t praise Laputa in terms of ergonomics: they’re heavy, they frankly shake on my head, and they’re not comfortable to handle while holding in my hands.

The cable, in short, is unusual and flat. I can’t guess why (except for the greater ‘audiophilia’?).

Let me repeat, the kit includes not only the cable with a balanced 4.4 mm connector, but also two adapters for connecting to various sources: XLR→4.4 and 6.3→4.4.

As for electrical specification, the manufacturer claims the headphones’ impedance of 60 Ohms ± 15%. My measurement was 68 and a bit more:

The sound

Standard links:

the description of my rig is here;
the audiogram of my hearing is here;
articles on measurement theory and the whole shebang are here;
some fine points of over-ear headphone measurements are here.

Frequency response of Moondrop Laputa headphones with perfect and real pressure:

You will never ‘hear’ the blue graph because, as I wrote above, the headphones are dangling on your ears, which always creates a significant air gap. And I suspect that the manufacturer was counting on exactly the sound that is shown on the green graph.

On this very graph, we can observe as follows:

The aim was a VDSF curve, which is obtained through measuring a sweep tone in a certain ‘ideal room’ through an ear simulator at the level of the eardrum, while the sweep tone was being played through some Genelec monitors (source). That is, Moondrop aimed at a certain neutral sound of their own. This is murky waters…
The graph at 20-1000 Hz is incredibly smooth for headphones of this type, but it droops by 3.3 dB at the maximum in relation to 10db Tilt DF (if you combine the headphone graph and the target curve at 1250 Hz).
The right part of the frequency response graph, from 1000 and above, also looks great in the sense that it doesn’t have the accents that are typical of planars in the range of 5-9 kHz, and even the classic peak around 13 kHz is not large at all.

Besides, I’ll note that there are practically no changes in frequency response due to rotation and/or shift, which is typical for all open-back planars.

The result is as follows: with a realistic fit, Laputa provide a sound where there is either more top or less bottom, which is the same thing – the assessment depends on the specific preferences and ears of the listener. Along with that, the frequency response of the headphones deviates from 10db Tilt DF very smoothly. Actually, in order to bring the Laputa sound to neutral, you need exactly one parametric equalizer filter: Freq=120, Gain=3, Q=0.2:

If we apply this filter, we get a really ‘smooth’ neutral sound. If we don’t apply it, we get either a lack of lower middle and middle sections or slightly emphasized medium and high frequencies.

Personally, my perception of the ‘pure’ Laputa sound is somewhere in between one and the other. This is a very technical sound delivery in terms of frequency balance, free from tartness and sharpness at upper frequencies, but emphasizing the upper middle. The headphones sound full-fledged, the information from the track reaches the listener in full, but the sound is frankly ‘light’. But it’s subjectively ‘smooth’. The sound stage is wide, but the instrument positioning is mediocre. I would say that these are the most ‘normal’ ‘light-sounding’ open-back planars that I’ve ever heard, although I’m not a fan of such sound delivery.

The ‘hump’ on the subbass depends entirely on the fit: the larger the air gap, the stronger the emphasis on the subbass, but the more it is shifted to the right and the sharper the decline on the subbass. Again, this is true for almost all planar headphones.

Is there any reasonable way to influence this sound signature by changing the earpads? I checked it out.

The results can be found in detail at my squig.link. I see no point in commenting on each measurement separately. I’ll focus only on one Laputa frequency response graph with standard earpads from XK Audio Fluxion headphones:

You can read more about these earpads in my respective review. Let me remind you that their feature is special dense tubes, which, in fact, ‘forcibly’ preserve the air gap at any type of the fit.

These earpads significantly change the Laputa sound: the upper middle is muted and comes to a kind of consensus with the lower middle and the lower frequency range as a whole, while the 5-9 kHz range is emphasized. Laputa start sounding more balanced, but inevitably sharpish. Of all the earpads that I have tried with Laputa, I would recommend this particular option to get an alternative sound hue of these headphones. The trouble is, I don’t know where to get these earpads separately from the headphones.

Next come the technical aspects.

The left and right channels match perfectly in volume over almost the entire frequency range. This happens very rare:

Nonlinear distortion at 94 dB with the ‘Use harmonic frequency as ref’ option turned off and on:

The manufacturer promises that distortion at 1 kHz at 94 dB volume will not exceed 0.05%. I got 0.06%. The result is excellent, but not outstanding. Fluxion, for instance, have 0.04%.

There’s no disaster visible on these graphs, that is, at 94 dB of volume, the distortion is well below the threshold of perception by the human ear.

Nonlinear distortion at 104 dB with the ‘Use harmonic frequency as ref’ option turned off and on:

If you decide to say goodbye to your ears and turn up the volume to 104 dB, then yes, you can hear distortion around 3.3 and 5 kHz. But, to tell the truth, I can hardly imagine a person who would seriously listen to headphones at such a volume.

Group delay and spectrogram in the ‘Burst decay’ mode:

The ‘Burst decay’ graph can frighten with long ‘flashes’, but this is a fairly common pattern for open-back planars, and this is a good result. You can look at a bad one, for example, in the measurements of HiFiMan Edition XS. And to be clear, as always, I don’t have an anechoic chamber to make truly accurate measurements like this, so the microphone of the measuring rig can catch re-reflections of sounds from the walls and ceiling of the room (although I struggle with it as best I can).

Comparisons

As an example of good open-back planars, which many people are familiar with, I chose Moondrop PARA:

PARA are much more full-fledged in the low frequency range, not so aggressive in the middle (to put it mildly, just because they actually have a dip there), but they have an unpleasant accent around 12-13 kHz. From my point of view, PARA are ergonomically superior to Laputa – their headband is stiffer, the headphones don’t shake on your head so much – but their sound is inferior to the Laputa’s one, as the latter sound more natural. On second thought, the prices of these models vary threefold.

Frequency response of Moondrop Laputa compared to Moondrop Cosmo:

I’m not a fan of the Moondrop Cosmo sound. I don’t understand why the manufacturer wants a pot of money for such a mediocre sound, in my opinion. The middle of the midrange dominates in the Cosmo sound – there is a clear accent around 1000 Hz on the graph, while 3 kHz, on the contrary, comes ‘short’ in volume. The weight and ergonomics are comparable, but note that Cosmo are more expensive and more difficult to equalize. From my point of view, Laputa win in almost all respects.

Frequency response of Moondrop Laputa compared to XK Audio Fluxion:

On Aliexpress, Fluxion cost some $900 now, that is, they are more expensive than Laputa. Fluxion are 50-100 g lighter (depending on the earpads), even more inconvenient to handle, and made incomparably worse, there’s nothing to speak of. However, I would put them above Laputa in terms of sound: the upper middle is less bright, the dip in the lower middle is smaller, and the overall sound is much more balanced and natural.

Frequency response of Moondrop Laputa compared to XK Audio Avalon mkII:

Avalon have been discontinued, but they’re still available for purchase for some $570. These are fully open-back dynamic headphones with huge 70mm drivers. In this model, XK Audio wanted to get a neutral sound, and, in general, they succeeded. There are still some caveats, there are flaws in their sound, and as for design and usability, this model can be criticized at great length and with abandon. But I’m unaware of any headphones that match the 10db Tilt DF target curve better than Avalon. As you can see, if you don’t pay attention to the 20-30 Hz range, then there are practically no errors in the Avalon tuning. The main drawback is the peak at 13 kHz. It’s really there, it’s not a measurement error, and it’s clearly audible. And, well, as for the subbass distortions, they… could be better, to put it mildly. But if I only evaluated the sound, I would choose Avalon.

The next step was supposed to be a comparison of Laputa with the new HiFiMan HE600, but, unfortunately, I don’t have them.

Summary

Let me tell you what everyone is thinking of: all open-back planars from Chinese manufacturers are very similar in sound. The price does not always correspond to the conditional ‘sound quality’ of headphones, but if we consider merely Chinese open-back planars, then their prices seem to be set completely randomly. Well, not always, but exceptions are very, very rare.

And the sound of Chinese open-back planars has features that repeat from model to model: a lack of low frequencies (compared not even with the Harman curve, but at least with the diffuse field curve), accents at the upper middle and a range of about 12-13 kHz. Again, there are exceptions, but they are rare.

It’s clear what good open-back planars should sound like – the above disadvantages need to be fixed. I make no strong claims that this is easy to implement from an engineering point of view. But why continue to produce models that are as like as two peas in their sound?

Laputa, I believe, are taking a step in the right direction. They give a light but natural sound. There are no narrow accents in their frequency response that are unpleasant to the ear, and the ability to build an expansive sound stage complements the generally decent picture. Plus, there is a possibility to easily change earpads of a common diameter, very decent build quality and overall workmanship of the device, and an interesting cable with adapters. And a case! What also seems important is a very simple equalization with literally one to three points of the parametric equalizer due to the overall smoothness of the frequency response.

However, the scales can be turned by poor ergonomics caused by both considerable weight and the headband too flexible.

To buy or not to buy: to buy if you want exactly that kind of sound and are ready to put up with an unstable fit.

P.S. I’ve listened to a lot of open-back planars over the past few months, and to my own surprise, I have to admit that the Fluxion’s sound (and only sound) can be called outstanding when compared with technologically similar models.

P.P.S. I’ll just leave this here: