ZMF Caldera Measurements and Impressions

As an insane hobbyist, I’ve been always highly curious what multi-grand headphones bring to the table. After experiencing a good number of hyper-expensive headphones from major/boutique manufacturers myself, I had to admit that they all differentiated themselves enough from similarly-priced competitors as well as less pricey categories.

Nonetheless, still spending 4k+ dollars for a pair of headphones could not be a trivial decision for me. Only those meeting my personalized set of requirements in technicality and tonality with preferably one extra punch or two are eligible to be shortlisted. In other words, they should be super special in the ways I can appreciate.

ZMF Caldera effortlessly circled or double-circled all the items in my checklist. Needless to say, it joined my exclusive club.

Thanks to @zach915m I will be able to evaluate Caldera for the next few weeks. Will update this thread with measurements and subjective findings during my time with this super-capable pair of headphones.

Relevant links

• General impression thread: https://www.superbestaudiofriends.org/index.php?threads/zmf-caldera-review-and-impressions.12693/
• Product page (stock wood): https://shop.zmfheadphones.com/collections/stock-headphones/products/caldera
• Product page (LTD): https://shop.zmfheadphones.com/collections/stock-headphones/products/caldera-kingwood-ltd

 

Subjective thoughts after one week

I’ve been enjoying Caldera for one week. And it’s growing on me every day. Hence, I don’t want to be judgemental or conclusive at this point by any means. That said, my early thoughts on Caldera’s sonic traits below.

Aside: I received three different sets of perforated pads (lamb, cowhide, and suede) and will surely discuss how they differ later. For now, I’d stick to common things.

1. Downward tilting tonality. Strong and rich bass, a bit laid-back yet clean midrange, then slight touches of sparkles on top of mildly uneven highs. Very harmoniously voiced.
2. To me, Caldera’s technical prowess punches up its price point. Switching from/to my other planar references (Susvara and Stealth) failed to prove informational losses.
3. One of Caldera's traits that impressed me most is its capability to make everything 3D. Even with poorly or closely mic'd records, Caldera nicely creates a big spherelike headstage and places phantom images orbitally -- like a wraparound movie screen. With higher quality tracks that take great care of spatial components, all the complexity and variability are audibly and explosively increased.

I really have way much more to ramble.. But I will save my words for later posts.

Measurements will be up tomorrow.

 

Measurements Part 1: Base measurements / Stock pads

Ok, let's look at measurements. I am going to begin with stock pads ("Caldera Perforated Lamb"; as usual ZMF, please do not confuse with headphones name and geometry name).

Measurement Configurations

• Measurement rig: MiniDSP EARS + SMSL HO200 + SMSL M500.3
• Measurements were calibrated at 94db SPL @ 300hz.
• 5 different positions (center + up + down + front + back) were measured then average values were taken per channel.
• My target curve is based on a piecewise combination of builtin HPN and HEQ curves. Highly close to SBAF compensation v1.0 as a result.

Frequency responses (Top: L+R average, Bottom: L vs R)




Comments/Thoughts

• Subbass attenuation is exaggerated in the graphs above, and likely a seal issue. I found the bottom of pads was a bit loosely attached to EARS's silicon plate. Pushing inward immediately recover a few db. They seal better with my actual head/skin (YMMV). I still hear minute attenuation by 1-2db at the 20hz end though.
• The 4.5khz peak suggested by the graph is likely an artifact, too. It seems associated with internal resonance created by EARS's cylindrical ear canal, earcup geometry, and etc. Not very hearable in the way it is measured. I suspect a more accurate (i.e., pricey) ear couplers would address this issue.
• Slightly recessed region around 1.5khz is audible. It looks like HFM scooping pattern but with a narrower bandwidth and less insensitivity. Subjectively this is associated with the perception of laidback and distant vocals. Fortunately, this cut relatively boosts 1khz and 2khz, which helps to maintain vocal clarity.
• Bass and lower midrange is slightly emphasized, which I believe is correlated with ZMF house sound. Subjectively Caldera sounds closer to earlier ZMF tone than more recent Verite or Atrium.
• Highs are fairly neutral, which is consistent with what I hear. Not overly bright, but fairly aggressive for ZMF.

A few comparative plots will shortly follow.

Spoiler: Other measurements

Warning: The following sets of measurements could be experimental at best and misleading at worst. I highly recommend ignoring them.

 

Measurements Part 2: Caldera Pads




Comments and thoughts

• Like I said in the previous post, bass attenuation is highly influenced by how well pads are sealed with plates. Cowhide pads look the most attenuated. But please note they are the thinnest in depth and the most inelastic in cushion, which troubles solid attachment to the plates. They're more or less the same to me in subbass quantity.
• I'll update my subjectives in how differently they sound later in another post. They sound much more different to my ears than the curves indicate.

 

Measurements Part 3: Comparisons with other ZMFs




Comments and thoughts

• These are all with stock pads. Caldera with the Caldera lamb perforated pads, Verite with the Universe lamb perforated pads, and Atrium with the Universe lamb perforated pads
• Curves are normalized at 600hz, which I believe the closest to how I hear their differences
• For unknown reasons, my measurement system exhibit a big dip in 9khz with Atrium, which seem partly associated with some artifacts
• I'll revisit how they subjectively differ in later posts. Please do not think graphs are everything we hear!

 

Measurements Part 4: Comparisons to Other TOTLs




That's it for now. More updates (subjectives, comparison plots, etc) to come later.

 

Per-Album Listening Notes: Part 1

One thing I am recently very into is to express my audiophiling joy with specific albums. I strongly believe hifi headphones -- acoustic transducers in general -- are merely a tool in the end. And hence meaningless if they don't improve lisner's music life. This along with some relevant thoughts made me motivate to write about my gear enthusiasm in the musical contexts.

The following notes come from my serious listening sessions done with Caldera last week. Qobuz links are attached for your references. I'll update more Caldera-specific listening notes in this thread down the road.




https://open.qobuz.com/album/0073145377982

On the Caldera, I consistently heard solid imaging and natural-ish timbres across this album. It also delivered an expansive, spherelike headstage, throughout which sounds suddenly popped in and/or slid along the outer edges of a multilevel 3D circles. There's clearly something I am inspired to say beyond merely deep and wide soundstage. Studio effects were reproduced in a highly euphonic and effortless way. This presentational nature seemed to make the sound a tad diffused to my ears tho.

Another thing that impressed me was how Nil's trumpet was rendered with Caldera. The trumpet sound spreaded like a spilled drink across the free space, flooding it. Plus, for unknown reasons, the reproduced trumpet felt like a morphing mixture to some extent. There was a great amount of information I could hear: shifting shapes, subtle sonic contours of tones, and rushing breath. It's really an immersive experience.




https://open.qobuz.com/album/0798387952066

Throughout the whole album, Caldera carefully revealed that there was a lot going on with successfully disintegrating all the messiness. Everything sounded transparent and orderly. For example, in the first track ('Chubb Sub'), Caldera reproduced organ, double bass, and drum with abundant musical storylines as well as barreling-down momentum. Caldera didn't quite replicate clearer sightlines into the paths charted by each instrument like Stealth, but it did offer the bulging and air-chuffing beat. Double bass notes wove dexterously through the busy passages. Organ notes beamed with color and kept their composure regardless of their pace. The hi-hat took its place amidst the percussive space. Caldera also helped me clearly identify John switched to piano later in the track, without throwing timbral differences at me.




https://open.qobuz.com/album/ca0s3vcak7nka

On Caldera, Patricia's voice was highly intimate. It came with not only sexiness but also a sense of air temperature. It was also easy to hear lots of breathing room where notes traveled. Inside the venue, Caldera reproduced airs around the crowd, the instruments, and Patricia's voice. Her voice also exhibited a good heft as if it was hard hitting the microphone diaphragm. How physical it was.

On the other hand, John's guitar plucks sounded fully alive and right in front of me. The phantom image was convincingly realistic with his breathing, the air encircling him, and his playing movements. There seemed to be more lower midrange energy with the Cowhide Caldera, which made the hand booms were more visceral. Needless to say, everything sounded very harmonic.

 

Per-Album Listening Notes: Part 2



https://open.qobuz.com/album/0030911113629

I revisited my favorite Organ symphony rendition today. Listened twice -- with Caldera and Susvara (as a control/reference) respectively. Susvara is still awesome, but Caldera is no slouch in resolving every micro-detail as well as nuance.

I'm really surprised that Caldera was textually more transparent-ish and realistic overall. Susvara might be tad easier to pick out and follow the organ in slow or soft passages. But Caldera came with more heft in a more vivacious fashion. Caldera also could maintain the balance between organ and orchestra in the finale as equivalently as Susvara. The impact of opening and the finale's run-up to the conclusion was TBH more goosebumping with Caldera. I'd say Caldera's reproduction is more closely associated with what I deem realism.

 

Per-Album Listening Notes: Part 3



https://open.qobuz.com/album/0801061025632

Revisited this highly experimental, complex, and profound jazz work today with Caldera, Susvara, and Stealth. It's a large-scale electronic music with deep bass and maximal effects, which I do want to listen to with the most technically capable headphones. Caldera equipped suede pads (I love these pads whenever analytic capability is called for) btw.

This album is very challenging because at any given moment it spans most of the audible frequency range and fills the whole headstage with clutter. Nonetheless, all the three pairs of headphones impressively had little trouble sorting it out. Definitely more resolving than any other totls I had in the house.

I carefully focused on the following three aspects: (1) how large area each distributed informational piece over and (2) how separately, orderly, and distinctly they did.

Caldera and Susvara casted wider headstages while Stealth reproduced higher ones. Similar depths among these three. Stealth made me recognize each sonic particle more clearly with the least diffusions. Susvara was the most distracting. Caldera was in-between. I'd note that both Susvara and Caldera felt like "something else" was also filled, which I believe was associated with some diffusions and distractions. I'm very unsure if they are either good-ish (plankton) or bad-ish (suboptimal separation) things. Maybe both. Excellent bass articulation and delinearation among all the three. Very satisfying summit-fi moments.

 

Caldera Pads

The three Caldera pads seem well-optimized for the Caldera performance. This statement should be surely granted for all the headphones.. but I've witnessed many products whose 1st party pads just disserviced. Yeah, I'm looking at you, HFM lol.

Anyway, after spending some time with each of those three pads. Here is the summary.

Caldera Lamb Perforated (stock)

• Personally the least preferred
• In many aspects, somewhere in between Suede and Cowhide
• Jack of all trades, master of none

Caldera Suede Perforated

• The most resolving and analytic (resolution per se is not quite associated with pads.. but suede pads presented every sonic component with the least ambiguity)
• Smoothest tonality in highs
• The most linear
• Slightly rounded + fast decaying in transients

Caldera Cowhide Perforated (thinner than the other two)

• The most engaging and dynamic presentation
• A little low-mid bass hump
• V-shaped sounding
• Powerful and energetic sounding
• Slightly ediger and sharper in highs

While Caldera is caldera regardless of pairing pads, to my liking, I obviously prefer suede and cowhide to lamb. The ranking between suede and cowhide is rather subtle and can be overridden by the mood, genre, etc etc. The two feel pretty complementary ime.

 

Per-Album Listening Notes: Part 4



https://open.qobuz.com/album/w5wd2t3ctif8b

Wow, Caldera (equipped with suede pads atm) impressed me in the Firebird this morning. Caldera's very powerful and soulful presentation exhibited a fantastic mariage. For example, in the beginning of the first track, Caldera's reproduction presented that very subtly rumbling -- almost murmurings -- in the lowest strings were clearly separated from accents of the lowest winds and brass. They were also well-arrayed across a huge soundstage. The transition to full orchestration came soon successfully with heft, weights, and impacts to die for.

Switching to my Stealth, I found the orchestration became a bit more colorful and clearer. Tad better location, too. But I also lost a bit of extra warmth (not only tonally but also harmonically), power, and so-called soul I heard with Caldera.

 

@zach915m posted his measurements on various pads here. Copied and pasted below.

Stock lamb (solid) vs Suede (dashed):


Lamb (solid) vs Cowhide (dashed):


Lamb (solid) vs Lamb-thin (dashed):


Lamb (solid) vs Lamb-thick (dashed)


I haven't heard or measured lamb-thin or lamb-thick. By the way, I'm assuming lamb-medium is stock. Please correct me if I'm mistaken.

My measurements on pads a few posts back. Could be interesting to see how the difference varies across different ear couplers and/or measurement fixtures.

Some quick observations/thoughts

• As I said before, I don't bother variability in sub-bass measurements. I still believe it's more influenced by the seal rather than inherent sonic nature.
• Zach's 5128 measurements and my EARS measurements respectively and partly account for what I hear. But neither perfect enough. I'm suspecting there are several latent variables associated with Caldera's back damping that cause additional variability on real ears (over measurement couplers). That said, I'm finding my measurements closer to my perception, perhaps because I'm smoothing out some of position-dependent effects by averaging 5 different measurements.
• None of the pads above seems to address 1-2khz scoop enough. It's way less severe than any HFM (Susvara included) tho.

 

Yeah that's all correct. I would also say that I hear the way the suede sounds more in line with your measurements too. I measured twice with the suede to make sure I had it right, because generally the suede does sound smoother to me.

 

FAQs: Part 1

Finally, Caldera is officially launched today. While a bunch of early reviews begin to pop up here and there, it’s still not sufficiently clear which audiophile cluster Caldera will serve for. Below are my answers to some of the frequently asked questions I’ve witnessed in other forums.

-----------------

Q1. Is LTD (Kingwood atm) a must?

There are surely a lot of sonic and aesthetic considerations. People often vary in opinions of hardwood value/worthiness. I used to be a hardcore hardwood advocate and I still am.

Nonetheless, potential buyers should be aware that Caldera is inherently heavier than other ZMF models, ceteris paribus. This is probably due to larger radiating space and magnets. Choosing too heavy headphones without taking one’s long-hour neck tolerance into consideration would result in less headtime, which happened to my hardwood Auteur and Verite previously.

The demo unit I’m having is stock wood and weighs 525g (wo cable) -- lighter than hardwood ZMFs I owned before lol. I can attest it can exhibit whatever magic Caldera is capable of. If I pick up my own pair today, I will very likely go for the stock wood. YMMV.


Q2. Can Caldera replace my Verite/Atrium/Auteur with?

Well, not exactly.

This question is probably asked hoping to recoup (daunting) Caldera costs. Unfortunately, if you like the way how your ZMF sounds, I can see you will find some areas Caldera can’t fulfill as much as the ZMF you’re having.

Don’t get me wrong. Caldera has audibly greater technical prowess such as resolution, separation, and 3d headstage. No doubt. But ZMF products come with distinct sets of strengths and presentation styles. And in particular, if you highly value organic expressions and dynamic driver timbres, Caldera may not be substitutional enough -- different at best.

For the reference, if I’m asked for HFMs (e.g., “Can Susvara replace HE1000 or HE6?”), I wouldn’t hesitate to answer yes any day.

That said, I’m perfectly fine with Caldera being the only ZMF pair I’d keep (even in timbre or organism). But it’s really subject to what you’re looking after.


Q3. Do I need hyper-powerful amps to drive Caldera?

I’m not chasing headphone amplification rabbit holes these days as much as I used to be. I still believe amps make differences.. But, beyond certain points, not proportional to the benefits I’d expect for the price deltas. I’m also a little disagreeing with general underpowering concerns when it comes to orthos/planars. I am rather concerned with overpowering cases because surplus current does more harm than good to me.

That being said, Caldera has not shown any hints of underpowering with moderate SS amps I hooked up with. The current Caldera requires is way less than Stealth or Susvara by lots of decibels. I believe Sonic synergies or flavor alignments should be prioritized in choosing amps for Caldera. I’m planning to test with some exotic tube amps my local pals have down the road. But for now, I do recommend not to rule out Caldera solely because you don’t have monster-power amps.

As an aside, I’m personally liking when Caldera is driven by amps with low noise floor, harmonically colorless, and clean power supply.


Q4. Which secondary pads should I choose?

I’ve briefly discussed this previously. Choose suede if you want to hear all the sonic components more distinctly and clearly. Or, go for cowhide if you want more engaging styles.


Q5. Who should shortlist Caldera?

One audiophile cluster I’d recommend shortlisting Caldera to is those who are considering Susvara.

I’m saying neither that Caldera is better than Susvara nor that Caldera perfectly satisfies this group. But I found these two headphones were more closely matched to each other than to anything else in this price range. And for this reason, I do believe Caldera is the closest alternative to Susvara.

To me it’s pretty obvious they both have similar capability realm and presentation style in detail retrievals, soundstages, and liveliness. Tone, warmth, and/or planar-likeness (or lack thereof) can be deciding factors in final purchases. I am currently debating whether to replace Sus with Caldera for myself. Sigh.

 

Thanks for the great review @Vtory

Do you have thoughts about the size of the stage between the Caldera and the Sus? And/or the HE1000SE?

 

To me HE1kse (from the recent memory) reproduces the widest and tallest stage but the flattest in depth by comparison. Not bad at all but the other two are in a little different league in spatial stuffs, relatively speaking.

Susvara's soundstage is almost as wide. But much more deeper in z-axis, which makes a little more convincing dimensions overall. Also the most expansive + the most open feeling. The most lively presentation, too.
Caldera is 99% Sus in terms of sizes. Not as much open-up, either (almost there tho). By far the finest in lateral and depth definitions (i.e., placements and layers feel very precise and evenly distributed). The least diffusing among these three, too.