Dynamic Range in Headphones, or lack thereof

After spending an evening with something that wasn’t a headphone I was struck upon return at how constrained the dynamic range felt on my system, even among the better examples like the Utopia and HD800. They just couldn’t quite reach those dynamic highs during the most dramatic passages, they didn’t sing, and felt somewhat muted in comparison. I’ve been feeling quite dissatisfied with them since then with particularly dynamic recordings. Though I suppose this could also in part be a psychoacoustic effect caused by headphones’ natural limitations, for one thing the lack of physical bass sensation.

Alternatively macro-dynamics is another term used for this, though that term combines Slam/Impact in it’s meaning. Which is perceptually related, but not the same thing. Many also refer to poor dynamic range as compression, or sounding flat.

Tracks I’ve been using;
In The Air Tonight by Phil Collins - The cavalcade of drum hits around 3:38
Star Wars IV: A New Hope Theme by Stanley Black & London Symphony Orchestra - Instrumental highs reach a satisfying volume increase around 1:50.


I’d be interested to hear any of your thoughts in and around this, so here’re a few questionable questions of questionable quality to start things off (Tagging @Garuspik @zach915m):


1. Are headphones fundementally deficient vs speaker in dynamic range, and if so why?

2. Modern planars have a reputation for compression. What is the primary impediment for planars, and how could they be overcome?

3. Which planars and electrostats possess the best dynamic range, and do they meet or exceed the HD800 or Utopia? For that matter are there any dynamics better than those two?

4. In terms of amplifiers and DACs which types generally exhibit notably better dynamic range, and what are some specific pieces of gear that demonstrate excellence in this area?

5. Using my chain as a base for upgrade how could one make more than slight improvements in macrodynamic performance, with the least expenditure: Eitr -> DEQ2496 -> Gungnir Multibit A1 -> Mjolnir 2 w/WE396A -> Utopia (stock)

 

Very interesting and complex questions, I'm not sure to know the answer of questions 2, 3, 4 and 5, but my theory (not entirely sure about this) regarding question 1 ( "Are headphones fundementally deficient vs speaker in dynamic range, and if so why?") is: surface area of the drivers.

For example, Utopia's driver is 40mm, that's more than 6 times less than a 10 inch (254mm) woofer, which in a 3 way speaker is used just to play bass only, thus free of doing midrange and treble duties unlike a single driver headphone.
Such a big surface area plays a huge role, the woofer barely moves to play bass, this IMO translates to that "unstressed" dynamic range that speakers have vs headphones.

PS: this is an overly simplified answer, as this subject is too complex, for example two speaker drivers of the "same size" (6.5 inch) according to the manufacturers, such as the Scanspeak Illuminator 18W and the Audiotechnology C-Quenze 18H have different surface area (154 cm2 for the Illuminator, 137 cm2 for the C-Quenze). There're also crossovers involved with multi driver speakers, which is another complex subject but also has an influence on the sound.

Mass and operating principles may be another factor, I'm sure that 15 inch woofer will sound punchier and more robust than a planar bass panel of the same surface area, assuming that both are driven by the same amp and have the same efficiency.

 

It might just feel that way because you can't feel the impact of the bass on your body.

 

@9suns I am seeing a corollary between these things, possibly long-throw drivers, and speakers/headphones believed to have powerful magnets. Lending themselves to characteristics like slam, and tactility. Which I believe are contributory to a sense of dynamics. Rather than, or perhaps in combination with, literal dynamic range.

@riker1384 I believe that would be one aspect among a number of traits that create a sense of dynamics.

I’ve not had any time to explore this further and follow up, but myself and @E_Schaaf recently had a conversation on-profile that touched on this. Here’re the pertinant parts:

Re. ADSR Envelope
https://blog.landr.com/adsr-envelopes-infographic/

 

I'm not sure I want more dynamic range in my headphones. There are times in music and movies where I think I have it just right. Enough volume for the quiet moments and then something happens and it gets way too loud in lots of different frequencies. And this sound is right on my ears. Ouch! With speakers I just worry about the neighbors and not the pain in my ears.

 

Everything below and the quotes above are of course grossly oversimplified, merely reflecting things I've noticed from my own experience.

Another important factor for both dynamics and bass response, from what I can tell, is the size of the transducer vs the volume of space requiring pressurization. This is proportionately much higher in headphones compared to most floorstanding speakers in a medium to large sized room.

Hence, a 40mm+ headphone transducer has no problem reproducing bass down to 30hz or lower given it also doesn't have to have massive excursion to pressurize this small, sealed volume within a headphone pad... this is much like how it's easier for a larger woofer to reproduce low frequencies in a larger space with less excursion breakup compared to one or multiple smaller woofers. Also explains why larger planar drivers in headphones have no problem reproducing bottomless bass with much lower measurable distortion in comparison with dynamic drivers.

The most radical example is in IEMs, where you can have a transducer larger than your actual eardrum sealed within less than half an inch of space. I've always felt IEM bass, when done right, has the deepest reach and tightest and most immediate dynamic profile (I guess you could say the cleanest spectral decay?), even though it feels the most "compressed" given there's effectively zero bodily conduction happening.

All I know is that the most macrodynamic experiences I've ever had of music has been in clubs with massive speakers, high volume, and poor acoustics. Low and slow, boomy and bodily. Microdynamic... probably those ATCs I mixed with in the college studio a few years back. Headphones always feel "faster", "leaner" and "cleaner", but always less fun with less ineffable physical presence than most quality loudspeakers can effortlessly produce.

 

I have observed this also. The Utopia quickly loses it’s dynamic reputation when changing pads to something that increases the internal volume or are more air-permeable.

What’s confused me was why the Utopia with it’s long throw 40mm driver out punched planars with their much much larger surface areas, even though they extend lower. Other headphones with Bio drivers have exhibited this also, the commonalities being a stiff diaphragm plus real surround, and high gauss magnets. I suppose this is due to the thin, flexible, low mass nature of planar diaphragms making them less forceful.

But all this is almost entirely focused on low end aspects and slam.

LOL, I think I might just see your point. Does this example extend out of the low frequencies?

 

Once you join the BWC you kind of have an intuition for such things. Shame big woofers aren't as popular in contemporary commercial designs.

Bigger midrange woofers are nice too.

 

Headphones have obvious advantages over speakers in terms of detail and so on.

But speakers have psychoacoustic advantage over headphones and that helps them sound bigger and more dynamic. They tick off more checkboxes in terms of fooling your brain, to some extent, into feeling like you're really there.

• Speakers provide full-body tactile sensations, as everybody's already said
• In real life, when you turn your head, the sound source doesn't move with you like it does with headphones
• In real life, room reflections play a big part in how our brains understand sound. Speakers are a little closer to reality here.
• In real life, HRTF plays a major role in how our brains understand sound. Again, speakers are a little closer to reality here.

Speakers are far from perfect at those things, just a bit closer to reality than headphones.

 

I think this concept of "reality" or "realism" when it comes to sound reproduction is very problematic.

All recorded sound is acousmatic. Often times sounds are recorded in a fairly dead space and the sense of space is purely manufactured artificially. Can the reproduction of an artificial sound ever be anything but artificial?

Even with live recordings, microphones are usually multi-tracked and mixed with fairly heavy-handed processing... unless the engineer is using a spaced pair or ORTF / Blumlein / etc etc mic configuration. You'll usually only see that with few classical recordings.

With most actual live performances (when you are present in the room) you won't have the microscopic focus a treated microphone recording can provide, either in the studio or in a live room. Recordings tend to sound cleaner than the actual thing.

At some point I gave up on the idea that gear was meant to reproduce "the real thing". It's a fruitless effort. That would only be possible if every recording ever followed the same exact mic protocols, and were recorded in the same room, with the same processing, and if your speakers had the same response as the speakers used at the time of the performance, with no room response whatsoever in your space.

Mics themselves have phasing incoherence baked into them independently and relative to other mics in the same space. This is unavoidable. Sense of perspective will always be skewed as soon as something hits a mic.

Mics have coloration. Mic preamps have coloration. Cables have coloration. Mixing consoles have coloration. Digital recording systems have coloration. Every instance of processing - analog or digital, hardware or software - induces coloration. Transducers themselves are filters. There is no "real thing".

"Realism" does not exist. I'd rather go for what I feel sounds good on a personal level.

This especially falls apart as soon as you consider any type of music aside from of purely acoustic live performances.

 

Headphones suck pretty much versus a reasonable pair of speakers for dynamics and linearity but your ears are inches from the driver so they can have some sort of more perceived “detail” due to being close up even if the actual midrange balance is more accurate on speakers. You also take the room out of the equation too. The “muh room sucks” or “muh roommates will bitch at me” argument is some internet wimp shit. Yeah maybe you shouldn’t put 15” woofer stereo subs on the floor to listen to anime soundtracks and trance if you don’t want to deal with the police but it doesn’t mean you can’t use speakers at a reasonable hour and volume.

That “everything is colored so it doesn’t matter” argument is ridiculous as some things are much, much more colored than others. There is still a range of “neutralish” and no it’s not biodym cans or single driver horn speakers with horrific beaming or Zu audio cancer.

Hyper reality is more real than real. It is impossible to know mostly what the recording really sounded like before it was recorded, mixed, and mastered but you can try reproduce the recording. Raw multi tracks of a drum kit are actually kinda dull much of the time and clean compressors to even them out won’t sound as real as one which have been run through standard produciton processes that are colored and more distorted and more artificial but sound more natural than the actual clean raw tracks summed together. It’s a lot like how people need makeup on TV to not look sick under the lighting but overdo it (as so many productions do) and they will look like mannakins in HD.

If you don’t give a f**k about trying to hear the recording then you don’t give a f**k but don’t pretend that you gave a f**k and what you ended up with was right. There are tons of dudes who don’t care about the panning and depth on the recording, who just want everything blown out and 10 feet away from them just like there are tons of dudes who don’t care about the levels of the instruments or reproducing the alignment of kick drum and bass guitar and you can buy transducers and gear that will just give you what you want rather than reveal the production.

Some transducers are far more real than others and you can hit the limits of transducers fairly easily as well as having to deal with “defective by design” bs. A three way to cover the entire frequency range will have an additional crossover somewhere that will f**k some stuff up compared to a two way and you’ll have to sit further back, meaning less perceived detail, but it will fill a room better than some small woofer two way. Some 6.5” woofer isn’t going to reproduce stuff below like 60-70 hz well at all and crank them as room speakers Zeos style? They’ll be awful. Big ass bass woofer in an untreated room tend to be mud city but a small woofer might sound better in a bad room due to having less extension in a bad room. Might as the smaller woofer in the range might not have the volume to fill the room anyway even with the less extended bass. Big ass 8-10” clean woofer in a two way ain’t reproducing midrange as well as a smaller woofer would no matter what marketing or “mids good enough” testimonials you read. We all know that’s bullshit and bigger woofer two-way vs smaller woofer two-way vs three-way is a compromise of extension, detail, dynamics, volume, and where you want to sit. Some people just want a PA system. Now all that audiophile old fart speaker show “muh magic 6.5” woofers fill the room with sound shit is hilarious. Yeah maybe those woofers can hit 60-70hz ported 3-4 feet away on a desk but these dudes claim they will fill the whole room. No way in hell you’ll hit a good bass volume cleanly 10 feet back sitting on the couch. Same with the active stuff that claims to do bass with a limiter and eq. Crank those 3-5” two ways and they all fall apart no matter how scientifically clean and long throw the woofer is even in nearfield. For example at CAF, Vanatoo tried the room filling sound bs with the Transparent Ones and Zeros and they sounded awful but weren’t even close to the worst thing there. On a desk as multimedia speakers? They’re not bad for the price.

 

Everything is colored, and coloration is not the same as distortion or accuracy of any kind either. The microphones themselves are colored, the amplifiers used by the guitarists are colored, the types of sticks the drummer uses and what the drum heads are made of cause coloration. Your own ears cause coloration. Your memory causes coloration. All sound, even acoustic, is a coloration of energy as expressed through waveforms of air pressure, resonating within a medium of variable parameters, and detected by inherently flawed transducers of some kind...

The question is whether or not you perceive certain colorations to be pleasurable or accurate according to your aesthetic biases and previous experiences. There is no standard. A paper cone will reproduce the sound of a large diaphragm condenser mic differently than a mylar panel. Maybe that coloration will be better suited to your tastes. Maybe one has a better distortion profile and measures with less dips and spikes in the FR but doesn't meet your preference.

I said nothing about neutrality at all. I also didn't claim you can't reproduce a *recording* in a faithful way. All I said is that any recording in and of itself is inherently too far deviated from "truth" to claim that any playback system can re-create the acoustic experience of an actual performance in your living room, as so many audiophool authors tend to write about. Hence, "realism" does not exist. Realism is not accuracy.

I also never made an evaluative claim about how processing is used in audio. I prefer the processed sound. I enjoy the sound of compression. I enjoy the fakeness of manufactured recordings. But it is not "real", and frankly I'm glad because most of my experiences of live music above a certain volume level suck ass in terms of the qualities I tend to enjoy from a playback system. I am not in search of truth. I am in search of enjoyment. Anyone who thinks they can have a "true" system is fooling themselves.

 

I agree with you there. Sorry, I misread your intentions.

That’s what the audiophile press has always tried to sell though and it’s impossible. They use a bunch of random positive descriptions of how boutique recordings nobody has ever heard of sound real with maybe an FR graph if you’re lucky. If it’s Stereophile, expect it to be some wax cylinder field recording of Pygmy folk songs from the 1920s that only comes to life on the uber-expensive wooden monstrosity being reviewed that usually doesn’t sound anywhere remotely normal with Taylor Swift.

 

@purr1n you should run burst tests or linearity tests to show how big speakers compare to little ones

 

I'm thinking about level linearity (Voltage In vs SPL out). Similar to what we see in DAC linearity (Input code vs Voltage Out) plots

Would be pretty easy to do with basic tools at any fixed frequency. Maybe do one in each region (50Hz, 2k, 10k). Could also sweep it and plot it in 3 dimensions (akin to waterfall plot)

I'm guessing most speakers will show good linearity within reasonable levels. You might see some compressive effect (curve goes <45 degrees) at high volumes particularly in the bass, but probably higher than you would want to test a speaker you own. Noise might dominate the low levels, but it would be curious to see if there was a compressive effect or any abnormal deviation from linearity.

 

Soundstagenetwork does that. IIRC it's mostly the crosover regions that show compression rather than the bass, aside from very small speakers. Bass should be mainly suspension nonlinearity me thinks.
Still, looking at their database it's honestly surprising how much compression some of these speakers have at not-too-unreasonable levels. 3db compression at 90-95db (vs 70db) isn't that uncommon. But then again maybe not that surprising considering how manufacturers cheap out on crossover parts even in megabuck speakers.

I tried to measure the effect with my HD800, but IIRC I couldn't measure any compression up to past 110db. Should be here on the forum somewhere. It's possible that we start to see some at 120db+, but I didn't go that far. 110db is good enough for me.

FWIW I don't think that headphones are inherently worse at macrodynamics than speakers, but I do think that the best speakers are more dynamic than the best headphones. I also think that a well driven HD800 will beat the crap out of 99% of speakers in the macrodynamics department.

EDIT: Will stop adding stuff now, lol