Sennheiser HD660S Speculation

Before I speculate, I'm going to talk about the HD650 and HD600. (There is some evidence that the latest HD650s, around the time of the HD6XX have higher third order distortion in the bass; but this is hard to say. I don't have my pre-HD6XX-era HD650 anymore, so it's impossible for me to take more controlled measurements.) In terms of frequency response, the HD650 and HD600 are not that different. From of the few scattered measurements I've taken, the frequency responses of the HD600 and HD650 can differ in two ways: 1) The same in the lows with the differences being in the upper mids and treble, with the HD600 having a bit more of a lift around 3-4kHz, and maybe more overall treble. 2) Roughly the same in the treble and upper mids, with the HD650 having more pronounced bass.

These frequency response measurement difference inconsistencies can be attributed to numerous factors: the pad wear, seal, revisions, etc. However the frequency response differs, an HD600 still sounds like an HD600 and an HD650 still sounds like an HD650. A quick back and forth listen to the HD600 and HD650 is all that is necessary to realize that their differences cannot be explained by only frequency response (or distortion, or CSDs). The HD600s have a more lively zippy quality. The HD650s sound more stately, more damped, even more boring. In other words, the HD600s sound faster. They have faster transient response. This is something that isn't necessary captured in measurements. (It might be, I have some ideas, but that is another topic of discussion).

The faster transient response of the HD600 is exactly why I started on the first set of HD650 mods where I ripped off the rear damping and cut a hole in the foam at the front of the driver. I wanted HD650 tonality, but with HD600 transient response. The Dynamat on the back of the HD650 magnet was just icing on the cake - to massload the magnet driver assembly and absorb extraneous vibrations. Solderdude did a really nice paper exploring the frequency response differences of each aspect of the HD650 mods. A few db more in the bass, slightly less than this in the upper mids. However, I feel Solderdude total missed the point. It was all about faster transient response and getting rid of that "boring" stock HD650 sound.

Now it sounds to me that the HD660S might be something along these lines. Maybe even superior to the HD650 with mods. If the HD660S can have the transient response of the HD650 (or HD700), but maintain the same frequency response of the HD650, then this might be something interesting.

It's important to point out that Jude said the driver is based on the one in the HD700. The nominal 150 ohm impedance of the HD700, same as the HD660S, also confirms this statement. My memory is foggy regarding aspects of resolution or transient response of the HD700, mainly because the spotlit treble was so horrible with some of my favorite CDs. But it wasn't all that bad. I don't remember the HD700 being particularly any more resolving than the HD650; however, it certainly did not sound boring - maybe too much the opposite of boring - Grado transients comes to mind.

Finally, I'd like to set expectations for two things: don't expect better bass extension and don't expect better bass quality. The HD700 driver (or variants as used in the HD660S) is a small driver like the HD650 and HD600. This means it will have a hard time pushing a lot of air, which is necessary for sub or low bass. The sensitivity of the HD660S is only 1db more than the HD650 therefore magnet strength / mass is likely more or less the same. This means total distortion will be similar. Besides the HD700 didn't exactly have good quality bass either. It measured just as bad in distortion as the HD650, but it sounded different. A different sort of bad. Would be interesting to see second and third harmonic proportions in bass distortion. Then again, HD700 cups and pads might have played a role in the HD700's bass presentation.

Finally, successor does not mean superior replacement. Remember, these guys are German. Germans are very literal and into small details and super worry warts. LOL, I can tell you funny stories of me assessing the security controls of German entities. I'm always assuring German clients: relax or don't take the "rules" so literally. Now, the French are the opposite.

 

The HD600 has similar distortion as the HD650. It's all close enough within a few db. The reason the HD650 has a reputation for more muddy or farty bass is because of its slower transient response and heavier bass to treble balance compared to the HD600.

 

Can someone tell evshrug on HF to stop using the term impulse response in place of transient response or speed. The latter two terms are actually very well established by people who design and make audio gear. The term impulse response is actually a very specific thing: a measurement from which a lot of other cool visualizations can be derived from (frequency response, phase, distortion, CSD, step-response, etc.)

Dammit. You see. Now you guys have me reading HF.

 

It was in the context of subjective impressions and aspects of sound was trying to describe. The dude actually rejected the use of the word speed as "too general".

No one says the ATH-AD2000, Grados, or REL subwoofers have great impulse response or fast impulse response. That would be erroroneous since the impulse response can be mathematically transformed into frequency response, distortion, CSDs, and even be used to predict sine burst behavior, etc. and could mean a million other things.

However people do say that ATH-AD2000s, Grados, and RELs are fast, or have fast transient response; or the Aficionado is faster sounding than the ZDS; or the LCD2 Fazor's bass is slow and muddy; or that STAX are super fast with great delineation; or that only the Edgar basshorns were fast enough to keep up with the REPs widebander drivers; or that most Martin Logans suck because their bass drivers can't keep up with the speed of the panels, etc.

 

It's erroneous to use impulse response in the place of "speed" because it doesn't necessarily describe "speed" and impulse response could mean many other things. For example, a Grado has shitty frequency response with massive CSD ringing as a result of its shitty impulse response with massive overshoot and ringing. But it's still a pretty fast sounding headphone.

 

We can argue semantics and meanings, but what it comes down it is that it's best not to mix subjective descriptors with terms that may have a technical or mathematical meaning or significance. I'd rather use the term decay than Energy-Time-Curve in the sphere of subjective impressions.

A refresher: an FFT / DCT on an gated impulse response results in a frequency response; a series of FFTs / DCTs on a series of collapsing impulse response windows results in a CSD.

I mean, what's next? Laplace Transforms of Pseudo Random Noise as a replacement for speed?

 

In electrical engineering, impulse response and transient response are very different things.

Impulse response is the response of a system to an impulse excitation (or a broadband flat signal with linear phase). This is a definition in signals and systems if I remember correctly. This terminology is used across control, DSP, navigation and so forth. It's very firmly established.

Transient response it the transitory and temporary behavior of a system before it reaches steady state. A junior in electrical engineering is exposed to transient stuff in electronics and also in networks. It is NOT impulse response. Transients depend on the initial conditions of the system and is related to initial conditions when solving linear differential equations, which are heavily used in solving networks voltages, currents and shit like that in electrical engineering.

So no. That dude is misusing the terms.

EDIT: Subjectively it's another story. That dude can use PRaT unicorns and proly not even his parents would understand what he means, other than he either jizzed or puked all over when hearing a system perform.