# MYSPHERE 3.1-UNIQUE in SOUND&DESIGN&BUILD QUALITY

Discussion in 'Headphones (full-size)' started by nomax, Sep 8, 2016.

1. So would a Mysphere 3.1 with added resistors be equivalent to version 3.2. Is the difference between 3.1 and 3.2 just resistance

2. Contributor
Mathematically YES!
But I (we) have to check this in detail much deeper!
As explained earlier, the most important Thiele Small Parameters except of the BxL are exactly the same (Mysphere 3.1 / 3.2).
However, when listening and comparing with different but well matched amps, there are certain differences to hear.

The next month will show more experience to this matter. And I will let all of you know...

BR heinz

3. Contributor
Now I'm confused. Why is it not desirable to achieve maximal damping!? We're talking of the electrical damping at the resonance frequency provided by the impedance hump in the context of back EMF. Why would we want a certain delay of decay?

Let me explain: The goal of a matching damping (factor) is an even frequency response – nothing else. You'd always want an impulse response as accurate as possible. «Too thin» is nothing that couldn't be perfectly equalized in the amplitude domain.

Maybe you remember the blessed Bose 901. A concept based on a small volume for nine rather small fullrange drivers resulting in a high resonance frequency (somewhere between 125 and 170 Hz, I don't recall exactly). The resulting «bad» frequency response and damping has been «perfectly» compensated by means of an adequate equalizer box. It produced maybe the best (fastest, most precise) bass reproductions at its time to my ears (the rest is a matter of taste).

Dobrescu George likes this.
4. The basic idea of any amp/transducer combo is to produce a flat response ( or predetermined other-than-flat ) - WITHOUT any EQ.

To cater for the proper EQ of an over/underdamped transducer a parametric EQ is a MUST. As frequency response of any real transducer is temperature AND amplitude dependant, no fixed parametric EQ is likely to cover for all the variables - even if sample to sample variations are neglected. The mentioned problem reared its ugly head in one of the full range Tangbang miniature speaker drivers; FAR too broad frequency range and Q of the resonance displayed under different power ( > temperature ) and amplitude ( the lower the frequency, the higher the amplitude ) for any reasonable solution to be adopted > discarding the said driver for "something else ".

I agree the Bose 901 ( in its multifarious incarnations ) is a VERY good loudspeaker if used properly. But, it does require its dedicated EQ box.

Analog parametric equalizers are few and far in between - and once more than 5 bands are required, the price of a decent analog PEQ dwarfs the price of the Mysphere ...

Digital DSP ( no other way than using PCM... - something I try to avoid unless absolutely unavoidable ) that allows parametric EQ exceeds - for the time being - processing power of most ( but not all, particularly TOTL designs ) portable devices and is therefore limited to home/desktop use.

Not to mention that any EQ has to be implemented individually by each user; how many can EQ in a way that does more good than harm is a valid concern.

Compare all the above to the simple combo of the amp with the proper output impedance for the headphones in question.

5. Contributor
Your post is shooting itself in the foot. If you're concerned about sample variation and heat-related aberrations, the «problem» isn't limited to an equalized system, but is valid for an untouched system as well. But don't worry: In the real world the problem doesn't exist to the degree you make it look like. I'm saying this as someone who's been excessively occupied with designing and building speakers during many years (on a hobbyist basis, but with serious audiophile ambitions).

«The basic idea of any amp/transducer combo is to produce a flat response»: a pious hope. I've never been satisfied with the available flatness, be it with speakers or headphones. Moreover: the reason why we have to care for the damping factor and the Q factor of the bass resonance is the fact that there is one at all. It forms the lower end of the reproduced frequency spectrum of a dynamic transducer and isn't a desired feature, but a physical restriction. The Q factor of the resonance (graphically displayed on impedance graphs) defines which damping factor is needed for a flat low-frequency response down to the final, unavoidable drop-off. Most (open) dynamic headphones fall off long before the 20 Hz mark, so if you want to cover the whole audible spectrum, there's nothing left but equalizing. Now if you have a headphone with a low-Q bass resonance – actually calling for a lower damping factor – there's no harm done if the drop-off becomes smoother and begins earlier than desired if you wouldn't compensate it. Quite the opposite: It facilitates equalizing, since the compensating curve will be less critical. But in both cases you'll be able to «perfectly» flatten amplitude and phase response and by doing so eliminate the resonant behavior of the mechanical resonance frequency. So you gain both low-frequency extension and impulse accuracy. The same applies to the rest of the frequency spectrum, where's a lot to be gained in terms of amplitude and transient response by careful equalizing.

Ironically I don't use a parametric equalizer (allthough I've always favored these), but a (free!) 31-band/⅓-octave graphic equalizer (BTW developed by a Head-Fier) with foobar2000. I even use the FiiO players' simple 10-band octave equalizers with great success on the go. In any event I couldn't live without them anymore, I'm too much spoiled by their sonic potential. And I have a hard time understanding why many music lovers here renounce them.

«Compare all the above to the simple combo of the amp with the proper output impedance for the headphones in question.» A typical example may be the Sennheiser HDVD 800 / HDVA 600, specifically designed for the HD 800, with a built-in output impedance of 120 Ω (IIRC) to compensate for the relatively weak bass. It's not even an adequate compensation on the electrical level (resonance Q factor), but purely meant as a counterbalance for the HD 800's brightness. (Graphs from Inner Fidelity)

A crude and cheap method chosen by Sennheiser. It accentuates primarily the range around 100 Hz, whereas the frequency response would call for an increase below 100 Hz. That's what I do (and SonarWorks does), and it literally transforms the HD 800 into a high-end headphone.

Last edited: Jan 14, 2018
Matrix Petka likes this.
6. Contributor
Thanks for the link! – I'm (im)patiently waiting for the definitive launch of this fascinating new way of listening to music.

7. With pleasure !

D.M.

8. An owner and lover of the k1000 I have put myself on the list for their successors. I never thought it would come about. My dilemma is which model to get . Given that I have the Smyth Realiser A16 coming and the Chord Mojo and the HiFiMan megamini all of which should suit the 3.1 model it would be indulgence gone mad to get a 3.2 for my First Watt F7 power amp. I'd appreciate some advice on this. The majority of everyday listening is done with the megamini but for the perfect day it has to be my @dill3000 DIY f7 aka the "mini beast"

9. The article mentions a price of around \$3K but I don't know if that's accurate. My impression was that it would be \$4,500 to \$5,000. I obviously would be delighted if it was lower, but not sure whether there has been any confirmation of price. I signed up for the no obligation pre-order but haven't received any further notifications about it.

10. It probably seems like an eternity but it'll turn up before you're ready with the finances and that's when the hard choices will have to be made....

11. On the main page of the Mysphere it says that "Depending on the specification selected and any specific taxes of your country, the price will range from € 3,500 to € 4,500. Full details of any pricing are included in our quotation." - http://mysphere.at/
Obviously, you should just buy both.))
Actually, it's kind of funny. On one hand there is now some choice - you can buy the version which suits you best - and it seems like a great thing. But on the other hand now there is a painful dilemma - which version to buy - which is not so great. I myself also can't decide which version I would like to buy. Sometimes I use headphones with great solid-state amps and sometimes I usem them with portable devices. So they suit me both and none.

Last edited: Jan 19, 2018
12. Contributor
Hello,
basically the article is very OK and true.
However, one "typo" is in - it's the price point!
The final cost which customer will have to pay is not totally finalized yet. This is because we have not received all feedback from the logistic supply partners for word-wide shipping.
Actual calculation will end upon with a sales price landed at customer or total € 3900,-
But please be patient for a view more days till final price can be settled.
BR heinz

13. Contributor
Heinz would be much more qualified to answer this, but I'll give it a try.

On the one hand, both Mojo and megamini have very low output impedance (<=1 ohms), so they won't be a perfect pairing for the 3.1. On the other hand, the Mojo should easily have enough power to drive the 3.2. The only remaining unknown would be the megamini. I have no personal experience, but found one user review stating he drove his HD650 successfully (though not ideally) with it... so I think you'd probably be ok with the 3.2.