[blades]

  I have 3 headphone amps and the HD800 sounds different with each of them. The differences are not as trivial as the quantity of bass or highs - it goes down to finer details like control of decay, speed, resolution.  

 
You don't want to be using a headphone with an output impedance that is higher than the load impedanc.

 

[SilverEars]

   
You don't want to be using a headphone with an output impedance that is higher than the load impedanc.

I understand it's not efficient and with HD800's case the uneven impedance characteristic would change the signature.  But, what would this do to a flat impedance, resistive planar?

 

[davidsh]

Will do nothing to a planar, disregarding lowered power output

 

[SilverEars]

Will do nothing to a planar, disregarding lowered power output

What about the damping factor.  I know this factor is electrical, and there will be the "factor" since the value difference is there, but what is it doing to a planar?  What does this damping factor generally do?

 

[castleofargh]

 

 
a huge number of misinformed people here believe EQ to be the ultimate devil(most without any real knowledge or experience of a good EQ). but doing exactly the same **** with at least the same defects by looking for a "colored amp" is OK.
didn't you ever hear something like "EQ is destroying the sound, I want the real sound"?

Depends on the EQ.  An electronic device is characterizable by its frequence and phase response.  The phase response is, effectively, a time delay that varies by frequency.  A 'flat' phase reponse is effectively a propagation delay.  Different analog filter designs (butterworth, chebychev etc) have different characteristic phase responses.  To a large extent good analog filter design is about trading off undesirable phase response in favor of desirable frequency response.  The problem with non-flat phase response is that it changes the positioning of the frequency spectrum in the soundstage.  Your filter may attenuate around 1kHz, but it then causes 2kHz to show up later.  (In the time domain this shows as ringing. All filters ring.)  This is in analog filters.  In a DIGITAL filter, a so-called sinc filter, phase response is flat.  Always.  There is a propagation delay equal to the size of the filter kernel (5, 11, 64 samples, or whatever).  This is obvious to anyone well-versed in the art - convolution with a kernel of size N requires holding onto a window of N samples.  Phase response is flat at N.  Straight line.
 
The way you characterize an electronic device, typically, is by feeding it a unity impulse.  This is a brief spike (as brief as possible, less than one half sample).  You then sample the output.  From this you can create a filter where if you feed it the output you can deconvolve it into the input.  It works for any signal and effectively any device.  Through a complex Fourier transform of the collected data you get the frequency and phase response of the device.  It can be graphed and used as effectively as any sweep or analog method (except it avoids some phase problems in instrument amplifiers and such).  Similarly, a filter kernel can (easily) be produced that takes an input and replicated the behavior on the output, effectively emulating any device.
 
Well, any LINEAR device.  If you feed it a half-unity spike and get a different response you have a non-linearity.  But different I mean a different waveform out, not just a weaker version of the unity response.  The bad news is the real, physical world is non-linear and we makes simplifications where we approximate it to be linear.  Some things are close enough to linear that the error is irrelevant, such as transistors.  Tubes work by electrons beaming from a cathode, gated by a grid, and collected onto an anode where they form the output signal.  This is not a very linear process.  Because tubes aren't linear a characterization in terms of unity frequency and phase response they can't easily be characterized, and the physical mechanism is too complex to easily produce a mathematically accurate model.  It's truly woo.
 
So "coloring" can mean many different things.  It can be a digital EQ, an analog EQ, complex impedance (i.e. reactive load), the response of a linear device, or the response of a non-linear device.  For the non-linear kind, the only really practical approach is to swap things out and see what happens.
 
Non-linearities also come in many different kinds, with different subjective results.  Some produce ugly distortion.  Some, not so ugly.

I had to ask google for help as I didn't remember writing this. but I really did, on the first page of the topic ^_^(luckily I still agree with my former self).
thanks for the detailed explanation, indeed different type of EQ will have different undesired effects of different magnitude depending of how much we abuse them(that's probably the most important part). my point was that people ignore the magnitude of those effects compared to the rest of the audio chain and wrongly imagine that a 3% THD amp does less damage to the signal than a good EQ.
of course we can have luck and get something that we like after randomly picking non linear gears(probably), but I'm guessing most people bought a hd800 because it was a "hifi" product, the sennheiser flagship, one of the best dynamic driver headphone on hearth, or any combination of those reasons. so it doesn't fit very well with gambling on a happy massacre of sound from an overdriven tube amp with a massive roll off. or maybe my definition of hifi clearly isn't good, and all people really ask for is to have a relaxed low-fi rolled off sound but on a headphone they can show proudly as being "hifi". that sure would explain a lot, but wouldn't be related to the actual quality of sound.
 
 
 

  Some more examples of EQ changing decay, this time with normal EQ.
 

Spoiler: Images

Unmodded vintage planars:

 
The same planars, now with some physical mods to remove the hump in the lower mids/upper bass:

 
The same planars, unmodded like in the first picture, but with regular EQ used to roughly match the unmodded response to the modded response:

 
The decay characteristics below 2 kHz are fairly similar whether modded/EQd, and in both cases notably better than unmodded. The decay isn't identical between modded and EQd, but neither is there a complete match in responses, something that could be tweaked with more careful EQ. I can't remember if I used parametric or graphic EQ, but it was either Electri-Q or xnor's, potentially xnor's.
 

Spoiler: Images

Another pair of old planars, with physical mods:

 
The same pair of planars and with the same mods still in place, but with EQ (Electri-Q) added to correct the tonal balance:

 
As in the convolution example a few posts up, and what castleofargh was wondering about, the frequency response was scaled up (relatively) with EQ in the upper bass area, yet the apparent decay there went down, at least in part.

and thank you too, the second series certainly speaks to me. "if" listening to it could make the decay variations clearly noticeable, I would certainly have convinced myself that the bass was "masked" by the sub bass boost(so blaming my brain for something really happening on the headphone). but here the boost really does have a reducing impact on the neighboring frequencies decay.
I don't really get why. maybe the push on the driver to try making the sub bass louder get to a point where it cannot maintain its movement, and that reduces the mobility for the close by higher freq, making it stop that oscillation faster? I'm clearly wild guessing here and I'm probably totally wrong, but it doesn't matter why, the result is what matters most here. thank you for taking the time to post this.

 

[castleofargh]

 

   
You don't want to be using a headphone with an output impedance that is higher than the load impedanc.

I understand it's not efficient and with HD800's case the uneven impedance characteristic would change the signature.  But, what would this do to a flat impedance, resistive planar?

more than the damping factor for the phone, there is the problem of energy no getting dissipated in the headphone. so the amp would have to deal with it probably making it get hot(the most obvious way to dissipate energy).
at some point if the headphone impedance is a lot lower than the amp's, from the amp's point of view, it will start to look like you just short circuited it with a wire.

 

[davidsh]

Damping factor means nothing for planars, though it matters with dynamics. It has to do with the induced current when the driver moves, as I understand it. Not sure about the details. Anyway, planar drivers induce no or a negligible small current when the diaphragm moves.
Probably a very simplified explanation.

 

[ab initio]

Damping factor means nothing for planars, though it matters with dynamics. It has to do with the induced current when the driver moves, as I understand it. Not sure about the details. Anyway, planar drivers induce no or a negligible small current when the diaphragm moves.
Probably a very simplified explanation.

I think this is an incorrect explanation.
 
The traces on a planar diaphragm move through a magnetic field, which induces back emf. If there was no magnetic field present, then running current through the transducer wouldn't induce the the force that creates physical motion of the transducer.
 
Damping factor is just as relevant to planar magnetics as it is to traditional dynamic drives when it comes to electrical damping!
 
No matter how times I link this article, it doesn't seem that anybody bothers to read it.
 
Cheers

 

[SilverEars]

No matter how times I link this article, it doesn't seem that anybody bothers to read it.

Well, sometimes, interpretation helps more than reading something and not getting the idea because there isn't something extra there that somebody explaining it may provide to trigger the light bulb.  

  The light bulb regarding electrical damping to mechanical has not hit yet.
 
Dynamics have coil.  The Planars don't.  The coil I can understand current through it for back EMF.  Two types have difference in design.
 

 

[Dogmatrix]

 
I think this is an incorrect explanation.
 
The traces on a planar diaphragm move through a magnetic field, which induces back emf. If there was no magnetic field present, then running current through the transducer wouldn't induce the the force that creates physical motion of the transducer.
 
Damping factor is just as relevant to planar magnetics as it is to traditional dynamic drives when it comes to electrical damping!
 
No matter how times I link this article, it doesn't seem that anybody bothers to read it.
 
Cheers

The article is not really relevant to headphones
Because of  the low mass of a headphone driver and the small low energy coil the forces discussed in the article do not apply in any significant way

 

[davidsh]

@ab initio
I did read said article right after posting, actually.
 
I suppose you are right.
 
I don't think electrical damping should matter much wrt headphones, though. Don't know really. I wonder whether a 500 ohm in-line resistor with whatever planar will affect sound significantly.

 

[SilverEars]

  @ab initio
I did read said article right after posting, actually.
 
I suppose you are right.
 
I don't think electrical damping should matter much wrt headphones, though. Don't know really. I wonder whether a 500 ohm in-line resistor with whatever planar will affect sound significantly.

This is what I've been asking all along.  If somebody that knows this please explain it.  The planars are interesting because the source is seeing a resistor at every frequency, unlike a HD800 where it varies in impedance so it doesn't see it as resistor.  The coil is pretty much an inductor.  The Planar does induce magnetic field from the magnets, not coils and there is a thin and very light free moving diaphram in between.
 

 

[castleofargh]

 

  @ab initio
I did read said article right after posting, actually.
 
I suppose you are right.
 
I don't think electrical damping should matter much wrt headphones, though. Don't know really. I wonder whether a 500 ohm in-line resistor with whatever planar will affect sound significantly.

This is what I've been asking all along.  If somebody that knows this please explain it.  The planars are interesting because the source is seeing a resistor at every frequency, unlike a HD800 where it varies in impedance so it doesn't see it as resistor.  The coil is pretty much an inductor.  The Planar does induce magnetic field from the magnets, not coils and there is a thin and very light free moving diaphram in between.
 

even I don't look as obsessed by impedance. and that's saying something man.

you already had all the answers over several topics, you know from measurements that planars tend to have a flat impedance response over freqs. so no problem with coloration.
  what do you want more? if the damping ratio gets bad, the efficiency of the amp falls down, as I said it will get hot for the amp(maybe to a damaging point here if the amp has a really high impedance output, given the power usually used on planars). the drivers will have less control, but on planar who really cares as less would still be amazing transients.
so you'll waste more power when they are already very demanding drivers in optimal situations. and that's certainly not super clever or interesting. hay I got those they're super fast and super power hungry, what can I do to try make them even more power hungry and ruin the transient response?
 having a bad damping ratio is never a good thing for drivers(at least for optimal signal restitution). a few guys do that on speakers because they prefer more boomy bass(is that still a thing or do they just buy a subwoofer nowadays?), some do it to change the signature of some custom IEMs and it's mosty ok because there is close to no energy used for those IEMs. and everybody else avoids bad damping ratios or doesn't know what he's doing.
 
 

 

[SilverEars]

  even I don't look as obsessed by impedance. and that's saying something man.

you already had all the answers over several topics, you know from measurements that planars tend to have a flat impedance response over freqs. so no problem with coloration.
  what do you want more? if the damping ratio gets bad, the efficiency of the amp falls down, as I said it will get hot for the amp(maybe to a damaging point here if the amp has a really high impedance output, given the power usually used on planars). the drivers will have less control, but on planar who really cares as less would still be amazing transients.
so you'll waste more power when they are already very demanding drivers in optimal situations. and that's certainly not super clever or interesting. hay I got those they're super fast and super power hungry, what can I do to try make them even more power hungry and ruin the transient response?
 having a bad damping ratio is never a good thing for drivers(at least for optimal signal restitution). a few guys do that on speakers because they prefer more boomy bass(is that still a thing or do they just buy a subwoofer nowadays?), some do it to change the signature of some custom IEMs and it's mosty ok because there is close to no energy used for those IEMs. and everybody else avoids bad damping ratios or doesn't know what he's doing.
 
 

So basically, don't use bottle head crack with LCD-X.  

  If you don't get the humor, the LCD-X is 20ohms and crack has over 100ohms output impedance.  The tube should be nice and toasty.  Would be nice to have during the winter to keep you warm and cozy.
 
Anyway, ok, the electrical damping in the case of resistive load is just power delivery issue.  That's good to know.

 

[ab initio]

  Well, sometimes, interpretation helps more than reading something and not getting the idea because there isn't something extra there that somebody explaining it may provide to trigger the light bulb.  

  The light bulb regarding electrical damping to mechanical has not hit yet.
 
Dynamics have coil.  The Planars don't.  The coil I can understand current through it for back EMF.  Two types have difference in design.
 

If you can ask questions about the parts you don't get, then we can move the conversation forward. However, continually asserting that planarmagnetic headphones aren't affected by the amplifier's damping factor because they don't exhibit a strongly frequency dependent impedance is both wrong and not helping anybody address the question in a meaningful way. I am happy to try and elaborate on the points that need elaboration, just point me where the pieces are missing. I link the wiki article because it introduces what the damping factor is and how it relates to speakers. It's a good foundation to start the discussion where all parties involved can use common terminology.
 
There is a "coil" on planardynamic drivers as well. It's the copper/aluminum/whatever trace etched into the driver. You can clearly see it in the picture you've included above. The fundamental operating principle of operation is exactly the same in both dynamic and planardynamic drivers. You might notice that the two names differ by only the addition of "planar" in the name "planardynamic". The difference comes from the way the magnetic field lines are distributed in the planarmagnetic drivers compared to the classic drivers. Rather than having one large magnet with a coil on the driver as in classic dynamic drivers, planarmagnetic drivers have an array of magnets with the etched coil winding through the magnetic field. Innerfidelity has a nice little description about it here. you can also read a wiki excerpt here.
 
 
 

 
The article is not really relevant to headphones
Because of  the low mass of a headphone driver and the small low energy coil the forces discussed in the article do not apply in any significant way

Please point to anything that corroborates this claim. Thanks. The fact that my headphones be effin' ringin' suggests that the dynamics of headphone speakers could benefit from electrical damping.
 
 
 

  @ab initio
I did read said article right after posting, actually.
 
I suppose you are right.
 
I don't think electrical damping should matter much wrt headphones, though. Don't know really. I wonder whether a 500 ohm in-line resistor with whatever planar will affect sound significantly.

The driver in a speaker, whether of dynamic or planardynamic desgin, will have inertia. Remember, it's the motion of the speaker cone that generates the sound waves. Whether or not that speaker motion requires electrical damping depends on the mechanics of the speaker. The zeroth order model for the driver is a mass-spring-damper system with external (the electronics) forcing. Unfortunately, I don't know what "typical" parameters are for a headphone speaker. As for the 500-Ohm resistor, you can always try it and see.
 
Cheers