What and how do we know what IS ‘Better’?
or
It’s all in our heads, or is it?
Part 21 Spikes, Overshoot, and Wonderful.
I’ve been playing with my 800’s for a while now and have tried a variety of ‘adjustments’ in various ways to ‘see what happens’ and to hear what the changes made, have yielded.
Up until a while back I didn’t have access to acoustical measurement equipment.
But most of what I didn’t have was results that were really significant enough to warrant the time and effort needed to pursue measurement results.
Until recently.
As some may know I have been a proponent of ‘hardwiring’ for quite some time.
My SAA modded 800’s are hardwired and have been further tweaked with vibration management techniques.
These changes were and remain significant to the extent that they were the impetus of my recent experiments and investigations.
Which have been measured on the same equipment, which in turn yields a direct comparison of responses and behaviors from stock.
IOW using the ‘stock’ measurements and being able to directly compare these 2 sets of measurements allows for certain ‘conclusions’ which can be formulated and then further tested to see if these changes are truly ‘Better’, or not.
And even if, the ‘or not’ is the resounding result, there is Always something to be gained/learned/perceived within the measurements themselves.
This post is about one of those “gained/learned/perceived” aspects, and what I’ve gleaned thus far.
If you have ever studied the set of ‘stock’ HD800 measurements, especially the transient and distortion responses subset of the measurements (impulse, square wave, THD), you’ll see that there are some rather ‘unsavory’ indications of aberrant behavior.
But there are also some indications of absolute wonderfulness.
So this raised a question in my mind, is there a way of reducing the “‘unsavory’ indications of aberrant behavior” and at the very least maintain the wonderfulness and perchance actually improve it?
And at this point I’m not too concerned with the freq response curves.
Why?
Well mostly with no exceptions I can think of, ALL headphones exhibit a roller coaster ride of an EQ, and now with DSP EQ we have the means to begin to deal with it.
IOW what we have been hearing ISN’T anywhere near FLAT, unlike what speakers can be.
But these days with DSP based EQ this can now be addressed and is readily available for both speakers and headphones.
Thus as long as we have a representative response curve from which to work it can be corrected and brought back into ‘flat’ or at least a MUCH closer approximation of flat than ever before.
But to return to the primary subject, that being Spikes and Overshoot what I noticed when I first looked at the stock measurements were these…
pic of stock 30hz square wave courtesy of Inner Fidelity
Notice that initial overshoot spike on the leading edge?
See how fast, and what its magnitude is?
This to me says classic ringing/something is not properly damped enough, as in the entire mechanism is resonating (and at a fairly high frequency) when given an impulse signal (a square wave or single impulse, much like a fast transient).
This spike is a direct source of
tLFF all by itself.
It also can act as a mask or as a source of inner detail veiling.
So I tried ALL of the dampening mods I could find and they indeed helped in a variety of ways, not the least was a major reduction in
tLFF.
I made a set of Anax v.1 and v.1b and v.2 and wound up using shelf liner material with additional dampening over the driver clamp ring.
It was quite nice, but by comparison a bit subdued and soft when dealing with quick dynamics.
Then I went with the SAA cables and while they were being hardwired they also performed their modifications to the 800’s.
Out came ALL of the dampening material, as did the clamp ring, and the 2) 28gauge interconnect wires to the drivers along with the connectors themselves.
And there were other changes as well but these were the ones directly related to the electrical connection and those components that directly resonated in the plane of the driver.
These changes were substantially better than any of the dampening mods I had fussed with and in ways I didn’t expect, and they were Wonderful.
Wonderful enough to live with for quite a while.
Until I managed to convince a local Head Fi’r to loan me his HD650’s (thanks Big Poppa!) while mine were being measured.
Once the measurements came back I could then evaluate the differences between ’stock’ and my modded set of 800’s.
And while some of the changes weren’t ‘Better’ some were pointed in the correct direction.
So with these new insights into the measured behavior and performance I began analyzing what had changed.
But what was interesting, at least for me, were the changes that these mods had made to the overall SQ and more pointedly what these changes made in terms of all of the aspects I use to help define what truly is ‘Better’.
There truly was more there, there, and in ways that enhanced their scaleability to a remarkable degree and in ways I didn’t expect, nor suspect.
Indeed most of all of this ‘Better’ series is based upon all of these scaleability factors which were further enhanced by the initial SAA mods.
But now that I had some actual measurements I could target specific behaviors with further mods.
Specifically the transient impulse responses as shown in the square wave and impulse measurements.
So I further modified my 800’s to see if I could retain the improvements while improving other aspects of their responses.
And since this is an ongoing series of experiments, and I haven’t even been able to try the next series of mods, thus my existing results are ‘mid stream’ so to speak.
But the results I’m getting have reached what I call Gen-2 in terms of the overall acoustic net effect.
JJ
End Part 21
Next up Fuses and current draw
Make sure to share what you think of it in a couple weeks or whenever you get it hooked up.
