Latest Thread Images

TOTL HP Shootout 2: Abyss, HEK, HEX, HD800S, HD800 Heavy Mod, HD800 Stock - YGG/ROK Stack
Mar 1, 2016 at 9:39 PM Post #31 of 133
torq said:
I spent some time looking at off-the-shelf, configurable, hardware (or appliance-like) DSP solutions since making my original comment ...

snip

I did much the same, in that I went searching for parametric EQ plug-ins that would perform well, provide the degree of control and functionality that I needed.
I accumulated a whole bunch of plug-ins, mostly free or low cost ones and none of them were satisfactory for my intended end use.

I then looked at the pro level plug-ins, and quickly realized they were all designed for pro level use (well D'uh…) in that they had all this additional functionality I not only didn't want or need but they were quite complicated as well. Which raised the bar in terms of learning curve and figuring out how to actually attain the results I was looking for.
Very few had the flexibility and ability to tailor the waveform shape needed to match (in reverse) a 'known' headphone curve.
IOW they were 'awkward' in terms of compensation curve waveform shape matching at best, at worst completely unsuitable.
So I abandoned that entire approach at that time, and just concentrated on the SSBB emphasis.

That is until Sonarworks popped up.
It is MUCH easier to fuss with, albeit it's more limited in its ability to tailor the response curve, mostly because it again is a pro level plug-in mean for their stated goals which don't really address our end user requirements.

But coming up with a 'universal' compensation curve generating DSP system that can be used to help reach a 'flat' response, I would think, would be quite popular.

The capabilities would let you do the phase-reversal trick and sub-sonic-bass-boost, but you'd only have four more bands for your EQ after that - which falls a good way short of the number of adjustments that Sonarworks seems to be able to apply for some headphone profiles.

The more I think about this, though, the more I think it would be fun to put together something a bit more powerful that could handle 24/192 and more bands of EQ, along with phase reversal, and then provide them as a low-cost module to go between source and DAC that could be tweak-able, or offered "pre-configured" for one or more sets of headphones. 
Click to expand...
One of the tricks, or complications seems to be in being able to 'insert' a program or process between the output of any player and the output data stream.
Especially with windows complicated audio path, but even in a mac or unix environment this might prove to be problematic.

Which is why utilizing the DSP function within a player has been the de-facto method being used thus far.
I suppose a 'standalone' device could be physically inserted between the computer and the dac but that also raises yet further complication in terms of interface and control and I/O configuration etc.

If there was a plug-in that could use a 'translated', stock or custom headphone response curve AND allow for further adjustments (the harmon curve for instance) along with the rest of the tweako stuff headphone usage could readily use such as SSBB, PRT, just to name a couple more …
THAT would be killer. :atsmile:

JJ
 
Mar 2, 2016 at 1:37 PM Post #32 of 133
torq said:
 

Abyss – Revisited

 
So, a week later, the “bass distortion/seal issue” I found, and that other’s heard/felt, with the Abyss turns out to have been down to driver damage.  Specific cause unknown, but since that particular set of headphones is part of a loaner program and get shipped all over, that seems entirely reasonable and is not something I consider a reflection on the quality or solidity of the headphones in general.
 
And … just that one week later … with the fault confirmed …
 
I wound up purchasing my own set of Abyss AB-1266, which I have on my head right now. 
 
There is a night-and-day difference in the bass performance here – and no more distortion or flutter.  I don’t need to break the seal to get the best bass performance; it’s cavernous and rich and detailed and visceral in ways I have never heard from any headphone.  While I’m sure the effect is largely psychoacoustic, it is as if I can feel the bass in my chest, as when I run my speaker system, rather than just hearing it.
 
I’ll have to update my original post with this new information, but with the bass where it is supposed to be these are an even more impressive headphone and an entirely different overall experience.  If this is what I had heard on my initial trial, I'd have ordered them there and then.
 
As a result, we may have to re-do this shoot-out now to give everything a fair shot once again.  Ideally we’ll time that to have the LCD-4 on hand, though that may take a bit!  Perhaps we can include some more people as well ... :wink:
 
Anyway, I’ll update my original post with proper impressions on the Abyss later today, so as not to lead anyone reading astray unnecessarily.
Click to expand...

Added a note to the first post. Good to hear that you are enjoying them. Look forward to your updates.
 
Mar 3, 2016 at 7:11 AM Post #33 of 133
torq said:
   
I'd want to know more about the clocks on the HiFiBerry Digi+ before I'd look into it more seriously.  As it stands, I've personally never managed to hear a difference between properly-done TOSLINK and coax.  If Mr. Moffat says coax is better, I'd believe him - but that doesn't mean I can personally tell the difference.  And, as with most things, the devil is in the (implementation) details.
Click to expand...


This article discusses the clocks in the Raspberry Pi:
https://hifiduino.wordpress.com/2014/11/13/raspberry-pi-b-digital-audio/
 
 
But I'm not sure how much (any of) this applies directly to the Hifiberry Digi+ boards. I could only find this reference on their support forums:
"The HiFiBerry Digi outputs an SPDIF signal. There is no "master clock" on SPDIF."
 
As for Toslink vs Coax, from what I see it is important to (1) use a good quality, glass optical cable or (2) have a Coax out connector with galvanic isolation. In both cases any noise picked up upstream would be filtered away, yielding only the relevant bits to the DAC SPDIF input. Here's a nice read on galvanic isolation:
http://www.epanorama.net/documents/groundloop/digital_audio.html
 
Mar 3, 2016 at 3:02 PM Post #34 of 133
Interesting reading, thanks!
 
The Raspberry Pi itself I'm very familiar with as I've used them in a number of non-audio related projects, and its clock implementation is not really targeted at audio.  It's certainly not something I'd use that way at least - I'd want to move all of the audio-centric hardware to a dedicated, task-specific, implementation (which is what the Digi+ is one version of).
 
From your links, it shows that the Digi+ is using a standard implementation of the WM8804 interface/transceiver.  Per the datasheet, given a stable 27 MHz external clock, it's internal PL is good to about 50ps.  Much better than the Pi can handle on its own.  So that's not bad.  It's not in the same class as more recent audiophile products, but certainly beats the specs that some papers have suggested are all that's needed to reduce jitter below audible thresholds.
 
Still, to put together a decent streaming implementation, assuming I'm good with an existing audio-centric stack/distribution for the Pi, you're looking at~ $100-150 in parts (including a case and a decent power supply, and a few hours to get everything set up.  Factoring my time, that comes out right at the same level of investment as an Aries Mini or other, basic, commercial streamer.  So, for me, it's easier just to go with the commercial solution - and enjoy all the incremental features that come with someone else doing the work on the software side.
 
Mar 4, 2016 at 1:55 AM Post #36 of 133
gutb said:
  So the Abyss completely obliterates the HD800, as expected.
Click to expand...


Stock, and at 5x the price, sure.
 
With @johnjen's hardware mods, and some software tweaks, the consensus was that his HD800 was the best headphone of the evening.
 
Since I can't buy John's headphones, and I value not having to have a computer in the mix for all my listening, the Abyss is the best available option.  And it is truly superb.  With my new pair I'd say the Abyss just ousts John's cans in the bass and loses a little in detail to them in the higher octaves.  But then John's HD800 are so far above the stock HD800 it's not funny.  More than one of us opined that he should start offering his mods commercially.
 
But yes, ignoring price (which I'm lucky to be able to do), the Abyss beats the stock HD800 fairly easily.
 
Mar 4, 2016 at 7:18 AM Post #37 of 133
The one thing that really struck me when I 1st heard the Abyss's, was they changed the harmonic content of all 'voices'.
I attribute this to the fact that they are an asymmetric design.

Yeah they do have 'kick' and a plosive attack when the music calls for it and sometimes even when it doesn't.
It certainly is entertaining and 'fun'.

So to me it boils down to the strengths and weaknesses of each of the 3 'types' of drivers used in speakers and headphones.

Planars can move air, there is no question about that, but they often lack finesse and inner detail, mostly because the entire surface area of the driver is trying to respond to the entire signal regardless of frequency.
This tends to add some degree of smoothing or a bit of fuzz to the small signal dynamics since the membrane is large and high freqs tend to have fewer 'problems' with smaller surface areas.
I noticed this with the HE1K's where they are really smooth and enticing, but lacking in inner detail and fine resolution.
Audeze's follow suit, albeit in their own way, as well.
It’s a planar thing.

And dynamic drivers do have more mass than planars so they tend to be slower in response, especially when compared to e-stats.
But they have an advantage in that they act, or tend to act like pistons so the break up (non pistonic response) of the radiating surface is less of a problem, especially in headphones where the excursion/amplitude of the radiating surface is very low.
This reduced excursion factor also 'helps' in the response time and control of the driver at higher frequencies.

And there are some tricks in how to get a single driver to 'localize' hi freqs to a smaller area, which I figure Sennheiser is well aware of.
But pistonic action, especially when it's coupled well, can have as powerful an amount of bass, but it can also have better inner detail and resolution and extension into the subsonic region than a planar and certainly better low end extension than e-stats.

And an asymmetric design (which applies to all planars but to differing degree's and in different ways) means, to me, it will, or can be a pleasant and fun experience, but ultimately not an accurate transducer.

So in my investigations and experiments the 800 was an optimal choice, and the results I'm achieving, which was only partially in evidence at the meetup, are most gratifying.
And the 800's lend themselves quite readily to tweaking, as can be demonstrated by all of the mods and attempts at resolving the few problems they do have.

It's like a dilemma where they are so good at so much, but for a very few 'problems', which tend to stick out like a sore thumb, which results in a disproportionate 'frustration factor'.
This is, at least it seems to me, why they are being fussed with so much, because they hold up such promise, if only…

And yeah many simply don't want to tweak or modify them for a variety of reasons, but for those who do, the results can be quite gratifying.

And lastly it has ALWAYS been a desirable design goal to use just one driver for the entire audible spectrum.
Regardless of it being speakers or headphones.
Case in point…
Quad esl 57's when they were tweaked, were simply magical, and nothing could touch them, except for they had very limited dynamic range, no real bass and little top end extension, and could/would die if pushed to hard, but in their day (and even today) they sucked you into the music like nothing else could.
It is a single e-stat driver that emulated a radiating sphere.

And in a similar vein dynamic drivers when used in headphones, for a variety of reasons, are a more ideal driver choice than the other two, IF the full audio spectrum+, and maximally accurate resolution are the primary design goals.

At least these are the conclusions I have arrived at based upon the experiments and results I've conducted thus far.

JJ
 
Mar 4, 2016 at 1:21 PM Post #38 of 133
johnjen said:
 
Click to expand...

The one thing that really struck me when I 1st heard the Abyss's, was they changed the harmonic content of all 'voices'.
I attribute this to the fact that they are an asymmetric design.

Yeah they do have 'kick' and a plosive attack when the music calls for it and sometimes even when it doesn't.
It certainly is entertaining and 'fun'.

So to me it boils down to the strengths and weaknesses of each of the 3 'types' of drivers used in speakers and headphones.

Planars can move air, there is no question about that, but they often lack finesse and inner detail, mostly because the entire surface area of the driver is trying to respond to the entire signal regardless of frequency.
This tends to add some degree of smoothing or a bit of fuzz to the small signal dynamics since the membrane is large and high freqs tend to have fewer 'problems' with smaller surface areas.
I noticed this with the HE1K's where they are really smooth and enticing, but lacking in inner detail and fine resolution.
Audeze's follow suit, albeit in their own way, as well.
It’s a planar thing.

And dynamic drivers do have more mass than planars so they tend to be slower in response, especially when compared to e-stats.
But they have an advantage in that they act, or tend to act like pistons so the break up (non pistonic response) of the radiating surface is less of a problem, especially in headphones where the excursion/amplitude of the radiating surface is very low.
This reduced excursion factor also 'helps' in the response time and control of the driver at higher frequencies.

And there are some tricks in how to get a single driver to 'localize' hi freqs to a smaller area, which I figure Sennheiser is well aware of.
But pistonic action, especially when it's coupled well, can have as powerful an amount of bass, but it can also have better inner detail and resolution and extension into the subsonic region than a planar and certainly better low end extension than e-stats.

And an asymmetric design (which applies to all planars but to differing degree's and in different ways) means, to me, it will, or can be a pleasant and fun experience, but ultimately not an accurate transducer.

So in my investigations and experiments the 800 was an optimal choice, and the results I'm achieving, which was only partially in evidence at the meetup, are most gratifying.
And the 800's lend themselves quite readily to tweaking, as can be demonstrated by all of the mods and attempts at resolving the few problems they do have.

It's like a dilemma where they are so good at so much, but for a very few 'problems', which tend to stick out like a sore thumb, which results in a disproportionate 'frustration factor'.
This is, at least it seems to me, why they are being fussed with so much, because they hold up such promise, if only…

And yeah many simply don't want to tweak or modify them for a variety of reasons, but for those who do, the results can be quite gratifying.

And lastly it has ALWAYS been a desirable design goal to use just one driver for the entire audible spectrum.
Regardless of it being speakers or headphones.
Case in point…
Quad esl 57's when they were tweaked, were simply magical, and nothing could touch them, except for they had very limited dynamic range, no real bass and little top end extension, and could/would die if pushed to hard, but in their day (and even today) they sucked you into the music like nothing else could.
It is a single e-stat driver that emulated a radiating sphere.

And in a similar vein dynamic drivers when used in headphones, for a variety of reasons, are a more ideal driver choice than the other two, IF the full audio spectrum+, and maximally accurate resolution are the primary design goals.

At least these are the conclusions I have arrived at based upon the experiments and results I've conducted thus far.

JJ
Click to expand...

 
 
JJ, I love these discussions, they really make us go really deep into the listening experience. I hadn't really thought too deeply about the mechanical differences of the drivers before.
 
I agree that the HD800 is more versatile in its ability to change its sound as you have proven and is more resolving at the expense of being delicate when stock or tuned that direction - losing its texture/dynamics in favor of nth degree of detail. But I do find that fully modded, your HD800 sounds very close to the Abyss with minimal differences. That is a very good thing considering the price differences. In taking on the dynamic impact of the Abyss, the HD800 does lose some of the nth degree of detail just like the Abyss where the detailing is very close, at least to my untrained ear. As for the planer fuzz, I get it with the HEK/HEX/Audeze with a softer more euphoric SQ. However, the Abyss is surprisingly transparent and lacking of that fuzz. But its dynamics do diminish that amount of delicate detail that pops forward just embedding them into the music.
 
I should also point out the HD800/700 have a treble glare that acts much the same as the planer fuzz. Toggle the SonarWorks plugin on and off to hear how much detail is hidden behind that treble glare. The HD800 sound stage is also artificially wide which enhances the detail retreaval too which gets reduced a bit with the full mods bringing it close to the Abyss sound stage size which is reasonably large too. The Abyss can be tuned to be very wide as well by fiddling with the ear cups, but the HD800 is still the champ. However, in my opinion, your fully tweaked HD800 and the Abyss sound very real to life with a realistic full sized sound stage and lifelike dynamics and without artificial pushing out of detail. There are times that I am digging into the music and can appreciate the stock HD800 analytical powers, but for listening to music and enjoying, the tweaked HD800 and Abyss with lifelike detailing is what puts a smile on my face.
 
The real question is, if the Abyss starts where the HD800 fully modded/tweaked has gotten after years of research and tweaking, where could you take the Abyss if that was your platform. How much more does the Abyss have to give when optimized? Will there be a Franken-Abyss in your future - it has the appropriate look? 
evil_smiley.gif
 
 
Mar 4, 2016 at 9:20 PM Post #39 of 133
The Abyss certainly does look the part of frankephones or cybersonics run amok. :atsmile:

As for tweaking them…
Unless someone wants to give me a pair to fuss with, I doubt it.
5k$ is way to much to risk frack'n them up.
Having said that it might be possible to adapt the techniques I use on the 800's on the Abyss's but it would take much fussing to figure all of that out.

And planars usually don't easily lend themselves to tweaking, other than removing parts.
The drivers are somewhat even more fragile than most drivers (except e-stats) so the cost of replacement parts would need to be factored in as well.

And the 800's very easily lend themselves to the sorts of tweaks I am using, much more so than just about any other headphone I'm aware of.
They are built to be easy to work on and my current mods are fairly easy to implement as well, so the difficulty level is low in fussing with them.

But the asymmetric design of the Abyss is built in and it isn't something that can be 'fixed'.
What this means is the recreated waveform shape is, by definition, going to be different than the original signal.
And there is no way to correct for this type of 'coloration' or deviation away from the original signal.

So I see it as it would be a little like the lipstick on a pig scenario, since their basic nature isn't going to change.
And for me if the drivers simply can't re-create a 'proper' harmonic structure to begin with, then all of the adjustments and corrections which follow will never be able to 'fix' that deficiency.
And a proper harmonic structure is for me a key aspect in terms of being able to successfully tweak any system, since what I aim to do, on the whole, is to remove or reduce un-harmonic responses from the system such that the system gets out of its own way.
And much like if a digital track has been reduced in resolution, that information is simply lost and can't be brought back into the equation, no matter what techniques are employed to try and recover the missing information.
But allowing the existing information to not be 'blocked', in the first place, (getting the system out of it's own way) so that more (all?) of the inherent information CAN be presented to us, is my goal and means and method of tweaking.

And while the Abyss certainly can be entertaining in their own right, one of my goals is to have a 'one size fits all' system where any type or kind of music is recreated with as much of a compelling nature to it, as is inherent to the music in the first place.
IOW it won't matter what music is being played, the presentation is just as compelling and IMPERATIVE such that I get sucked into it regardless, even if it is a terrible recording with all kinds of distortion and can sound 'ugly' on some systems.
I don't see how the Abyss's can achieve my goal, regardless of what mods or changes I or anyone else could make.

And while my 800's may not fully achieve my lofty goal, I am much closer to achieving it now than I ever have been, and I still have a few tweaks to experiment with which are further along the lines of those I have already made.

And so it goes… :atsmile:

JJ
 
Mar 4, 2016 at 11:34 PM Post #40 of 133
johnjen said:
...Planars can move air, there is no question about that, but they often lack finesse and inner detail, mostly because the entire surface area of the driver is trying to respond to the entire signal regardless of frequency.
This tends to add some degree of smoothing or a bit of fuzz to the small signal dynamics since the membrane is large and high freqs tend to have fewer 'problems' with smaller surface areas.
I noticed this with the HE1K's where they are really smooth and enticing, but lacking in inner detail and fine resolution.
Audeze's follow suit, albeit in their own way, as well.
It’s a planar thing...
Click to expand...


and this along with their weight is why i'm not a planar kind of guy
 
Mar 5, 2016 at 1:35 AM Post #41 of 133
up late said:
and this along with their weight is why i'm not a planar kind of guy
Click to expand...

well the abyss and audeze are stupid heavy...
but not the HEX, HE560 etc...they are quite nice if you havent tried them.
 
 
I like my HEX and HD800's, both are nice and light, yet I still prefer the comfort of the HEX slightly more.
330grams for HD800
399grams for HEX
 
 
LCD4 680grams WHAT?! yes... it's stupid. so was that carbon fiber really needed or what?!
 
Abyss is 620grams WHAT!? yes... neck pain as well.
 
 
I can't stand anything much over 400grams. I actually have a neck condition and any weight on my head for extended periods actually is painful. 
 
I can listen to the HEX longer than any headphone as it's got a relaxed yet detailed presentation and is uber comfortable.
 
the HD800 would be OK if the stock sound signature wasn't like screw drivers in my ears. With sonarworks it's really nice! 
 
Mar 5, 2016 at 3:08 AM Post #42 of 133
i haven't tried the hex or he560 but any reduction in the weight of a headphone can only be a good thing imo. however, it wasn't the weight alone that led me to conclude that planar magnetic cans aren't for me.
 
Mar 6, 2016 at 7:05 PM Post #43 of 133
So, the main take-away I'm gathering from this thread is that the Abyss sounds the best. Great hopes were had for a modded HD800, but planars still beat dynamics, and nothing changes that fact.
 
Mar 7, 2016 at 12:12 AM Post #44 of 133
gutb said:
So, the main take-away I'm gathering from this thread is that the Abyss sounds the best. Great hopes were had for a modded HD800, but planars still beat dynamics, and nothing changes that fact.
Click to expand...


The take away for those that can't afford the abyss is that the modded and tweaked hd800 reaches abyss SQ levels at a fraction of the cost. But we are talking about a stock abyss so I wonder how far it can go with similar tweaking.
 
You must log in or register to reply here.

Users who are viewing this thread

Total: 2 (members: 0, guests: 2)
Back
Top