Regarding Stax - I don't own, nor have I tried the realizer. But I have a couple of thoughts..
 

• Very low distortion - we are talking about 0.1-0.3% 20-20kHz for the basic 207
• FR without extreme peaks and dips that are hard to equalize, as seen with many dynamics
• Treble extending all the way to 20 kHz, which add to spacial cues. 'Stats oftenmost have very impressive high treble extension
• Last, and most importantly, the 'stats I have tried (especially the lambdas) sound like speakers next to your ears. Try comparing that to hd800, which by itself tries to emulate a proper soundstage by displacement of the driver. I guess one would (ideally) want 2 uncolored planar wave units right next to the ears with the realizer for optimum results.

 
Just some thoughts.

The electrostatic headphones are lighter to wear then regular driver headphones.  I like the stax not because of sound but because of comfort.  They do sound good though.  When you wear headphones for 10 hours a day because you HAVE to then some priorities change.  With that said Im wearing dt770 right now because I do need a closed back headphone until I get into a quieter room.  I will be off the realizer as heavy usage in 2 weeks though.  The dt770 is pretty comfortable but they do hurt (sore around the head where they press) when you wear them so much. 

I'd think Stax do well because of their good transient response in general:
- fast rise time, discussed above as top end extension
- fast decay: typically not significant ringing over most of the band

They have low distortion and very good / clean bass. They're extremely chameleon like, which is the most suitable for a virtualisation.

Having said that, having heard (not compared the two directly though and not used in the same rooms) both stats (sr009) and orthos (pm1) with the realiser/HPEQ on, non-stat phones can do well too.

Arnaud

I have a rather serious question that I would like to hear other people's input.
 
Before I acquired the Realiser, I had owned and auditioned many TOTL headphones (including the legendary Orpheus and SR-009/BHSE), but I came to the conclusion that HD800 and flagship orthodynamic headphones are better suited to my tastes than their electrostatic counterparts due to the latter's relative lack of weight and viscerality in sound. In other words, even the best electrostatic headphones sounded too polite that made music listening not as engaging as dynamic headphones. My current favorite is the HD800, which scales indefinitely with better source components and headphone amps, followed by HE-560 (Hifiman's new headphone) which IMO is the best orthodynamic headphone I've heard to date (better than LCD-X and LCD-3 to my ears).
 
It was after I had settled on these headphones that I acquired a Realiser A8, simply out of curiosity. Imagine my surprise when it turned out to be the most shocking, game-changing, revelatory piece of audio equipment I had purchased throughout my entire journey into Head-Fi. Unlike other gears that came and went by in my system, I can't possibly think of selling the A8 as I've been completely spoiled by the way it presents 3D holographic music.
 
That being said, I've read in many articles here that electrostatic headphones are inherently a better match with the Realiser than any dynamic headphones out there, that even a low-cost vintage Stax headphone could perform just as well or even outperform a high-end HD800 setup (which I happen to own), when both are fed by the Realiser. Just when I thought the HD800 was a permanent keeper, and reflecting back on how not even the best electrostatic headphones out there failed to captivate me, I began to wonder: will the Realiser change all that?
 
As I've said the main reason I prefer my HD800 and other TOTL dynamic headphones compared to stats is because the latter doesn't sound as engaging, as they lack the physical, visceral weight of sound that conveys a sense of being one with the music. Electrostatic headphones just don't make me tap my toe like the dynamics do. But if what people are saying is true about the Realiser, will it change all that? If Realiser inherently works better with stats due to some design or technical reasons, what are they? Now that I own the Realiser, should I sell my HD800 system and replace it with an electrostatic system?
 
By the way, for the record, I preferred the sound of SR-007 mk1 to SR-009 over the BHSE. The BHSE/SR-007 mk1 was the best electrostatic setup I've heard (prior to auditioning the Realiser), followed by the Orpheus, than the BHSE/SR-009. The BHSE/SR-007 mk1 sounded most natural and balanced out of these three. Even so, a high-end HD800 setup, with its massive soundstage, full-sounding mids and bass, and the ability to convey pinpoint imaging and extract microdetail won me over. But like I said that was before the Realiser, so I wonder if things will be different now that the Realiser is working its magic in the upstream...

Just got some HE560's so gonna see how they fare up to HD800. While the Senns are super airy and accurate, there always seems to be that tiny extra sibilance/harsness.. probably need to spend few more hours experimenting with the manual eq routine..
 
As it happens, the store I bought my HE560 is one of the best and friendliest around with a great new listening room rotating often some new stuff. I think I'll try to do some PRIRs today from Goldnote A-6 XL's.. the owner is also quite interested in trying Realiser.. 
 

hekeli said:

  Just got some HE560's so gonna see how they fare up to HD800. While the Senns are super airy and accurate, there always seems to be that tiny extra sibilance/harsness.. probably need to spend few more hours experimenting with the manual eq routine..
 
As it happens, the store I bought my HE560 is one of the best and friendliest around with a great new listening room rotating often some new stuff. I think I'll try to do some PRIRs today from Goldnote A-6 XL's.. the owner is also quite interested in trying Realiser.. 

 

Click to expand...

 
What a coincidence, I own the HD800 and HE-560 too along with the Realiser too.
 
Unlike you, I always prefer the HD800 to HE-560 with or without the Realiser. In the past I used to dislike the HD800 because of the very sibilance/harshness you mentioned, but with the right source and amp (and cabling too), this could be remedied. In fact, I do not hear sibilance or harshness in the HD800 with my current system, which comprises of a NAD M51 DAC and Allnic HPA-3000 amp.

Ok I ended up doing some HPEQ analysing with HD800... if you are not into tech ignore this post. 
 
Check out how the different xfact coefficients work... (read A8 manual HPEQ OPTIMISATION chapter if you have no clue what these are)
 
This is one of my Siba PRIRs... I think the default 1/0.7/0.3 setup (or any for that matter) doesn't handle the 6k peak / 8k null gracefully. Maybe this is the source of the slight harshness.
 
Another thing I noticed is that it really doesn't touch anything under 1kHz! That explains why especially the subbass is quite different on many phones regardless of HPEQing. TH900 had way too much bass, HD800 could use a little bump..
 
I'm going to try to make a custom hpeq filter to use in software.. since I can get both PRIR and HPEQ data separately I can let REW calculate some stuff from it..
 

Hekeli, you're the one who kindly pointed out how to recover impulse response of PRIR and HPEQ filters using hdmi out, aren't you?
I am still debating about buying a realiser, but one of the motivation was to use it for properly compensated headphone measurements. You're proving that's quite doable indeed .

Also, fantastic investigation on what the HPEQ is doing! I don't quite why the 8kHz is so accentuated with the default HPEQ. Assuming 1kHz is the ref. point, I get why the correction is such until 6kHz, but I am at a loss with the 8kHz.

It's as if the HPEQ system was analyzing the phone response as a set of troughs and peaks and coming up with proper peq filters but the compensation values seem to be based on local amplitude / Q factor than trying to achieve a target flat curve like reqw does for instance.

I am looking forward to read more about your findings! PM is fine if that's too boring for others .

Arnaud

Just came back from a quick trip to LA and San Diego to get PRIRs done at AIX and two places arranged by darinf (Acoustic Zen and PBN Audio).  I have had the Realiser since last February, but this is the first time getting custom PRIRs done.  I have enjoyed using one of Darin's Acoustic Zen PRIRs that he shared on the PRIR dropbox folder, but once I heard the custom PRIRs there is no going back.
 
Some quick notes:
 
1)  Custom PRIRs are essential to get the Realiser to sound its best.  You can get somewhat close if you are lucky enough to find someone else's PRIR that sounds good, but the room/speaker combo has to be really good and their ears need to be a close match to yours. If you do find one of those, it can sound better than a custom PRIR done in a poor room or with poor speakers (or measured improperly).  But if you get a custom PRIR done in a good room with good speakers, it is more true in frequency response, localization, etc.
 
2)  If you are local to SoCal or have the time and means to get there, I highly recommend going to AIX and one or more of the places that Darin can arrange.  
 
3)  Lorr from Smyth and Mark from AIX were super friendly and patient in making sure the PRIR was done correctly.  Mark has a fantastic room at AIX and it is a great PRIR to use for mixing if you do your own (naturally as that is what the room is used for).  Lorr and Mark did a great job with my PRIR at AIX.
 
4)  Darin was also super friendly and patient.  He has clearly mastered the art of doing PRIRs especially the method of doing 7.1 PRIRs with just two speakers.  The 7.1 PRIRs sound as if there were true 7.1 setups.  He has access to some really great systems.  If you like some of the Acoustic Zen PRIRs on the shared dropbox folder (as I did before), then I highly recommend going to Acoustic Zen with him to get a custom PRIR.  I also like the PRIR I got at PBN Audio.  Darin has access to a number of different sounding systems, so there should be something that is close to your liking.  Darin did a great job with the two PRIRS he did for me.

arnaud said:

It's as if the HPEQ system was analyzing the phone response as a set of troughs and peaks and coming up with proper peq filters but the compensation values seem to be based on local amplitude / Q factor than trying to achieve a target flat curve like reqw does for instance.

Click to expand...

 
Well I asked Mr Smyth, it's probably ok to paste here..

Quote:

HPEQ has always been a bit of a black art for us, and I think we really designed it around the Stax headphones that we developed the system with.
1. It's not easy to correct low frequencies, retain linear phase and maintain low latency - which is way we don't bother - as you can see.
2. You can try altering the degree of correction by altering the inverse compression factors of the four bands in the auto-eq.
3. Also, we have found that the best way of optimizing the HPEQ is to use the manual EQ method. For example even with the Stax I normally find the final auto-
HPEQ to be too bright, and therefore I use the manual eq method to modify these coefficients. It's quite tedious, since you need to go around all the frequenc
y bands at least twice and possibly three or four times - this is because the bands overlap somewhat.
 

Click to expand...

 
So yeah the auto eq is not a one for all solution. Not even manual eq especially for bass due to a large 500hz band adjustment. So more flat and less peaky the phones are really affects the outcome. Of course you can overcome this slightly with some LFE volume adjustments etc.
 
I'm going to do some super tedious manual eq testing and see how it does in the graphs... I think in theory you could do something even without listening, just need to count where you are in the bands and adjust using the raw headphone measurement as reference.

If anyone is local to the DC area and has a nice speaker setup (preferably in a nicely treated room with room correction), and doesn't have a Realiser, i would be happy to talk about doing a custom 7.1 PRIR for you in exchange for letting me do my own PRIR there. That way you can test the Realiser. If the Realiser is too much money, you could also use your PRIR with Darin's Out of Your Head software subject to his price to make the custom preset ($149 I believe). Doing my own custom PRIRs with AIX and Darin was fantastic and makes me want to do even more! Please PM if interested. Thanks.

hekeli said:

  Ok I ended up doing some HPEQ analysing with HD800... if you are not into tech ignore this post. 

 
Check out how the different xfact coefficients work... (read A8 manual HPEQ OPTIMISATION chapter if you have no clue what these are)
 
This is one of my Siba PRIRs... I think the default 1/0.7/0.3 setup (or any for that matter) doesn't handle the 6k peak / 8k null gracefully. Maybe this is the source of the slight harshness.
 
Another thing I noticed is that it really doesn't touch anything under 1kHz! That explains why especially the subbass is quite different on many phones regardless of HPEQing. TH900 had way too much bass, HD800 could use a little bump..
 
I'm going to try to make a custom hpeq filter to use in software.. since I can get both PRIR and HPEQ data separately I can let REW calculate some stuff from it..
 

Click to expand...

OK, now we are getting some real answers as to why the Realiser is recommended for electrostatic headphones: verified by hekeli's measurements and confirmed by Dr. Smyth himself, the HPEQ doesn't even attempt to correct for much of anything below 1khz.  If you look at these measurements from Innerfidelity you can see that the electrostatic headphones have very linear and neutral FR from around 1khz and below (if anything they are a bit anemic in that range, but that's probably OK because most speakers/room interactions will create excess energy here anyway):
 
Stax SR-009:

StaxSR-007:
 

Stax 404:
 

 
Audeze LCD-3:

Compare those measurements with the best that the dynamic drivers have to offer:
 
 
Sennheiser HD800: (actually not too bad, but not quite as flat as the above)
 
 
​
Beyer T1:
 
 

AKG k812:

In addition, the machine was designed using Stax headphones so it would make sense that it's optimized for such.

Yeah.  With a system like this you are starting flat. It isnt about adding your own nuance and color via different components.  Its about getting those rooms accurate.   About rooms again.  I would much rather hear a $70k room with a $6k speaker/amp/dac then a $70k speaker/amp/dac in a $6k room.  When looking for rooms to get impulsed look for the serious build outs.  Thats what you are paying for.

Hi guys,
 
There seems to be a bug in the ONE measurement procedure. 
 
The left channel is virtualized at center and moves to the left when turning head to the left. Right speaker is at -90° and moves to -120° when turning head left.
I've checked all cabling and settings but can't find what's causing this misbehavior.
 
I've contacted Lorr and he acknowledged the bug but I don't get any response from Smyth UK. Tried several times over the last year (!).
 
Maybe someone knows more about this bug and/or has contact to Smyth UK? Right now my Realiser is sitting next to me, collecting dust
 
Thanks, Markus

markus7 said:

  Maybe someone knows more about this bug and/or has contact to Smyth UK? Right now my Realiser is sitting next to me, collecting dust

Click to expand...

 
It's well known and to be fixed in the next firmware. Only way to use it is to downgrade to old firmware, which obviously is a silly idea. I've emailed with Mike Smyth many times, there is supposedly a new firmware coming really soon, he said I could beta test it, but haven't heard anything in a while.. maybe I'll ping him again.
 
Surely you can make a two speaker measurement?? Or did you steal only one from somewhere.