• Welcome to The Audio Annex! If you have any trouble logging in or signing up, please contact 'admin - at - theaudioannex.com'. Enjoy!
  • HTTPS (secure web browser connection) has been enabled - just add "https://" to the start of the URL in your address bar, e.g. "https://theaudioannex.com/forum/"
  • Congratulations! If you're seeing this notice, it means you're connected to the new server. Go ahead and post as usual, enjoy!
  • I've just upgraded the forum software to Xenforo 2.0. Please let me know if you have any problems with it. I'm still working on installing styles... coming soon.

Surround Sound in your Headphones

Razz

Well-Known Member
So I was at an audio show in Irvine CA yesterday and came across this guy who has made software for your headphones that makes your headphones play "surround sound". It was very impressive. It also breaks down 5.1 and 7.1 music. As of now, they are only PC and Mac based but he says they are working on other platforms like IOS and Android.
Flint, I thought you might enjoy this as you listen to a lot of music from your laptop right?

Go check out the demo.
It priced right too!

https://fongaudio.com/demo/
 
Well... I'll check out the demos, but I've developed a very strong opinion based on experience, science, interactions with others, and more that convincing surround sound in headphones, regardless of the power of the DSP used, can be accomplished unless some sort of head tracking technology be implemented which compensates for even the slightest movement of the listener's head while processing a surround signal. So, these demos have grown increasingly unimpressive with each successive "breakthrough" claimed by the inventor.

I've written about the reasons why it simply isn't possible to take a normal set of headphones and process the signal and get convincing or even moderately effective surround effects many times on the forum, sharing the research, physiological issues, and perception issues which make it impossible.

So, now that I've completely been a horrible forum buzzkill jackass over this, I'll make time today and check out the demo and write my impressions.
 
awwwww... good! Looking forward to what you think Flint. I remember you talking about this in the past but I found this one very impressive. I don't care for them trying to duplicate a speakers sound. I think that's BS, but the surround effect is the best I have ever heard with headphones. The Pink Floyd demo is pretty f'n cool!
 
Don't visit the headphone section much, so I just saw this. Being at work, I couldn't just sit and listen to the whole thing, but skimming through it, got to hear most parts of the demo. That was pretty cool. There was definitely some depth to it. Maybe not as much as actual speakers placed around the room, but the sound didn't sound like it was coming from the headphones, unless they turned the processing off.
 
Here’s my impressions of the Out of Your Head DSP software for headphone listening after listening to the Demos and downloading the free trial with it’s very limited 2-minute trial time for the various presets.

Setup
First off, I love the sound of my various headphones. For the listening tests I used my Sennheiser HD650 headphones and my Shure SE846 headphones with Grace Design DAC/headphone amp to ensure the most accurate reproduction on my end.

Preface
Unfortunately, only two of the demo tracks were content I was very familiar with – the Master and Commander scene in 7.1 surround, and the Rebecca Pidgeon “Spanish Harlem” tune in 2-channel stereo. I completely ignored all the marketing content and descriptions of what the processing does and just listened. I put in about an hour specifically on the demo tracks then played with the trial version of the DSP with content I know ever better and discovered quite a bit about what the DSP is doing by listening before reading about what the inventor claims it did. I also avoiding looking at the screen during playback of the demos, specifically the video-based demos, as our brains are VERY susceptible to the power of suggestion when it comes to what we hear. I didn’t want to be tricked by visuals or words into thinking I was hearing something my ears and auditory processing was, in fact, not deciphering on their own.

Initial and Immediate Impression
My initial impressions were that the stereo processing is interesting, but the surround processing is confusing and constrained, at best. More on those two impressions later.

So anyone here can also experience what I did, I will specifically use the two demos I am familiar with: Rebecca Pidgeon’s song and the Master and Commander scene. From the demos I could not ignore that the overall tonality and balance is changed completely and there is a sense of a hollow-tube “honky” tone that cannot be ignored. The high treble was altered significantly in both demos from the original content I own (assuming they are using the same source content I have) and the deep bass is vastly different – not even close to the same depth and levels as the original. The midrange, as mentioned, has a strange “honky-ness” to it which I can only describe as something like listening to a speaker at the end of long narrow corridor. It isn’t as bad as actually listening to a speaker at the end of a corridor, but that is the characteristic I heard.

7.1 Surround Sound Simulation
With the Master and Commander soundtrack, I was quick to hear how the DSP added ambience to the sound. That said, much of the impact was reduced and the dynamics seems lacking by comparison to the original soundtrack on my BluRay disc. I also couldn’t help but notice the difference in frequency response balance, which for me was distracting. As the scene progressed to the explosions the sound from the processed signal was significantly muddier, sloppier, and more difficult to make out any detail at all. It was just a massive confused stew of noise and rumbling without any detail I expect to hear either in my HT or over my headphones. Basically, the processing seemed to decrease the dynamics, muffle the detail, and alter the timber of the original content. I don’t know why anyone would want that.

By comparison, enabling the Dolby headphone processing in my PC software-based BluRay player resulted in the requisite added ambience but the balance of the frequency response remained essentially the same as the original, the dynamics were only slightly affected, and the detailed clarity was affected considerably less. Based on the listening experience, I assumed the Out of Your Head filter was, in fact, modeling specific speakers and rooms the inventor believes are “ideal”. I have complaints about this approach, both in theory and in practice, I’ll address below.

Finally, using the demo where a voice announced the channel as each of 7 channels were played I found things very strange. Everything that was said to come from in front of my head seemed to come from over my head – not in front of me. Everything intended to come from behind me seemed to come from either the left or right of me. In other words, what should have been a flat circle of speakers around me was perceived as an arc over my head and extended out to the left and right when it was supposed to be “rear” speakers. The DSP succeeded in taking the sound out of the center of my head, but it failed completely in making sound seem to come from the speakers in a real home theater. The same demo using Dolby headphone processing was better at placing speakers where they should be, but it was still far from perfect. So, yes, the processing influenced my perception of space, but neither DSP was realistic – to me at least (and everyone is different).

2-Channel Stereo
Using the demo of Rebecca Pidgeon’s, “Spanish Harlem,” I was less disappointed than with the attempts at 7.1 surround. The processed demo pulled the sounds out of the center of my head to a virtual space above my head. The frequency response obviously changed and the dynamics flattened, both of which were very distracting and disappointing, but the soundfield effect was somewhat pleasing. I pondered if I could make a habit of listening with this sort of processing as I would sacrifice details, dynamics, and the frequency timber I like from my headphones to experience the music from outside of the center of my head. That said, when I played a dense recording using the trial version of the software in stereo, the loss of detail and dynamics wasn’t acceptable to me. Like I said, I love listening with headphones and rarely feel I am missing anything when doing so.

Again, I got the impression the DSP was modeling a set of speakers and thus changing the frequency timber, phase shift, group delay, and so on. Between that and the added ambience, I was convinced this was removing detail, dynamics and even bass performance. Compared to Dolby Headphone processing, I still preferred the latter.


continued in next post...
 
...continued from above

Summary and Comments
So, I surmised the processing was not just adding ambient content to create a virtual 3D environment, but also some sort of speaker modelling was going on – like they may have been modeling real speakers in real rooms. This was proven correct when I read the marketing, FAQ, and played with the trial software which allows one to pay a hefty price to license specific speaker modeling algorithms. I am really opposed to this approach to reproducing audio. Here’s why…


All transducers fail at perfectly reproducing audio signals in the form of sound energy. Speakers and headphones add distortions in the form of deviation from linearity in frequency response, phase response, harmonic distortions, ringing & resonances, and inter-harmonic distortions, and other added noises. They also remove content due to the various response issues listed above and their failures at perfect dynamic reproduction and cancellations in output from resonances. Add to that all the acoustic issues which add and remove from the acoustic energy as it moves from the transducer(s) to the ears and you get a very complex transfer function which makes every speaker and headphone sound different as they fail to be perfect. That is what we contend with in this hobby – finding the most pleasing form of audio performance failures. So, you take a headphone which already introduces changes to the sound and virtually model a speaker and all its failing and you get deviations from linearity piled on top of deviations from linearity multiplying the aforementioned reproduction failures. Add to that attempts to add ambience which requires adding content to the signal to recreate the reflections in a room and cross-over of sound from the left to right (or vice-versa) to give the impression of a virtual room and virtual speaker placement and things will inherently deviate substantially from the original.


Dolby Headphone only processes ambient information with some EQ to ensure the sound doesn’t get boomy or brittle, and it is less than perfect but retains the general sound of the headphones you chose and likely love. Out of Your Head processing adds speaker modeling to the process and flies even further from attempts to accurately reproduce a signal and the sound of your headphones is masked and removed.


Additionally – and I haven’t written about this very much, but here it goes – adding ambient information to a signal increases the difficulty for a transducer (speaker or headphone) to accurately reproduce the signal. The simpler the signal is, the easier it is for these imperfect devices to reproduce it. A simple sinewave is easier to recreate acoustically than a ramp, sawtooth, or square wave, which are effectively impossible for a speaker or headphone to reproduce. If you consider the inherent resonances in every transducer, adding more excitement in the form of additional information through delayed echoes and phase inverting reflections and so on to increase the perception of ambience with intentions to virtually reconstruct the reflections in a real room, like a movie theater, concert hall, and so on, simply increases the complexity of the signal being fed to the speakers/headphones. This causes the negative effects from the limitations of the speakers/headphones to be amplified and thus hurts the accuracy. This is why many of the songs I chose to put on my Speaker Demo Discs are extremely dense and complex – I am testing speakers’ abilities to handle difficult signals which is a very real limitation of all. We measure this ability in the form of impulse response, step response and waterfall charts, and no transducer is perfect in all areas.


So, this exercise of forcing a speaker/headphone to recreate a vastly more difficult signal to virtually produce an ambient environment which seems to be like a room, well, I find it unnecessary and just a waste of good listening time. When wearing headphones I’d rather discover the subtle sound of a musician exhaling during a break in playing his nylon string guitar from a more accurate experience which may lack a sense of the speakers being in front of me than add a bunch of processing to give the impression the sound is coming from speakers and never hear the musician exhale at all.


VR Audio
For true virtual reality audio, which I see value in for VR games and movies, the placement of the virtual sounds need to move in the headphones to compensate for the listener moving his head around. This is already being done primitively with VR headsets, but it can be improved and requires considerable real-time processing and extremely accurate positioning and movement sensors on the headset. It also requires regular calibration of head location, which isn’t difficult. Attempts to make a truly high-end audio version of VR audio has failed a few times and it raises the common question – do we need more goofy stuff to get lost in music. Stereo was a great improvement in the listening experience, but I find surround music is usually very distracting and takes away from the experience of getting lost in the music (with very few exceptions). So, adding the concept of VR to high end critical listening seems unnecessary.


Conclusion

I am happy with basic stereo over headphones and the Out of Your Head processing doesn’t change my opinion. However, this is just a hobby and we all have things we like. If anyone enjoys this stuff, go for it! Have fun with it!
 
Last edited:
>>>The midrange, as mentioned, has a strange “honky-ness” to it which I can only describe as something like listening to a speaker at the end of long narrow corridor.<<<

That's something I've found with many DSP processes which try to use head transfer function processing. Its distracting to my ears, but obviously it doesn't bother most people.
 
>>>The midrange, as mentioned, has a strange “honky-ness” to it which I can only describe as something like listening to a speaker at the end of long narrow corridor.<<<

That's something I've found with many DSP processes which try to use head transfer function processing. Its distracting to my ears, but obviously it doesn't bother most people.

I've often thought the years of studying sound through headphones in the studio where I had a live reference to compare to the microphone/preamp/mixer/processing/recorder process has trained me to hear through the issues most people have with headphones. That combined with the ear training I did with headphones and I think I prefer non-processed audio when it comes to reproduction.
 
I'll try it out, I just need to find a good way to get sound from my laptop to my dac in my headphone rig. (I normally use squeezebox... which of course is now known to be the worst delivery system :laughing: )
 
OK I got it working. Playing the various demo tracks... I did not get a strong impression of front and back surround. Maybe a little extra ambient echo like you would expect in a 3d space, but... not nearly the same as actually having front and rear speakers in an HT. I think I get what Flint was saying about frequency response "tunneling" and sorta felt that the stuff that was supposed to be center (e.g. vocals from Hotel California) were ... constrained. Maybe felt a little like the sound was coming from directly above me, rather than forward or back? The whole thing felt kinda "wobbly" to me, and it was a little unsettling. Sorry I know that's a horrible description. And I'm not trying to be negative here, it's just that I didn't hear front/back separation, at least not any more than I can get from a really good 2ch stereo recording.
 
Right after this, I put on Norah Jones' "Painter's Song" and was (again) blown away by the presence and layering there, in a regular 2ch recording (via squeezebox).

But I admit maybe this is just because I'm used to hearing this song, as it's one of my long time demo tracks for speaker/headphone evaluation. Maybe I'm making more of it than is really there... don't know. That's part of what makes audio evaluation so hard, we have all sorts of personal preferences and expectations that try as we might, are really hard to get away from.
 
I played all of the demos directly off the web page, through my Windows 7 hp desktop computer's headphone jack, to my Denon AH-D5000 headphones.

Ignoring all that Flint posted about any other effects on the sound, I must say that the surround effect / localization is incredibly impressive, accurate, and "real." Both for music and movies. Even more impressive was plain old stereo.

Even taking Flint's excellent criticisms into account, and ignoring any cost to acquire the software, if I were someone who sat at a computer with headphones on watching movies on the screen (which I never do), or listening to music while I worked (which I extremely rarely do), then I would not hesitate to use this software. The surround / placement effect is so "good" that it vastly overrides any of its limitations.

But since I'd never use it, and since it is priced far beyond what I'd even consider paying, I won't ever buy it.

Jeff

ps. to Razz: sorry for the tardy post. I only saw this thread this evening. Thanks for drawing it to my attention. It was a very neat experience listening to those demos.
 
Thanks for responding guys. That's a hell of a review there Flint!

I take this software as "A hell of a lot of fun". I felt the sound was very well separated, I too did not get sound from directly behind me but did visualized in my head that sound was somewhat behind me to right or left. I also felt like it was over my head a lot, above me. The opening Pink Floyd "Money" track was VERY impressive to me. I felt cash registers were well localized around my room in very specific locations. I then put on my regular 2 channel version (not using the software) and felt I was missing out on the separation. The details might be more defined in my 2 channel original, but the software breaking down a 5.1 version was much more fun and interesting. I would use this software if I listened to music from headphones.
I use my headphones late at night when watching movies sometimes but have no practical way of implementing it into my current HT setup :suspicious:
 
I recommend you listen to some surround content using the Dolby Headphone surround processing. I think it is more impressive.

But still, the issue with head movement is a vastly limiting problem that simple software with normal headphones cannot address, at all.
 
the issue with head movement is a vastly limiting problem
Flint, you are VERY focused on this issue. And that's ok, I understand that. But... I don't care about this so much. I have - at least I think I have - evolved a mental way of listening through headphones where I don't care so much about that, but accept and realize that the headphone "sound field" is always fixed regardless of my head position. One could even be so bold as to suggest this might be an ADVANTAGE of headphone listening... albeit not a "natural" one in the strict physiological sense.

And I think this isn't the only issue inherent to headphones that one's brain has to adjust to. The presence of ZERO cross-talk between channels is also inherently different from speakers. I think this is partly why I got such a huge headache from listening to headphones early on. But now my brain has adjusted to that, at least to some degree, such that what I listen for in headphones isn't really exactly the same as what I listen for in speakers. Not unrelated, certainly, but not exactly the same thing either.

Sorry, I ramble. Carry on.
 
Pauly - I am only focused on that issue in regards to simulating speaker placement thru headphones. I, too, have no issue with stereo thru headphones. In fact, I sometimes very much prefer it. But with attempts to simulate location of speakers I find the effect is actually very distracting and detrimental to the listening experience - it becomes less about getting lost in the sound and more about trying to recognize the effect of virtual speaker placement. So, that's why I get hyped up on that issue. It is essentially impossible to create a completely convincing virtual speaker soundfield over headphones, but that is exactly what tech like this claim to precisely accomplish. That's the issue.

If they can accurately track the head location and alter the virtual placement accordingly, the distraction and brain power required to enjoy the sound is eased. I've heard some amazing demos which come close, but they are yet to be practical for actual production.

Binaural recordings of live performances work better, but even those effectively breakdown with head movement, thus killing the benefits. Listening with your head in restraints, as the early demonstrations employed, work near perfectly.

But, this is a fun hobby, so my rants are just intended to be informative. If one likes something, then use it.
 
But with attempts to simulate location of speakers I find the effect is actually very distracting and detrimental to the listening experience - it becomes less about getting lost in the sound and more about trying to recognize the effect of virtual speaker placement.

I find the exact opposite as you in this case Flint. If I don't concentrate on exact placement I find the effect very pleasing, even enhancing. I'm not a critical listener so I found the effect more stimulating and I paid more attention to sound than I usually do.
 
Ha! Good point .

I find anything that adds to or subtracts from the original audio distracting.
 
Back
Top