Want to build a "binaural listening" rig!
Sep 11, 2020 at 7:42 AM Thread Starter Post #1 of 12

W1ll1eTheP1mp

500+ Head-Fier
Joined
May 3, 2020
Posts
644
Likes
257
Location
Norway
Hello fellow head-fi'ers! It would be awesome if someone could enlighten me with their knowledge/experience of this topic.

I want to build myself I headphone rig where I can listen to my music "binaurally. Something similar to Smyth Realiser. I have read some topics and instructions on different places, but they get quite technical fast. I want to be able to do this budget style. Listening to some binaural recordings is awesome and it's a shame that the producers/record companies isnt making more albums available in binaural. There I want to build a rig where I can listen to binaural to f.ex Tidal.

What software/hardware is available out there that are similiar to Smyth's Realiser? It doesn't have to be as complicated as the Realiser

I own a "old" Creative Aurvana X-FI. And it was that who started my little obsession for headphone audio. When it comes to the Aurvana X-FI, the Crystalizer effect is quite okey on bad recordings of low bitrate MP3's. The other effect on that headset sound just thin,metallic and reverby as most of those surround effects do to my ears.

I own a LG V40, laptop and tablet. I haven't bought any separate DAC's or AMP's yet since i don't know if it will sound better than my V40 due to the headphones and IEM's I own( modded M1060,PM3,IE800 and waiting for Audiosense AQ7). I really want a desktop rig, but it's a waste of money if it doesn't improve sound quality and it isn't easy to try due to where I live. There I thought that I could buy I desktop rig if needed for the binaural experience:)

I've been looking at:

Software's:
OOYH
Impulsifier
HeSuVi

These software requires alot of stuff to do measurements etc.

Hardware:
JVC XP-EXT1
DSPeaker Headspeaker
Focus by S-Audio Systems
Creative SXFI.

The JVC seems pretty good and can convert 2ch and 5.2ch to 7.1.4 ch.. it's a hefty price and the downside is that I am limited to use a specific headphone.

DSPeaker isnt produced anymore and Ive read that it can only play lossy music???

Creative SXFI has mix reviews. Some are really blown away by it and some days it is the same thin,metallic reverby sound effect.

So what do u guys think? What is your experience?

This was a very long post and I had to write as can while having fever, so hope it's not a very exhausting read!

Thanks:)
 
Sep 11, 2020 at 2:37 PM Post #2 of 12
What you're talking about isn't really binaural. Well it is in the sense that you want something that ends up as 2 channels in your ears, but what the term is used for in this hobby is specific to stuff recorded that way. Not what you're discussing or try to achieve here.
What you're talking about is speaker simulation or some more general 3D simulation on headphone. For that, one way or another you need some impulse responses to characterize the sound source(virtual, or copied from some speakers), so you can have that sound applied to your headphones through convolution. The basic principle will always work that way.

Any solution that doesn't require you to measure stuff at your ear, or at the very least to take pictures or video of your head/ears, is going to apply some generic stuff instead of something closer to your own HRTF. The usual consequence is something that doesn't work as well subjectively. Some people get lucky, but on average, even though it's a PITA, you should aim for as much customization as possible and that does start with getting your very own HRTF or at least your very own IR for the various channels you plan to reproduce(can be just 2 impulses if you're fine with stereo albums and no head tracking).
I fully understand getting cold feet when checking Impulsifier and all the personal work, installation, configuration, having to get or make your own binaural microphones to capture the impulses, etc. Even though it's such a great idea, you can see from the thread that the people who really went all the way with it can be counted on your fingers. Sadly that's the cost of trying to get custom sound. There are people who offer to create a model of your HRTF(no idea how accurate it is), from video(or was it pictures?) of your head. But it's expensive, and once you have that, you'll still have to work to extract the impulses you wish to use and have that implemented in the playback solution that won't sound stupidly sterile(as on their own, they won't have room reverb or anything like that).
Even if you invest in the A16, you still have a lot of reading and personal work to put in before you'll really achieve a fully customized(and most realistic) simulation. I don't want to claim to know what I don't, but based on the general unwillingness of audiophile to do personal experiments if they require even a tiny amount of personal work, I wouldn't be surprised at all if more than half the actual owners of the A16, never got beyond using the default room that was delivered in it. It's a room that obviously wasn't captured at their own ears, so the simulation isn't going to be right for most people. But the head tracking works well, the channel separation is at the very least better than most crossfeed effects, so I do expect that many people got satisfied with that much and ultimately payed a lot of money for not using half of what the A16 is about. Nobody other than me would bother with the A16 in my family. My mother wouldn't even endure reading the manual. And she's not dumb, it's just too technical for her. If you really want to, the manual has all the steps and the dedicated thread here has people who can help with the more abstract ones. But it's work, that again, not everybody will be willing to put in. Then some people don't have a speaker rig, or only a subpar one in a crappy room(me! Me! Me!), which makes getting the reference impulses even harder because now we need to go find a place willing to let us waste their time and play annoying test signals at fairly loud levels, many times.
It's not for everybody, and at the same time, it's basically the one device that is for everybody because it will make the sound you, specifically, you with your specific head, need.

Compared to that Impulsifier has harder first steps(getting everything installed and working, getting you're own binaural mics and whatever ADC solution to record the signal from them). But after that, you'll only need 1 impulse response for one playback channel(so 2 for stereo, and 6 for 5.1(or is it 5 and just a low pass of the signal for the sub? IDK). As there is no head tracking involved, you don't need to multiply the takes for each speaker with your head in various positions like on the A16. Overall I would say that it's a better solution for a beginner. Plus it's a lot cheaper than an A16. You put the amount you want into the mics and that's about it. Plus it gives you a real sense of what can be done, and lets you practice and understand the basic principles that will help you if you ever go deeper down that rabbit hole.

Something like OOYH or any solution that doesn't customize anything or only uses some simplified range of options, has obviously the advantage of being pretty much plug&play. that's a powerful argument. But as I mentioned, no customization drastically increases the chances that you won't get a good experience. Some lucky guys are absolutely thrilled with OOYH, in fact some people were happy with the old Surround DSPs for headphone, or even with basic crossfeed effects. Different people, different results. The more customization, the more personal work, the greater the odds of getting a convincing imaging on playback. At some point it might even deserve to be called accurate. Not everybody cares about that, but some clearly do.
 
Sep 11, 2020 at 6:51 PM Post #3 of 12
It’s an older and more rudimentary solution, but some people like the “3D” effect amps with crossfeed have. Check out the ifi iDSD micro (or anything of ifi’s that has the “3D+ Matrix” which is their fancy term for crossfeed.) It basically takes a little bit from the left channel, time delays it, and feeds it through the right channel, and vice versa, mimicking the way our ears naturally hear sounds produced in the real world.
 
Sep 15, 2020 at 6:24 PM Post #5 of 12
A bit more than just a simple crossfeed, but I'll take it :)
Hey! Is that 3d effect the same on all of your products that has it?
 
Sep 16, 2020 at 8:49 AM Post #6 of 12
What you're talking about isn't really binaural. Well it is in the sense that you want something that ends up as 2 channels in your ears, but what the term is used for in this hobby is specific to stuff recorded that way. Not what you're discussing or try to achieve here.
What you're talking about is speaker simulation or some more general 3D simulation on headphone. For that, one way or another you need some impulse responses to characterize the sound source(virtual, or copied from some speakers), so you can have that sound applied to your headphones through convolution. The basic principle will always work that way.

Any solution that doesn't require you to measure stuff at your ear, or at the very least to take pictures or video of your head/ears, is going to apply some generic stuff instead of something closer to your own HRTF. The usual consequence is something that doesn't work as well subjectively. Some people get lucky, but on average, even though it's a PITA, you should aim for as much customization as possible and that does start with getting your very own HRTF or at least your very own IR for the various channels you plan to reproduce(can be just 2 impulses if you're fine with stereo albums and no head tracking).
I fully understand getting cold feet when checking Impulsifier and all the personal work, installation, configuration, having to get or make your own binaural microphones to capture the impulses, etc. Even though it's such a great idea, you can see from the thread that the people who really went all the way with it can be counted on your fingers. Sadly that's the cost of trying to get custom sound. There are people who offer to create a model of your HRTF(no idea how accurate it is), from video(or was it pictures?) of your head. But it's expensive, and once you have that, you'll still have to work to extract the impulses you wish to use and have that implemented in the playback solution that won't sound stupidly sterile(as on their own, they won't have room reverb or anything like that).
Even if you invest in the A16, you still have a lot of reading and personal work to put in before you'll really achieve a fully customized(and most realistic) simulation. I don't want to claim to know what I don't, but based on the general unwillingness of audiophile to do personal experiments if they require even a tiny amount of personal work, I wouldn't be surprised at all if more than half the actual owners of the A16, never got beyond using the default room that was delivered in it. It's a room that obviously wasn't captured at their own ears, so the simulation isn't going to be right for most people. But the head tracking works well, the channel separation is at the very least better than most crossfeed effects, so I do expect that many people got satisfied with that much and ultimately payed a lot of money for not using half of what the A16 is about. Nobody other than me would bother with the A16 in my family. My mother wouldn't even endure reading the manual. And she's not dumb, it's just too technical for her. If you really want to, the manual has all the steps and the dedicated thread here has people who can help with the more abstract ones. But it's work, that again, not everybody will be willing to put in. Then some people don't have a speaker rig, or only a subpar one in a crappy room(me! Me! Me!), which makes getting the reference impulses even harder because now we need to go find a place willing to let us waste their time and play annoying test signals at fairly loud levels, many times.
It's not for everybody, and at the same time, it's basically the one device that is for everybody because it will make the sound you, specifically, you with your specific head, need.

Compared to that Impulsifier has harder first steps(getting everything installed and working, getting you're own binaural mics and whatever ADC solution to record the signal from them). But after that, you'll only need 1 impulse response for one playback channel(so 2 for stereo, and 6 for 5.1(or is it 5 and just a low pass of the signal for the sub? IDK). As there is no head tracking involved, you don't need to multiply the takes for each speaker with your head in various positions like on the A16. Overall I would say that it's a better solution for a beginner. Plus it's a lot cheaper than an A16. You put the amount you want into the mics and that's about it. Plus it gives you a real sense of what can be done, and lets you practice and understand the basic principles that will help you if you ever go deeper down that rabbit hole.

Something like OOYH or any solution that doesn't customize anything or only uses some simplified range of options, has obviously the advantage of being pretty much plug&play. that's a powerful argument. But as I mentioned, no customization drastically increases the chances that you won't get a good experience. Some lucky guys are absolutely thrilled with OOYH, in fact some people were happy with the old Surround DSPs for headphone, or even with basic crossfeed effects. Different people, different results. The more customization, the more personal work, the greater the odds of getting a convincing imaging on playback. At some point it might even deserve to be called accurate. Not everybody cares about that, but some clearly do.
Thabks so much for your reply! This was a good read:) Im not afraid to put alot of work into it nor the technicalities i dont understand yet(alot of these terms and processes are quite new to me).. Its my least concern. My biggest concern is how much money i atleast have to spend to be able to get good results since i own a minimum! Downloaded OOYH yesterday, and just " out of the box" it didnt impress me at all. I prob need to eq my 'phones flat or to Harman target via Eq APO?
 
Sep 16, 2020 at 5:39 PM Post #8 of 12
Thabks so much for your reply! This was a good read:) Im not afraid to put alot of work into it nor the technicalities i dont understand yet(alot of these terms and processes are quite new to me).. Its my least concern. My biggest concern is how much money i atleast have to spend to be able to get good results since i own a minimum! Downloaded OOYH yesterday, and just " out of the box" it didnt impress me at all. I prob need to eq my 'phones flat or to Harman target via Eq APO?
It's difficult to say much of anything. The tech is fairly old, and open to anybody who can code and make stuff. So why isn't it already everywhere? I'm not sure. The need for customer work is probably a big hurdle but maybe it's not the only one, I really don't know. Maybe in 10 years we'll still be there with some version of the A16 and some generic Atmos and DTS headphone apps. Maybe next year half the planet will have advanced and slightly customized room simulation from apple earbuds and the latest Playstation? So I don't feel like I can tell you to invest in something like the A16 when I have so little idea of what the future has in store for us. My own expectations have been wrong or overly optimistic for the last 20 years. So at least we know that my predictions are worthless. "it's something" ^_^.

Now prepare for the short story long. If you're motivated, you'll have to understand this at some point anyway, so why not have it explained now by the worst educator in the entire universe(Venus included)?
You've been warned, you could just go google HRTF and sound localization to try and get the basic concepts from proper sentences and illustrations, and just forget about the rest of this post.
OOYH doesn't have your HRTF integrated to the sound. Some EQ can help compensate for the headphone's own FR fluctuations, and that's likely to be of help. But here is the problem and why I keep talking about HRTF instead of a simple EQ for something that already handles the room reverb: your head creates a specific frequency response for each angle around you(you locate sounds sources with that and with the timing delay between both ears catching the same sound). So if we imagine the most simple most typical situation of a pair of speakers 30° on each side, and you looking straight in the middle, so long as you don't move your head at all, the sounds will come from those 2 30° directions, and each speaker will reach both ears. To capture that specific and unmoving position, you will need 4 impulses. Left speaker to left ear, left speaker to right ear, right sp to left ear, right sp to right ear. And you're done. With those 4 impulses you have all the data you need to replay any music as simulated by those 2 specific speakers in that room while looking straight ahead. OOYH did just that for the rooms it captured, but the impulses obviously didn't involve your own head. That's the missing part.
Now from what I said you perhaps suspect that one EQ for the headphone might not be enough as FR changes with angle and you need 4 different captures for 2 speakers. For example, we do not know that your left and right ear have the same FR impact on incoming sounds. Most people have some slight asymmetries on the head and body. The result of sounds bouncing a little bit everywhere on us might turn out to be significantly different for each ear(at least at one or 2 specific frequencies areas and for a few specific angles). With some luck most of the differences will be small and EQing your headphone to perceived flat will bring almost all the improvement possible. But it's not a given that it will be enough to remove what troubles you with a stereo playback on OOYH for example. Ideally you'd want one EQ for each of the 4 impulses taken for your stereo playback(more if you use multichannel). Not only it's work, but it's really hard to guess what the right EQ needs to be for ourselves. One possible method would be to use the speaker at that angle as reference and trying to go back and forth between 1 speaker and one headphone driver while playing test tones or some noise. And hope to come fairly close by ear. But again, you'll most likely only have one EQ for everything, EQ APO can do one EQ per channel I believe but I have never used it. Still that's only 2 EQ when ideally you'd need 4 and with OOYH, there is no possibility of doing it(maybe if you can find the 4 EQ, then you could discuss with the owner how much it would cost to have them integrated into the room?).

This video give you the most straightforward example possible of trying to match headphone and speaker by ear.
note that because it's binaural recording on himself and he's usually at some fair distance of the orchestra he's recording, aligning his perception with a center/mono reference makes a lot of sense.
But with stereo speakers simulation, even the perceived center is born from equal output on both speakers. The sound sources are effectively always the speakers at 30°. We have to pursue the original sound source(s) if we want an accurate simulation. So his solution to target mono isn't ideal for room simulation. It might still be a good deal better than the default headphone's FR for many listeners, but it's not the right objective approach. Just a simple case so the example and why he's doing it that way is easy to grasp.
About Harman target, again, not the right objective approach, but who knows how you'll feel about applying it? Maybe you'll like that a lot. I say not the right approach because that target is a listener's preference target and one specifically researched while playing stereo albums on headphones the old/wrong way. So it's not the right reference for room simulation where we try to get closer to the "right" way to play back an albums usually mastered on and for speakers.

IMO go harass @jaakkopasanen in this thread https://www.head-fi.org/threads/recording-impulse-responses-for-speaker-virtualization.890719/ until you get his app to work, seems like the very best solution available today. He must start to hate me for always pointing out to him, but at the same time he's such a nice guy, he'll probably do his best to help anyway. Your only investment beside time and efforts, is to get a pair of binaural microphones or make some yourself.(so the price can go from maybe 40$ or material and a soldering iron, to thousands of $ for professional binaural microphones(I suspect that cheap ones can already do a splendid job).
You'll be able to test recording your own speakers and see how far you think you can get and if subjectively that seems enough for you or if you're left wanting more.
Or of course you could get an A16 at full price. Still a lot of reading, testing, recording, but then you have recorded your room or any room you can access with a cool sound system, and you'll have head tracking, smooth HRTF application(your own!), and maybe you'll even go get yourself a "bass shaker" or whatever it's called that you put on your chair or sometimes on yourself depending on the brand, to get some tactile vibration trying to give you the feeling of tactile bass to increase the illusion of listening to speakers.
No matter what you end up deciding, I repeat my warning: I have no idea what will happen in the next years. Maybe we'll have something 1/10th the price of the A16 doing the same with a better UI? Maybe we'll only have more of the same half done simulations. I just don't know.
 
Sep 17, 2020 at 3:33 AM Post #9 of 12
Its core is the same, but from one generation to another it changes and improves.
Thanks! So which of your products has the most improved 3d circuit? So is it excactly the same between Nano and Micro models? Just the generation that differs?You are pretty much biased,but is it good? Haha:) does it takes the sound " out of my head"?
 
Sep 17, 2020 at 4:46 PM Post #10 of 12
So which of your products has the most improved 3d circuit?

The latest stuff; hip-dac and our ZEN range.

So is it excactly the same between Nano and Micro models?

If we're talking nano and micro iDSD Black Label units, these came out in similar time, which would make their 3D tech similar if not the same, but I'd have to double-check this.

Just the generation that differs?

Yes, the generation of a product is more important than its price as far as our 3D/Xbass implementations go.

You are pretty much biased,but is it good?

Of course we stand by our own tech and everything we do is for a reason, but we're far from saying that our stuff is the best thing after sliced bread. Once in a while our customers say it for us :)

does it takes the sound " out of my head"?

Many people would say that our 3D tech does that, but you'd have to find this out for yourself :)
 
iFi audio Stay updated on iFi audio at their sponsor profile on Head-Fi.
 
https://www.facebook.com/iFiAudio/ https://twitter.com/ifiaudio https://www.instagram.com/ifiaudio/ https://ifi-audio.com/ https://www.youtube.com/@iFiaudiochannel comms@ifi-audio.com
Sep 18, 2020 at 2:54 AM Post #11 of 12
Thanks again for your reply. But i cant find any info about Hip-dac having the 3D circuit? Ony Xbass?
 

Users who are viewing this thread

Back
Top