[bigshot]

Primary depth cues are physical… reflections off the walls of the listening room, frequency response changes due to distance, etc. Real cues in the real world. Secondary distance cues are burned into the recording… hall reverb from the recording venue, miking distances, etc. I suppose there could be tertiary distance cues caused by time modifications from digital signal processing. But I don’t see how headphones could create any of these.

 

[IDrinkLava]

Primary depth cues are physical… reflections off the walls of the listening room, frequency response changes due to distance, etc. Real cues in the real world. Secondary distance cues are burned into the recording… hall reverb from the recording venue, miking distances, etc. I suppose there could be tertiary distance cues caused by time modifications from digital signal processing. But I don’t see how headphones could create any of these.

I've actually been thinking about this lately, although I didn't want to create another thread to flood this forum.

In my experience, I have tried headphones that have relatively accurate left-to-right stereo panning and I have tried headphones with less accurate (one could subjectively call it "blurry") left-to-right stereo panning. I used to think this is what people meant by "imaging" but I now know it could simply be referred to as time delay.

In one interesting case, by swapping the stock foam ear disks on the Koss KSC75 with Yaxi ear pads, I had the unfortunate side effect of experiencing this less-accurate left-to-right panning. My theory is that the thicker foam material on the Yaxi ear pads causes increased time delay. Sure enough, I looked up drivers of headphones I've used that have accurate panning like the HD 600-series and drivers with less accurate panning like the AKG K700-series and the Sony MDR-ZX100-series and this seems to confirm what I was thinking.

This objective design flaw might in fact be an intentional choice to increase perceived "spaciousness" of the music among listeners, which also seems to be confirmed by the fact that according to solderdude, the foam disk covering the K700-series driver doesn't actually have any effects on the frequency response at all.

 

[bigshot]

Time and distance are related. I can't imagine that there would be an audible time delay when the transducer is an inch from your ear. The threshold of audibility for phase shift is 1 to 3 ms, depending on frequency. Sound travels about 1000 feet a second. When a room is 10 feet long, you're into the range of a millisecond. Sound bouncing off walls and traveling back to you would make it even closer to being audible. But an inch?

My guess is that it wasn't timing, but rather response. The pads weren't aligned the same to your ears the same way and one was attenuating the sound a bit more than the other one. That combined with some sort of response shift from a bad seal or interaction with the shape of your ear canals might mess up the imaging without affecting anything to do with timing. But it would just be the way you were perceiving it. There really wouldn't be any blending of channels or timing error. Your brain would just be interpreting it as sounding that way.

 

[71 dB]

Primary depth cues are physical… reflections off the walls of the listening room, frequency response changes due to distance, etc. Real cues in the real world. Secondary distance cues are burned into the recording… hall reverb from the recording venue, miking distances, etc. I suppose there could be tertiary distance cues caused by time modifications from digital signal processing. But I don’t see how headphones could create any of these.

Okay, thanks. Now I know better what you mean.

Time and distance are related. I can't imagine that there would be an audible time delay when the transducer is an inch from your ear. The threshold of audibility for phase shift is 1 to 3 ms, depending on frequency. Sound travels about 1000 feet a second. When a room is 10 feet long, you're into the range of a millisecond. Sound bouncing off walls and traveling back to you would make it even closer to being audible. But an inch?

How does the ear know when the sound was generated somewhere in the distance? It doesn't. It only knows when the sound finally hits the ear. In time-based spatial cues relative time is what matters. The time difference between ears, or the time delays of reflections. I don't know what you mean by phase shift. Our ears do not measure the phase at the speaker and compare it to the phase at your ears to calculate the corresponding distance. Spatial hearing uses various cues it can extract from the sounds at both ears, such as ILD, ITD, ISD, amount of reverberation etc.

 

[bigshot]

Time based spatial cues are primary, not secondary. I mentioned the JDD for phase shift just to give some sort of context for scale.

 

[castleofargh]

We kind of suck at determining distance accurately by ear under natural circumstances, so what hope do we have of that getting anywhere close to something accurate when we screw up half of our HRTF with headphone/IEM?
From my last 2hours trying to find papers on the subject of distance judged by ear, I got:

primary audio cues for judging distance are loudness, reverb, and FR.

Loudness dominates when it comes to judging distance.
There are conflicting results about the role that binaural cues can play in estimating distance(as if it wasn't already complicated enough). And one paper suggests that the contradictions could come from the sound samples being normalized(they test 2 methods and both ruined the ability of the listener to estimate distance better than guessing.. or something like that. I only read a little of that one). It's from 2020, so I'm not super hopeful about finding clear consensus yet.
Another paper says that while they didn't find binaural cues to help judging distance, some listeners lost their ability to estimate distance if they had no binaural cues in the audio sample. Which really doesn't help me make up my mind on the topic.
Same paper seems to suggest that reverb is correctly used by some for distance. while another paper about simplified models of virtual rooms gave this little nugget

...the perception of sound distance in an enclosed room by human listeners can be quite simply modeled by fitting a temporal window around the ratio of direct-to-reverberant sound energy.

You know, "quite simply".




To cripple my argument that we suck at judging distances by ears, here's daredevil showing how far a "non standard" life experience and training can get you.

 

[CoryGillmore]

If you close your eyes, can you imagine the band playing out in front of you, maybe in a room. I suppose for some folks it works and for some it doesn't. In terms of in-ears you can get an approximation of sound in a room given the right tuning.. sounds crazy? Not for the folks that hear it I guess.

Yeah, I don't think everyone is capable of this. I think it actually took me a while to be able to do it, but now it just seems trivial to me. But a buddy of mine, who just recently got into headphones, has a hard time with it. I'm like "dude, listen to this track. You can't tell that her voice is coming from dead center?" And he has trouble with it with headphones. I'm trying to get him over to my house so he can hear my two-channel setup. It has an amazing center image, if he can't tell that certain sounds (main vocalist, kick drum, snare drum) are coming from the center, then there's no hope.

 

[redrol]

On my best IEMs for 3D staging, the Shuoer EJ07, something like at least 10 people I personally know bought them. Every one agrees they are extraordinary, and sound close to a surround sound system! I know I can't prove it and yet every one of those people have said they are taken back fully and it's got a special something going on.

As far as my theory, the Shuoer gets around a 20 in PTRF - 10k notch which is as good as IEMs seem to be able to get.

 

[bigshot]

Do any of those people actually have a multichannel speaker system? I think people exaggerate and call things “surround” without actually knowing what multichannel surround music sounds like. I’ve never even heard two channel speaker systems that come anywhere remotely close to multichannel.

 

[71 dB]

Time based spatial cues are primary, not secondary. I mentioned the JDD for phase shift just to give some sort of context for scale.

If all you have is primary spatial cues then of course also time based spatial cues are primary cues, but secondary time based cues are easy to produce thanks to modern digital technonogy. HAAS-effect for example is one very simple (but effective) way to produce secondary time-based spatial cues not to mention much more sophisticated methods.

In spatial hearing the JDD for horizontal angle is about 1° for sounds coming ahead corresponding ITD of about 7 microseconds, but JDD varies depending on what kind of time difference/spatial cue we are talking about. HAAS-effect for example works up to about 30 ms. The point is, it doesn't matter how far the sound source is from the ear (that's just the "bulk delay", the relative tine differences are what count.

 

[redrol]

Do any of those people actually have a multichannel speaker system? I think people exaggerate and call things “surround” without actually knowing what multichannel surround music sounds like. I’ve never even heard two channel speaker systems that come anywhere remotely close to multichannel.

Well yeah of course. Nevermind theaters and such. I'm not saying its as good, of course not. I'm saying its MUCH better than your typical in-head IEM sound. Drastically so. It's good enough to completely catch me off guard a lot. Sometimes while listening to music I think I'm hearing something outside my house, take the IEMs out of my ears, nothing. Then I realize the sound is in the track I was listening to. It's like that.

 

[71 dB]

Imagine yourself in a (not very small) room with two other people speaking with same exact SPL (e.g. 70 dB). The other person is near you (say 3 feet away) and the other person is on the other side of the room (say 20 feet away).

The direct sound for the person closer to you is 20*log10 (20/3) = 16-17 dB louder because of the distance attenuation rule.
The early reflection for the person closer to you are a little bit louder (e.g. 6 dB) depending on how close reflective surfaces are.
The reverberation level is the same for both speakers.

Comparing levels spatial hearing can estimate distances from sound sources, but there is more: The further away a sound source is, the smaller the delay between direct sound and the early reflections are for simple geometrical reasons. For a speaker 3 feet away, the floor reflection is delayed by 6-7 ms compared to direct sound. For a speaker 20 feet away this delay is only about 2 ms. So, when the direct sound comes a lot ahead of the "other stuff", it is an indication for the spatial hearing that the sound source is near.

 

[bigshot]

Real life distance cues are much more present than recorded ones. It’s possible to add timing adjustments to synthesize a larger space with a multichannel speaker system, but you’re building that on the physical dimensions of the room.

Mono isn’t the same as stereo and stereo isn’t the same as multichannel. But putting any of those into a sympathetic room will make them sound more spatial.

 

[castleofargh]

On my best IEMs for 3D staging, the Shuoer EJ07, something like at least 10 people I personally know bought them. Every one agrees they are extraordinary, and sound close to a surround sound system! I know I can't prove it and yet every one of those people have said they are taken back fully and it's got a special something going on.

As far as my theory, the Shuoer gets around a 20 in PTRF - 10k notch which is as good as IEMs seem to be able to get.

Based on crina's graph it looks like something I would enjoy. Somewhat Harman-ish response?
About personal impressions. If you have some time to waste on an auditory adventure, I'd suggest to download a few .wav from there, and play them with your IEMs:
http://recherche.ircam.fr/equipes/salles/listen/sounds.html

They're FTP links so depending on the browser and what not, it might not work at all. With my little Firefox I couldn't get any file, and telling it to open my antiquated Fillezilla, it asked for a login... Maybe I'm just too much of a noob, IDK.
With Edge(wow second time I use it! \o/) it asked me which app to use, I went to get some free app from the crapstore as was proposed and that finally worked. A few years back I didn't have to go through all that mess to get the files.

AAAAaaaniway, they're pulses noises that go "brrr brrr". It's the opportunity to find out how different the sound at the entrance of the ear canal really is for different people. I don't know of anything quite like that.
If somebody struggles like I did but would like to try, tell me and I'll upload some to Dropbox or something and then I'll get sued. And if all works just fine for you, well it means I'm now officially old.

As described, they're intended to give the impression of the sound source going around the head horizontally while keeping the same distance from you. That's how it felt to each listener while using their very own HRTF convolution. How you'll actually feel about the position of the sound source with your IEM and a given .wav is the surprise.

 

[bigshot]

A well balanced response will solve most problems you might run into in home audio. When you achieve that, the sound is clearly better and the mind can go to work attributing qualities to it that it doesn’t really possess. Then you can think up even more exaggerated ways to describe it. Goodness is like a snowball rolling downhill.