Headphone effect

[abbey road d enfer]

This subject is a sequel of a discussion between member George Toledo and me.
See the modern desk/console in 2021... starting at post #68.
Since it threatened to hijack the original subject, I suggested we opened a new thread.
The name came from Brian Roth.
It's a subject that crosses-over the Drawing board and Studio A, because it involves both theory and practice.

 

[Brian Roth]

And here is my remark that messed up the original thread:

I spent decades recording with analog desks and 2" multitracks with essentially 0.000 mS latency so I'm used to hearing my own voice via headcans with polarity flips. Every digital system I've encountered....no matter how low the latency claims are....sound funny to me with headcans regardless of polarity flips. Maybe I'm very sensitive to latency, or maybe I'm just crazy.

 

[PermO]

Nope,

I agree, when performing even the slightest latency feels weird and affect the performance.

For me zero latency is the only way to record.

I'm also the person that sees fluorecent tubes flickering, modern car dashboards flickering, displays flickering, even very expensive broadcast viewfinders are flickering to my eyes.

So my appartment is full of "old stuff" to prevent me from living in a constant migraine.

Also a lot of digital audio stuff to me sounds like there is a grinding mill grinding rocks in the background.

Can't help it, I'm a very sensitive person.

(short anakdote, a client and me where visiting a shipyard, I was going to do some antennawork there, and this (huge indoor) place was nasty, I immediatly felt "If you'll stay here, you'll get sick" so I told him I would only work outside.
He thought I was crazy and kept joking about it, untill 3 weeks later they shut the whole thing down because a report came out about the indoor air quality. So yeah, I trust my senses, I don't need an official report telling me what's going on, my senses tell me right awayand they're always right)

People are all different, so different that it's even impossible to find "a matched pair" on this planet

 

[Bo Deadly]

And here is my remark that messed up the original thread:

I spent decades recording with analog desks and 2" multitracks with essentially 0.000 mS latency so I'm used to hearing my own voice via headcans with polarity flips. Every digital system I've encountered....no matter how low the latency claims are....sound funny to me with headcans regardless of polarity flips. Maybe I'm very sensitive to latency, or maybe I'm just crazy.

Digital latency is almost certainly more than whatever is advertised but lets say its somehow only 1 ms. That's 180 degrees at 500 Hz which of course gives you a comb filter [1]:



As the delay gets longer, the comb shifts down. In practice, digital latency is probably more like 2-3 ms which corresponds to 250-167 Hz and puts the first / biggest notch in a bad place for audio and speech.

I've spent a lot of time thinking about the effects of short delays. There are no devices that can create really high quality short delays. Digital is limited by the latency create by the ADC and DAC. So short BBDs actually have an advantage over digital in this particular corner of audio processing. One day when I have an infinite amount of time and resources I want to explore this hole a little better.

 

[PermO]

I've spent a lot of time thinking about the effects of short delays. There are no devices that can create really high quality short delays.

There are,

These are called "reflective surfaces"

 

[Tubetec]

I had a chance sometime back to look over the setting up of a protools session in a well equiped studio .
Mics were taken first to API modules , the output was then split , one to protools hardware and another to a smaller analog mixer which ran the headphone mixes . The engineer who was maybe a little older than me and started out with analog tape explained he just didnt trust the H/P mixes through digital , made sense to me at the time , now were digging a little deeper technically into side effects of even very small delays the argument for an all analog path from the mics to the performer(s) ears is even stronger .

 

[GeorgeToledo]

The latency of low frequency from one player to another in a band rehearsal can be greater than the throughput latency of a modern daw system.

Lowest buffer settings, 44.1 sample rate, is typically 1.1ms or less to output with an oversampling converter. Raise sample rate, even less…getting into half a millisecond or less territory.

 

[GeorgeToledo]

It is honestly against better judgement to partake in this, in that I think the handling of this in the other thread was at a level which I would find completely unacceptable in first hand discussion. If someone made a snide remark about “name dropping” after attempting to explain something to them, it would certainly be the end of that conversation. Especially if they are simply an anonymous person to me with no real accountability for being rude, and potential motives for that being unknown to me. It is a true waste of time and energy to partake in a discussion on that kind of ground.

But to briefly sum up for posterity, and since there are others on the forum who may be interested to consider this…

When a singer is at a directional mic, and moves closer there is proximity effect. This effect is negligible with typical imperfect omni, and with a theoretical perfect omni, it doesn’t exist. The imbalance of pressure leads to this enhanced low end effect, and can only be present in a directional mic by definition. Singers typically sing at distance where proximity effect occurs.

If a singer complains about bass being shy, or a “funny sound”, it is typically helpful to invert polarity and see if that proximity induced low end can then be heard as they monitor. When the mic is in proper polarity the proximity induced lows couple with their head resonance in an additive way. When the mic is inverted polarity, the proximity induced lows become *subtractive*.

What happens to the proximity induced lows with omni? Easy, there is no proximity effect to begin with. There is no proportional increase of low end compared to the rest of the spectrum as a singer approaches the mic, because the mic is not directional. There is no pressure imbalance from front to back to create this proximity effect.

I quite simply *never* stated that there is lack of perceptual effect overall if a singer is singing into an omni and polarity is flipped. That seems to be a potential source of some nitpicking here, but it was projected onto my statement.

 

[PermO]

https://physicsworld.com/a/sound-breaks-the-light-barrier/
Don't know if relevant, always found it quite fascinating...

 

[abbey road d enfer]

When a singer is at a directional mic, and moves closer there is proximity effect. This effect is negligible with typical imperfect omni, and with a theoretical perfect omni, it doesn’t exist. The imbalance of pressure leads to this enhanced low end effect, and can only be present in a directional mic by definition. Singers typically sing at distance where proximity effect occurs.

True.

If a singer complains about bass being shy, or a “funny sound”, it is typically helpful to invert polarity and see if that proximity induced low end can then be heard as they monitor.

half-true, half-false. The effect is irrelevant of how LF is reproducesd, either via a directional or an omnidirectional mic. The only thing that counts is the relative level and phase at the listener's ear.

When the mic is in proper polarity the proximity induced lows couple with their head resonance in an additive way. When the mic is inverted polarity, the proximity induced lows become *subtractive*.

True.

What happens to the proximity induced lows with omni? Easy, there is no proximity effect to begin with. There is no proportional increase of low end compared to the rest of the spectrum as a singer approaches the mic, because the mic is not directional. There is no pressure imbalance from front to back to create this proximity effect.

True but irrelevant.

I quite simply *never* stated that there is lack of perceptual effect overall if a singer is singing into an omni and polarity is flipped.

Well, you stated it clearly and that's what started this fiery exchange. Your post:
https://groupdiy.com/threads/the-modern-desk-console-in-2021.77995/page-4#post-996443
"(An omni won’t exhibit that low end cancellation though.)"

 

[buckethead]

I'm thinking about this a lot as I'm neither having a low latency monitoring interface nor an analog HP mixer, so everything goes through AD-DAW-DA, around 3-5ms. It's been quite an experiment and I'm still waiting for someone to notice/beeing annoyed by it (sometimes singers hear it). What it came down to for me is that soundwaves take around 3ms per meter, so playing guitar with an amp 3-4 meters away already sums up to 10+ms in delay etc. Then take drum overheads that are 1-2m or more in distance to the snare which again is maybe 1m in distance to the players ears and a couple of meters in distance to the other performers - so I guess everything's a mess delay-wise anyways. Singers are the exception, the voice travels really fast from throat to ears... voices are a weird thing, so is bone conduction

Anyways, a compact HP mixer would be great to have

 

[Bo Deadly]

If you put a guitar amp in a sound proof room with a mic 1 meter away and then monitored that remotely using headphones, I would not be surprised if it was awkward to play. The sound of the guitar relies greatly on the room and echos bouncing around and vibrating the strings which are amplified which is to say the strings are individual resonators. The individual fretted string from one moment to the next is a resonator [1]. The delay will be so complex it will be more like reverb than a delay.

A singer in a booth with a mic a few inches from their face is a completely different setup. When the signers brain is trying to carefully control their diaphragm and vocal chords an echo can really throw off their brain. Try speaking while monitoring wet-only echo of your voice. It's so awkward its hard to just recite a sentence cleanly. So how short does the echo (aka delay) have to be before it's no longer awkward for the signer?

[1] Incidentally guitar string resonance might be why a lot of digital models of guitar amps don't sound that great. The feedback loop between string and amp cannot be digitally modeled because a digital model cannot provoke the string to vibrate. Maybe if you play in front of a speaker it would sound better though.

 

[GeorgeToledo]

Yes abbey, to start from your last sentence…the omni won’t exhibit the low end cancellation from proximity effect a singer is experiencing, which only comes from directional mic to begin with. I stated it clearly but for some reason you still fail to get it. Context is everything, you failed to understand it, and instead argue in circles.

You may not realize it but this is harassment at this point. I asked you to move it to PM, you neglected to. In the other thread I *told you* to stop discussing this with me. That is saying NO. Yet you keep on. I ask the identity of who I am talking to, you refuse to communicate on terms that give culpability for your loaded statements.

Your posts to me came in early with your personal insult of "name dropping" after explaining that Stephen Peus would make this same point often, along with various Neumann representatives. Again, in the real world would not result in any further discussion at all. Get ahold of yourself, you are supposed to be a moderator here. It is also quite an ironic attempt at insult given the HUGE abbey road namedrop attempt in your *screen name*.

Who is in charge of abbey here? I will also add that there was no visible moderator tag at the start of recent interaction.

 

[abbey road d enfer]

Yes abbey, to start from your last sentence…the omni won’t exhibit the low end cancellation from proximity effect a singer is experiencing, which only comes from directional mic to begin with.

Can't you understand that the cancellation will happen whether the sound is picked by an omni otr a directional mic?

I stated it clearly but for some reason you still fail to get it. Context is everything, you failed to understand it, and instead argue in circles.

You keep repeating that, and I maintain you are wrong.

You may not realize it but this is harassment at this point.

It is you who are insisting you are right, afgainst all scientific evidence.

In the other thread I *told you* to stop discussing this with me.

Well, stop discussing with me.

I ask the identity of who I am talking to,

I am Abbey Road d'Enfer, moderator.

you refuse to communicate on terms that give culpability for your loaded statements.

Who makes loaded statements?

Your posts to me came in early with the personal insult of “name dropping”,

"name dropping" is not an insult, it's just describing what you did in naming Peus, out of context and wrongly.

It is also quite an ironic attempt at insult given the HUGE abbey road namedrop attempt in your *screen name*.

The insult is in the ear of the listener.

Who is in charge of abbey here?

Why don't you report me?

I will also add that there was no visible moderator tag at the start of recent interaction.

 

[GeorgeToledo]

I was specifically referring to proximity effect getting cancelled out. That doesn't happen with omni. The point you are arguing is a strawman.

.

(the mod tag was not displaying here on my end, you may not realize that taking a screengrab now doesn't really prove anything in that regard one way or the other.)

 

[abbey road d enfer]

I was specifically referring to proximity effect getting cancelled out.

It's not the proximity effect that is cancelled out, it's the low frequencies, whether they are the result of the natural response of an omni mic or teh "effected-by-proximity" response of a cardioid.

That doesn't happen with omni.

Cancellation does appear with omni mics too. That's just plain physics. Two out-of-phase pathes combined result in cancellation.

(the mod tag was not displaying here on my end, you may not realize that taking a screengrab now doesn't really prove anything in that regard one way or the other.)

I don't want (don't need) to prove anything; I just want to show you where to look.

 

[JohnRoberts]

Yes abbey, to start from your last sentence…the omni won’t exhibit the low end cancellation from proximity effect a singer is experiencing, which only comes from directional mic to begin with. I stated it clearly but for some reason you still fail to get it. Context is everything, you failed to understand it, and instead argue in circles.

proximity effect wrt omni mics is well known and I predict even understood by Abbey.

You may not realize it but this is harassment at this point. I asked you to move it to PM, you neglected to. In the other thread I *told you* to stop discussing this with me. That is saying NO. Yet you keep on. I ask the identity of who I am talking to, you refuse to communicate on terms that give culpability for your loaded statements.

From what I can see he has been civil but persistent. You can not make a forum member go away just because you do not agree with his science. In fact he (we) could make you go away but are giving you an opportunity to calm down and perhaps see the logic of his arguments.

This is not a threat just a statement of fact. The forum does not have technical editors but the mods are pretty knowledgeable and resist leaving misinformation unchallenged.

Your posts to me came in early with your personal insult of "name dropping" after explaining that Stephen Peus would make this same point often, along with various Neumann representatives. Again, in the real world would not result in any further discussion at all. Get ahold of yourself, you are supposed to be a moderator here. It is also quite an ironic attempt at insult given the HUGE abbey road namedrop attempt in your *screen name*.

Who is in charge of abbey here? I will also add that there was no visible moderator tag at the start of recent interaction.

Abbey is a well respected moderator here for years. You have now involved all the moderators and admin by reporting him, which at first glance appears to lack any substance. Please don't waste our time by pursuing this as some kind of a rules violation because it isn't.

JR

 

[GeorgeToledo]

There is no logic in a continued argument about a scenario outside of the context I originally laid out. It is bizarre and unprofessional.

I was referring to a specific phenomena that only happens with directional mics. I set a scope which was focusing on that specific case. Cancellation occurs with omni mics? Yes it does, yet I was referring to proximity effect cancellation. I keep pointing this out, Abbey keeps arguing another point unrelated to my statement.

As a singer approaches a mic, proximity effect will be additive when polarity is correct. It is disorienting for it to *become less the closer an artist gets to the mic*. This effect is proximity driven and is precisely what I was initially speaking to. Regardless of any other cancellation effects that may happen for an artist when using an omni mic with inverted polarity, there will not be this inversion of proximity driven low end which can only happen with directional pattern.

Abbey asked me to report him above and I obliged.

 

[JohnRoberts]

I am not going to join you on that merry go round...

You reported him and I responded that he did nothing wrong.

You appear entrenched in your viewpoint, as does he.

Perhaps take a break and let the forum members make up their minds for themselves.

JR

 

[GeorgeToledo]

If someone states “stop contacting me”, continued contact by the other party is harassment. It may not be viewed as such on this site, but it remains an arguable viewpoint.