Headphone effect

[Rossi]

To come back to the original question. I just did a test, and the headphone effect occurs with a U 87 Ai in all three directional characteristics, including omni. So it is not dependent on the proximity effect. That's what I expected, but I still wanted to do a test as - sometimes - the unexpected happens, and if it does, there's something to be learned. But in this case, it's just what most people around here expected. Case closed.

For the record, I used analog monitoring (via a Universal Audio DCS) and a semi-open AKG K 240 DF, which is what I tend to use for vocals.

I do think that latency is a big issue. Not in terms of a time delay (which is barely perceptable at 1-3 ms) but in terms of comb filtering. A sure sign of latency problems is when you can't decide which position of the polarity switch puts your voice in phase with the acoustic path.

Although I have to say, digital systems are getting better. Some newer audio interfaces' internal (digital) monitoring is so fast, that it almost sounds like analog. I'm guessing 1 ms digital latency is acceptable. But as soon as you monitor through your DAW, I don't think that's attainable.

 

[morls]

I was told by the engineer on this particular session that a well-known rock singer here in Australia recorded his vocals in the control room, with a PA playing the band mix at live levels while he sang into an SM58. No cans needed because I think it could be heard a block away!

 

[Tubetec]

I was involved in a recording session years ago where the vocalist became really unhappy with the headphones , its so long ago now the details are sketchy in my mind . Basically a small Bose single drive unit speaker was set up in the vocal booth to reproduce the track , the vocalist sang without headphones and I think phase shift was used to (partially) cancel the backing music from the vocal mic . Not an ideal situation by any means ,definately something an engineer would rather not do , but you know if the artist is unhappy compromise is sometimes the only option.

I also heard about sessions in a place in Dublin called STS studios , U2 were their biggest customers at the time , the place was the size of a shoe box but they just liked the vibe . Apparently they would have their live sound engineer do monitor mix and the track would be reproduced over a massively powerfull wedges , spill all over everything, it didnt matter as long as Larry was getting into the zone.

definately agree with Permo's thinking on that horrible methodology of gating everything to prevent spill , often I saw bands at gigs have to sound check for an hour so the engineer could twiddle his knobs , there half burned by the time there done with soundcheck . My own methodology as 'in house' live sound was simple , no gates ever , minimise mics on the kit , get the sound as good as you can in as little time as possible , appart from vocals dont add anything to monitors unless its specifically requested by one of the musicians . You would get the odd 'special tool' with badly set up effects pedals, amps etc , buzz n honk all over the f**king place , if you happen to mention its sounding like shit and maybe make some simple suggestions to try and remediate the situation and your met with 'attitude' dont waste your time trying to polish turds.

 

[Bo Deadly]

You don't need to monitor the live vocal through the speakers.

Ah. Ok. That makes much more sense. Obviously I don't do much recording. I thought we were talking about monitoring a singers own voice.

 

[GeorgeToledo]

Analog headphone mixing affects performance the least but ditching the headphones and monitoring with a loudspeaker affects it even less! To me, that demonstrates that latency is not really the issue, it's the comb filtering! I did an experiment with some kids in a vocal class, and we found they all sang the best with none of their own voice in the headphones! Unfortunately, most experienced vocalists haven't the confidence to record that way.

Absolutely, Bob.

 

[GeorgeToledo]

The phenomena you maintain is only present with proximity effect is also present with omni mics. Just because in one case it falls above the threshold of audibility and in the other case it falls below the threshold of audibility doesn’t mean the same phenomena isn’t at work. Trying to separate the case of proximity effect versus no proximity effect is a logical error. Correlation is not causation.

If in one case it is below threshold of audibility, we are talking about two different cases.

Anyway, I was speaking to the inversion of proximity effect. It is anyone’s own prerogative to make a point of narrow scope, and that’s what I initially did. I was narrowing scope to a very obvious phenomena that does only occur with a directional mic by definition; inversion of proximity effect.

If someone wishes to discuss inversion of omni and the effect it has on a singer while they monitor through headphones, be my guest. I will suggest to them it won’t include proximity effect, save to note there is none in omni.

 

[GeorgeToledo]

Ah. Ok. That makes much more sense. Obviously I don't do much recording. I thought we were talking about monitoring a singers own voice.

No, this is about a singer monitoring their own voice.

 

[GeorgeToledo]

singer: “When I move up on the mic, the bass isn’t there like I am expecting.”

me: “Can you see if it is in omni?”

(Think about this. Omni, for the millionth time, does not have proximity effect.)

singer: “No, it is set on cardioid, I can see that.”

me: (flipping polarity) “How about now?”

singer: “Yes, sounds fine.”

—

yawn

Meanwhile, the assistant who brought up “name dropping” is taking the bus home.

 

[GeorgeToledo]

Sorry, but is Abbey (or anyone else here) supposed to know who you are?

Not particularly, but if someone wishes to engage in conversation, it is a good idea to not throw out jabs such as “name dropping” if you are actually having a motivation that is in good faith.

Since Abbey came in with jabs like that it seemed like a good idea to identify who I was communicating with, in case there was already some kind of personal or professional level background between us that would better explain the attitude and general attempt at distorting a brief parenthetical comment.

 

[JohnRoberts]

Sorry, but is Abbey (or anyone else here) supposed to know who you are?

Come on don't poke the bear.... he was almost calm, but you stirred him back up.
===
@George... please return to your calm inside voice, and stop the personal attacks.

JR

 

[Gold]

If in one case it is below threshold of audibility, we are talking about two different cases.

If the distortion in both cases is caused by the same phenomenon but you consider them different cases then you are making a subjective argument that audibility is the determining factor. That may work in a practical way but it's not a way to win a technical argument.

 

[abbey road d enfer]

Bleed? No, I think I mean feedback. You sing into the mic, it comes out the speaker and into the mic and out the speaker and around and around and so on. How does that not affect the recording? I guess it could be like singing in the shower but without the standing waves.

No feedback because there was only the playback in the loudspeaker. The singer heard their natural voice. Of course the playback did not need to be very loud.
A standard method at Abbey Road was to use a U48 in fig-8 pattern, with two loudspakers on each side of the talent, so they were in the null. IIRC it's described by Geoff Emerick in his book.. I'm not sure he used that technique with the Beatles, though...
Actually many voice overdubs were made with loudspeakers before the mid-60's. SE's knew how to play with the directivity patterns of the mics and the loudspeakers (very often an Altec 604).
Singers of the day knew how to listen, also. This is a lost art; most singers are deaf and don't want to exercise their audition. Audition is a tool that needs training.

 

[calaverasgrande]

In my experience feedback is as much a function of room modes as it is of the frequency response and directional characteristics of the driver and transducer.
In a small vocal booth that is intentionally deadened you aren't going to start hitting modes until you are at pretty obscene spl. At which point the driver and trnasducer are pretty effectively coupled anyway.
Suffice to say you can get pretty damn loud with a Galaxy or Fostex personal montor with little to no feedback.
Now stick that in the main room. The sound has more opportunity to propagate into the room, excite modes and stick around fro a while to build up on some frequencies and not others. Oh yeah, you have a much bigger volume of a room (with an audible RT60 cluttering up what is heard) so you turn up more to help the artist out.
Here you gotta put the mic in cardiod and stick that null in the direction of the monitor(s).

I recorded a lot of punk rock and metal in the 90's and 2000's. Those kind of vocalists are often insecure about vocals and not happy with headphones and a big LDC in front of them (which they can't touch or 'eat' as they would a 58).
So I gave them a handheld and stuck them a few feet in front of an LDC.
The handheld goes to the monitor. The LDC to tape (or DAW).
I actually liked to record both when possible in case they hit any peaks that exceed the headroom on the LDC. Handhelds usually deal with those kind of peaks in a more acceptable way.
I really think a lot of this monitoring awkwardness is less a function of latency and more of bone conductance.chest resonance. Not buying that peopl can hear latency under 5ms aside from the obvious comb filtering if you mix direct and indirect sources. However you get that kind of comb filtering every day all day, as a function of existing in 3D space and hearing your own voice reflected off of walls and floors. The brain can figure that out.

 

[Newmarket]

Ah. Ok. That makes much more sense. Obviously I don't do much recording. I thought we were talking about monitoring a singers own voice.

But, in any case, you're not bound to get feedback simply because you are putting the mic signal back through the PA.
You have to have enough gain in the system and be in the right (ie wrong) place with the right/wrong directional orientation of the mic to get feedback (or Positive Regeneration to get all 'Dr Who' about it ). Just like guitar feedback from strings/pickup -> amp -> Speaker -> strings/pickup.

If that were not the case then every mic'd up event - rehearsal / gig / bingo caller - would be wiped out by feedback.
And you might still want to add reverb etc for the singer's benefit wrt confidence, tuning etc.

As for method - as others have touched on - minimise cross talk by controlling volume / mic directionality and monitor placement to get as close to null as practicable. And I've heard of wiring monitors in anti-phase to reduce bottom end where cancellation can be more difficult to achieve. I think that came from a tip via either U2 or Duran Duran and I that they have already been mentioned here.
Obviously it's better to be doing this with a band arrangement and performance that you're confident of keeping. Always want to avoid the "Cracking Vocal but can we get rid of the ghost of a cowbell track that we somehow thought was good idea at the time..." scenario !