Skip to main content

Talents Headphone Bleed During Tracking

How much does headphone bleed concern you?

A folk group I was mixing months back had an upper freq problem that drove me nuts. After listening for hours, trying to de-ess but at the same time trouble shoot for the sake of "next time they go into the studio!" it will be better than this mess!

I continued asking the producer to keep trouble shooting what the high end swirl was caused from, he blamed the recordist, the gear "" then I finally discovered multiple vocal takes had various version of headphone bleed. So I asked him, "did you sing harmonies to lead lines that aren't here now, or that were changed? Pitches aren't lining up on bleeds." o_O

In other words, the final harmonies I am mixing, were sung to a lead line now scratched. This was the same for guitars, bass, drums etc! How do I know this? Again, because I hear those tracks from headphone bleed.

What this was is a work in process that became a song ready for mixing and mastering.
This is a common way many people create music in a studios today. We build a song and instead of actually learning the whole thing over, redoing it all over again start to finish, we leave it and mix it best we can.

There is a major problem with this and its in the headphone bleed.

The bleed from all these parts is still in the background. It can be so soft on one track but throughout mulitple versions as described above, will cause all sorts of upper freq swirling effects that become a de-essing nightmare..

Does this make sense?
I hear this all the time. The nightmare of digital editing and headphones bleed. Its in most tracks I get here and maybe we should be talking about this a bit more?

Usually the worst of it gets chopped out during the silent parts so as a mixer coming in on something blind, you don't hear it all.
Why do you think we have so many issues with sss?

Imagine Hi Hats ticking away in everyone's headphones. They are cutting through the headphones and capturing how many times?
They are in the lead vox, harmonies, guitarists playing to vocals and drums etc. Start over dubbing, moving tracks around, compounding this. Now we have parts of parts, aux bleed along with instrumental bleeds, even bleed being picked up from the acoustic guitar box's adding low end and out of sync, out of tune parts to this. I mean, they are acoustic instrument with different days that obviously aren't tuned exact to the last session.
It would be fine if everyone was playing all together but how many of us do that anymore?

The Harmonies and lead lines get sung and pieced together from all sorts of takes of guitars, drums, harmonies etc all the time but how many of us really know what this does to the phase and transients. You can't autotune out of tune vocal bleeds so be sure what you play to, is the vocal track you are keeping, because if you don't, even though you changed the vocal down the road, the deleted one will still be tracked on some part . These accumulates like a virus.

If you've ever lined 2 tracks up by finding the top of the peak. I do this with headphone bleed. I will often listen to the bleed on tracks and line the headphone ticks up. Does this ever teach you something about phase and where ss are coming from!

Comments

JohnTodd Wed, 12/17/2014 - 04:15
Thanks, audiokid.

My singer is an example, too. On stage he sings just fine, but in the studio, he is a bit pitchy. I know he does good on stage because I've reviewed videos of his performances. But in the studio, not as much.

But hey, we only have one pair of "studio" cans between us. Turns out both of us use my cans when tracking vocals.

I don't think it's the cans, but could it be? They are Senheiser HD215. Good sound quality, circumaural, closed back. They don't bleed at regular levels, but they do bleed at "my" loud levels.

Also, am I misunderstanding something about the cue mix? I just put stuff in the cans to listen to. I'll strip out some of the complex keyboard parts, and dwell on drums/bass/guitar and click in the cans. No FX on the singer's voice. Am I doing it wrong?

I've even done this: One can off the ear, and audio cue sent ONLY to the other one that is on the ear. Minimizes bleed nicely when he is singing, but my loud levels still bleed when I sing.

Somebody help! audiokid is gonna get tired of dealing with all my blood spilled out everywhere. Especially since I love to overdub lots of BGVs! With the same mic, preamps, cans, singers, and mediocre sounding rooms!

Boswell Wed, 12/17/2014 - 06:51
To cancel out the bleed, you have to perform a differencing operation between a performing take and a "quiet" take that's identical in terms of what's playing into and the position of the phones but the performer does not perform any singing or playing. If you sent exactly the same phone track both times, you would have to subtract the "quiet" take from the performing take to remove any bleed. If you were to invert the phone track before sending it to the phones, you could numerically add the two resulting tracks and the bleed would still cancel.

kmetal Sat, 12/20/2014 - 02:04
That's an excellent solution I've ised it for hums too.(the phase reversal trick) But it's attacking something that's already tracked problematically. If you have control over the tracks the idea is do t track the bleed.

John your singer is probably just as pitchy live as he is in the studio, the studio is just a microscope for it. I'd turn it down, but u should try those ISO phones, they really work quite well I know from experience. I personally hate phones so j track live when I can. I overdub exclusively wherever the monitor speakers are, and I find most people like the freedom of rocking out in the control room (sorry drummers). A bit of a late reply.

So, congrats John! If ur still uncomfortLe w lower volume maybe something mentioned would help. Cheers man!!!!!

JohnTodd Sat, 12/20/2014 - 06:10
When I turned down, I had to listen intently. Apparently studio recording requires hard work! As an artist, I'm not so good at that. :cautious:

As an engineer, I am. :whistle:

Who knew it would be such hard work?

Oh, wait ... you guys did! LOL!

Can somebody start a thread about what to put in the cue mix? A lengthy discussion on that would be very helpful.

pcrecord Mon, 12/15/2014 - 08:23
I had a few problems with bleeding but not to the extent you are talking about Chris.
Problems when artists would change the song and decide to do a quiet part with a recording that wasn't before.
When you mute the instruments to create that quiet part, those instruments are still present in the headphone bleed.
I also had a few deaf drummers who would ask for very loud clic.

I bought some vicfirth headphones that I use when ever I feel bleeding may become a problem (soft vocals, accoustic guitars and drums)
They are a pain to wear because they are tight and warm but they do the job well.



audiokid, post: 422305, member: 1 wrote: The bleed from all these parts is still in the background. It can be so soft on one track but throughout mulitple versions as described above, will cause all sorts of upper freq swirling effects that become a de-essing nightmare..

That is a serious problem and I'm not sure how I would go at it. Of course a gate or track edit would take care of silent parts but if you hear it with the performance, I would use minimal to no compression and maybe try to EQ out the most problematic frequencies. Or I would try to figure out something with RX4, like the dereverb and denoiser...
Attached files

Reverend Lucas Mon, 12/15/2014 - 08:44
Sort of a long shot here, but could you cancel out the track that's bleeding?

Here's the idea, assuming you have the track played through the headphones still on file.
1) Reverse-eq the track to account for the frequency response of the headphone drivers/muffling by head/response of mics as closely as possible
2) Invert polarity of the signal
3) Mix in the signal to as close to null as possible.

Has anyone tried anything like this before? Does it do more harm than good by introducing garbage to the mix? I suspect that it would be more effective on simple, repetitive signals (clicks and things) than full programs.

I'm not speaking from experience here if you couldn't tell;)

pcrecord Mon, 12/15/2014 - 08:57
Reverend Lucas, post: 422330, member: 48050 wrote: Sort of a long shot here, but could you cancel out the track that's bleeding?

Here's the idea, assuming you have the track played through the headphones still on file.
1) Reverse-eq the track to account for the frequency response of the headphone drivers/muffling by head/response of mics as closely as possible
2) Invert polarity of the signal
3) Mix in the signal to as close to null as possible.

Has anyone tried anything like this before? Does it do more harm than good by introducing garbage to the mix? I suspect that it would be more effective on simple, repetitive signals (clicks and things) than full programs.

I'm not speaking from experience here if you couldn't tell;)

Doing what you say, it would be better to record the instruments through headphones bleeds again (the same if possible) and then use that as a nulling track...
It's an interesting idea ! ;)

audiokid Mon, 12/15/2014 - 09:04
I've got a few EX 29, Extreme Isolation, they are really good headphones like that. They don't sound that great but are very quiet and avoid this, indeed. I'm sure the studio who tracked that band needed better for sure. And maybe the band wasn't holding then tight, who knows.

However, this is a problem with the mass most people aren't aware of, and what its doing. To what extent is dependent on how well you isolate and how much we overdub. I hear this all the time and have tricks to deal with it but the above was a pinnacle moment for me because I didn't realize how bad it can really get.

Bleed is a good thing when its together as a band. The small freq bandwidth that bleeds from headphone pushes upper mids throughout, but an EQ or side chain can deal with that pretty easy. But once we start over-dubing, those slight pitch and phase changes changed, including the now changed ticking on other tracks create phase, they are the worst.

audiokid Mon, 12/15/2014 - 09:19
Reverend Lucas, post: 422330, member: 48050 wrote: Sort of a long shot here, but could you cancel out the track that's bleeding?

Here's the idea, assuming you have the track played through the headphones still on file.
1) Reverse-eq the track to account for the frequency response of the headphone drivers/muffling by head/response of mics as closely as possible
2) Invert polarity of the signal
3) Mix in the signal to as close to null as possible.

Has anyone tried anything like this before? Does it do more harm than good by introducing garbage to the mix? I suspect that it would be more effective on simple, repetitive signals (clicks and things) than full programs.

I'm not speaking from experience here if you couldn't tell;)
I'm pretty certain many of us haven't thought of this, or hears it as a problem. Being a mixer, you are hearing other people mixes. When we record only ourselves, there are a lot of things we don't know about our room, techniques etc.

Consistent bleed isn't that bad and is easily dealt with. The culprit bleed is when we track to different version. The mix ends up having audible various versions of now removed overdubs. You cannot remove this. Its there playing in the background like a ghost.

Example: Band plays together. Acoust Guitar is out of tune but everyone else is good (for now) ;)
Recordist get the guitarist to redo the track. He plays to his last version and gets it close enough.

But, now his new version is different than what was bleeding on the original tracks of everyone. Follow? We now have two versions of acoustic guitar . One that is good on the Acoustic track and bleed from the bad track, still part of the rest of the bands track. Start compounding this now. Mashing up everyone like this. Its a huge problem.

Boswell Mon, 12/15/2014 - 09:23
I've done this sort of bleed correction in the past in the same way as I have taken out floor monitors: do a bleed take while replaying the monitor track at exactly the same volume, asking the performer(s) to keep completely still and silent. After the take, subtract the bleed take from the performance take. It's not perfect, but it can work where the bleed or feedthrough is really bad. It won't work to cancel out a badly played or out-of-tune part.

audiokid Mon, 12/15/2014 - 09:33
Boswell, post: 422338, member: 29034 wrote: It won't work to cancel out a badly played or out-of-tune part.

Right, which is usually why most overdub. Someone is off.
I've heard tracks where someone will autotune or use melodyne to correct the loud stuff, leaving the bleed still there. Client leave it like that and you now have pitch corrected overdubs with out of tune bleed. They'll do this to 20 tracks!

I've heard this on stems released by record labels for remixes. When I get the stems, its impossible to do anything with it then. At that point, its all mashed up and compressed to hell.

audiokid Mon, 12/15/2014 - 09:45
Spectra Restoration tools can get some of this but its pain staking and no matter how we try and clean this up, it effects something else. The best solution, once we have produced music that has all the creative parts right, if you can, retrack everything over with as little overdubbing as possible. It another reason why tape sounded great back in the day. It wasn't the tape, it was we had to be good enough to play from start to finish with less overdubs than today.

audiokid Mon, 12/15/2014 - 10:40
Kurt Foster, post: 422348, member: 7836 wrote: reading this thread it just occurred to me the tools have created the need rather than the other way around. ironic.
Exactly Kurt! , "less is more". Just because we can, doesn't mean we should. To me, its all about source and conversion. After that, its the issues with summing and capture and playback systems.
I hardly use plug-ins. 48 tracks running, I might have 8 plugs running. Samplitude is really the Cadillac for dealing with issues right on the suspect spots which is "Object Editing" . Basic EQ curves and filters are my thing. Even with my analog chain, very little is happening, its there for charter augmentation. Most everything I do is based around filters and M/S approach which is simply a way to tidy up the congestion we can't hear well on one box.
One of the most confusing things I just can't get my head around is why people need so much DSP with a list of 100 plug-ins. Pro Tools, what a marketing machine.

Marco, I've never looked at this:


[[url=http://[/URL]=""]View:[/]=""]View:[/]

audiokid Mon, 12/15/2014 - 11:02
Cool plug-in but I would do this process manually by lining up the peaks and listening to how each phrase glued to the rest of the audio in a mix. I will often have to crossfade sections, move and reverse as Bos described to null because bleed might not syncing with the current tracks that were replaced (overdubbed). Part of my workflow as a mixer is repairing bad punch-ins, clicks and tails left, bleeds effecting the phase, most often very obvious between words or beats.

This is a great audio example of what's happening with bleed, but the product doesn't address removing bleed because on tracks we want to keep. The direct problem is the someone during the recordings "bad out of tune bleed" is part of the tracks we want to keep. And so it goes.

pcrecord Mon, 12/15/2014 - 11:22
audiokid, post: 422350, member: 1 wrote: Cool plug-in but I would do this process manually by lining up the peaks and listening to how each phrase glued to the rest of the audio in a mix.

That's exactly what I do now. But the product seems to do it very well without noticeable artifacts.
Thing is, this discussion about bleeds make me realise that doing it manually or with vocalign may not be a good idea unless the headphone bleeds are very low.. When copy paste and sliding stuff in time. It's surely can build up a mess in the background...

audiokid Mon, 12/15/2014 - 12:49
Kurt Foster, post: 422354, member: 7836 wrote: wait! ......wait! ........ i'm getting something ....... channelling from a "different" wavelength .....

"downward expansion ........ sm57 .............. sm57 ....................downward expansion ....... noise gates ....... "

now i'm really scared!

I don't get it? :)

pcrecord Mon, 12/15/2014 - 14:28
Kurt Foster, post: 422354, member: 7836 wrote: wait! ......wait! ........ i'm getting something ....... channelling from a "different" wavelength .....

"downward expansion ........ sm57 .............. sm57 ....................downward expansion ....... noise gates ....... "

now i'm really scared!

Are you trying to revive Remy's account kurt ? ;)

Tony Carpenter Mon, 12/15/2014 - 18:25
I've been using Beyer Dynamic DT100 old school headphones for years. Had to replace the original worn out vinyl pads with the new comfy cloth ones a few years ago. Only exception, Sony HDR750 (I think ) for drummer, because the BD's are high OHM, I think 400, very quiet sometimes.. I have never understood how anyone could let bleed from headphones become an issue, always seemed natural to eliminate any issues from it, IF present.

kmetal Tue, 12/16/2014 - 03:36
It the case of a whole band playing together the bleed of the acoustic, proportional to the rest of the bands directly mixed sounds, and the new over dub, the momentary dissonance from the bum notes bleed should barely be audible, and if it is not a huge problem, in all but the slickest pop, which probably isn't using live tracking. Even then, you could just edit the live multitrack for that momentary sour not from another section.

I've overdubbed straight over rough tracks, cut in a live room bleeding into evrything, mainly oh and room, and been fine in rock and roll. The ghosting effect just blends into the ambience, and yeah gating and stuff can go a long way to making it disappear if u don't want it.

That said, just in a typical band in stage setup, it's quite remarkable how much isolation you get!

I like those ISO phones for drummers who want it reall loud, and need a have a click. I've also automated the click to fade or mite at the end of tunes.

It sounds to me like you got a sloppily eecorded set of tracks for the genre Chris. Over loud cans, or maybe they used open back cans. Ecessivly high recording levels?

I treat bleed as part of the sound, and usually try notching and de essing, the worst of it. One time I got a live recording w bleed over the leads, and I duplicated the track and nudged the copies a bit, and it "filled" in things a bit. I hope as it becomes more and more common to see an LDC at people's houses, and equipment is just taken for granted, basic recording considerations will be next to spread. I got faith in the future. Even if humans don't learn it, equipment will just do it for us. Yay.

Boswell Tue, 12/16/2014 - 03:47
In a live situation, bleed is something you work with. It's not usually a problem from headphones or even from floor monitors if you are careful, since the main trouble is original sound sources getting into mics other than the one you have set up for them. But the crucial point is that when it's live, it's all the same take (the only one). There is no difference in tuning, rhythm, solos etc that you get in a studio on overdubs by musicians who have come to see not getting it correct first time as their right.

audiokid Tue, 12/16/2014 - 11:18
Boswell, post: 422372, member: 29034 wrote: In a live situation, bleed is something you work with. It's not usually a problem from headphones or even from floor monitors if you are careful, since the main trouble is original sound sources getting into mics other than the one you have set up for them. But the crucial point is that when it's live, it's all the same take (the only one). There is no difference in tuning, rhythm, solos etc that you get in a studio on overdubs by musicians who have come to see not getting it correct first time as their right.


Bos is precisely following. (y)

And I'll say it again, Bleed is a good thing if its in time. But when its not, and when its accumulating by means of overdubbing, consider this an essential look into your next mix. Bookmark this thread.

Next time you are mixing. Take a deeper listen to the headphone bleed in all the tracks, normalize the levels to study how they are sounding and lining up.

Without being rude or suggestive we are all loser's here, I doubt many of us are fully aware of what this is really doing. I guarantee mass thinks they aren't subjected to this like I was. Think again. I make a living at this section more than I should even be sharing.
The moment two bleeds are out of sync, is 1 too many.

My point here, it took a bad version of this to get me to look deeper into something I am now blown away from. Knowing this will add another (that 2% towards wow). This is one of those dark distractions fooling us to look elsewhere when mixing.

Kurt Foster Tue, 12/16/2014 - 11:33
this is a perfect example of why we should make decisions and then stay with them .... practice "proper" techniques to avoid issues.

when i track, i like to get takes with the whole band and then go back and punch in to fix a few clams (usually the bass) .... in this case, phones are not necessary unless requested ... do several passes and then edit different passes for the perfect take. decisions made, problems avoided.

i think a lot of these kinds of problems happen to home hobby recordists who are recording by themselves. they may not be aware of what they need to do to get useable tracks.

it might not be a popular thing to say but if i were accepting projects, i wouldn't touch something with these kinds of issues. i would tell the client to do it over right. even when recording by myself, i can get tracks sans phone bleed ....

for example; when recording guitars or keys, record direct then re amp ...

pcrecord Tue, 12/16/2014 - 12:01
audiokid, post: 422404, member: 1 wrote: And I'll say it again, Bleed is a good thing if its in time. But when its not, and when its accumulating by means of overdubbing, consider this an essential look into your next mix. Bookmark this thread.

In theory, any bleeds (live through monitors or through headphones) could become a problem depending on the distance and how loud it is from the mic.
Of course it's a lot worst if you copy and paste stuff around and/or time drag ...

audiokid Tue, 12/16/2014 - 12:18
Kurt Foster, post: 422405, member: 7836 wrote: this is a perfect example of why we should make decisions and then stay with them .... practice "proper" techniques to avoid issues.

when i track, i like to get takes with the whole band and then go back and punch in to fix a few clams (usually the bass) .... in this case, phones are not necessary unless requested ... do several passes and then edit different passes for the perfect take. decisions made, problems avoided.

i think a lot of these kinds of problems happen to home hobby recordists who are recording by themselves. they may not be aware of what they need to do to get useable tracks.

it might not be a popular thing to say but if i were accepting projects, i wouldn't touch something with these kinds of issues. i would tell the client to do it over right. even when recording by myself, i can get tracks sans phone bleed ....

for example; when recording guitars or keys, record direct then re amp ...
Without doubt Kurt.

But, that pro world is diminishing; or lets say, the home studio and "2 track at a time" might be the most popular method for musicians to make music today.
I will say it again, there hasn't been one mix (recorded on a DAW) where I didn't do something because of some timing issue bleeding / phasing out.

To expand, one of the worst offenders, contributors to sss and spiky upper mids is a direct correlation between cymbals/ hats and vocals. Which, are often in bleed that has been altered from the original passes. So, what we end up with are bleeds that should be there to add a constant and natural sync now changed. This not only creates a weird change in volumes, the slight changes from overdubs now create a swirly effect like two instruments of the same duplicated and shifted by a byte that randomly swings all over the map!

Its a huge problem.

Maybe.... why vinyl, tape, analog, old school etc all sounds better. Follow me? The gear back then didn't allow us to edit so much. We played the stuff through instead of what I am now pointing out as possibly the sole reason why digital audio sounds so unnatural to us.

OMG, the new generation edits everything. Thats all it is. One big edit path where nothing is real. Listening to the headphones is a game changer.

Here is a secret weapon that I've done for years. When I produce something and have the opportunity to play my guitar, I will track my performance start to finish and hide it in the background. I do this so there is something that ( maybe even subliminally) adds a sense of real to the music. I swear its a mojo maker.

Member Tue, 12/16/2014 - 13:54
well, all that may be. But ...one of those demo tracks I posted that you said had audible heapphone phasing, I didn't even wear headphones on... the 414... Because I was only doing half a verse for sake of analysis, I didn't bother to monitor the DAW playback. In fact, I didn't even have playback running. I recorded it a ccapella to a blank track on a new project line... so, I dunno.

FWIW

audiokid Tue, 12/16/2014 - 14:41
DonnyThompson, post: 422418, member: 46114 wrote: well, all that may be. But ...one of those demo tracks I posted that you said had audible heapphone phasing, I didn't even wear headphones on... the 414... Because I was only doing half a verse for sake of analysis, I didn't bother to monitor the DAW playback. In fact, I didn't even have playback running. I recorded it a ccapella to a blank track on a new project line... so, I dunno.

FWIW
Well, there ya go :barefoot: . I said it sounded like it could be headphone bleed or combing? I hear walls and some weird phase or questionable combing that doesn't sound like the mic's alone. Just a guess but it sounds like around a 4ft to 6ft distance that is producing a combing effect on both mics.
I isolated the suspect freq, saw a similarity and posted this. So, I'm at a loss for you then. Its hard to pin down something I have know idea where its originating from, to boot... what you or others don't hear as well. Go figure... o_O
FWIW, headphone bleed has a very similar effect to what I'm hearing in the upper mids of both. You said it was a texture of nodes removed? I have no idea. I thought I heard this same upper mid on another vocalist (in a past mix of yours) so I was really hoping we nailed in down. Simple easy fix (y)
The vox's alone do include close walls and a low ceiling. I hear a combination of these issues in the both.

In a mix, it all gets washed around and maybe its hardly even worth worrying about.
x