Hi, I have a recording of guitar and vocal that has some headphone leak of the metronome/click track. Is there any way of removing this? I guess some phase inversion trick but will this deteriorate the other frequencies?

--------------------
www.myspace.com/shambolic-charm

If it's only in relatively quiet passages, somebody like the CEDAR bureau service could help (I've seen it done!).

If it's throughout the track however, it may be beyond recovery and will need to be re-recorded with quieter and/or tighter fitting headphones and/or more careful mic positioning to reject headphone spill, or to more closely mike the sound sources versus the headphone spill, or whatever.

--------------------
Dynamite with a laser beam...

Set up your recording space as close as possible to when you recorded the song and get someone to sit in precisely the same place with the headphones on.

Now get them to remain silent while you record a pass of headphone spill.

Now try mixing an inverted version of this recording back into the track.

If you’ve been careful with your re-creation, and you’re very lucky, you might get enough cancellation to make a difference. It’s probably a long-shot though to be honest.

This is something I would do at the time of recording, depending on the artist, once we’ve got a couple of takes down and they’re happy with their monitoring. I most often do it with drummers, since the click spill can ruin cymbal fades. The results are sometimes not great though and I have to resort to other tricks that will not help you here.

--------------------
An Eagle for an Emperor, A Kestrel for a Knave.

I get this problem a lot with drummers as they always need the click to be stupidly high in volume. Iv started to fade the click track down while recording, but they isnt always possible.

Elf - when you say mix in an inverted version of the click, what exactly do you mean?

'Inverted' means turn the waveform upside down. (If this still isn't enough info then ask for more!)

Quote chew_rocket:

---

I get this problem a lot with drummers as they always need the click to be stupidly high in volume.

---

...which will damage their hearing.

If the click has to be stupidly loud then they're not using the right headphones.

And/or are deaf, due to years of playing drums without, er, wearing the right headphones.

HD25s or IEMs FTW.

Izotope RX is very useful for doing this - i've used it for exactly this application on many times.

Yes the best solution I've found is also Izotope RX, using the offline Spectral Editor to select and zap the offending clicks with pretty perfect results.

The main problem is that you can only edit 10 seconds of audio at once.

It strikes me that someone out there, a gifted programmer, could create an algorithm which extracts tempo map info so that it knows where to find the clicks, and does it automatically.

My problem is violinists playing combinations of loud and quiet parts, needing a loud click to stay in time. I can see that the best approach would be to ride the click level during recording, according to how loud the section is.

Quote Dan_Pop:

---

It strikes me that someone out there, a gifted programmer, could create an algorithm which extracts tempo map info so that it knows where to find the clicks, and does it automatically.

---

Well you obviously have the tempo map and so know where the clicks are. The problem is what to do with that information.

You could side chain a gate to duck the track on each click, but that might sound too obvious when working at a level that would significantly reduce the click.

You could apply a phase-inverted version of the click, but the click spill won’t be truly aligned with the tempo maps, since it has gone on a round trip through a pair of headphones, back through a mic and through a lot of electronic wizardry. Maybe you could take a sample of some example spill and use that, but I’m still not sure the results would be so good. A drummer only has to move his head around and you’ve got the click arriving at variously different times in a pair of overheads. When you’re dealing with phase cancellation you are looking for sample-level accuracy – which you’re not going to get with any degree of accuracy. Might be worth a try though.

--------------------
An Eagle for an Emperor, A Kestrel for a Knave.

All of the suggestions above are to my mind useful - but bearing in mind the comments made about iZotope, I just wanted to say that in my experience there may not be a silver bullet solution here. When I've had to deal with click or metronome spill I've found the results to be completely dependant upon the nature both of the click and the material in which it is embedded.

Last year a job of this type went to Cedar, but the client felt that even with Retouch there was too much damage caused to the wanted audio. Wavelab's Spectral Editor (similar to iZotope) also caused significant disruption. EQ notches, ZNoise profiles, nothing really hit the mark. He eventually constructed a tempo map (the track came from a tape master) and played/programmed additional masking percussion...

Let us know how you get on.

Loopy

masking percussion, yes I have been thinking along the lines of something added to cover the click at the moment it's just a J-200 and voice track but it sounded rather nice on its own. I think even a low level single note organ would cover it. Might check Izotope first.

--------------------
www.myspace.com/shambolic-charm

I'm totally shooting for the hip here, but have you tried using a de-esser? I'm sure it won't remove it completely, but it could make it much less offensive without reeking havoc on the rest of your track-- plus you probably already own one.

Just a quick rider - you mentioned a held organ note.

For maximum effect when masking you'll likely benefit from spending a little time analysing the frequency composition and location of the click, then targetting that area with any additional material.

Loopy

Quote Nolum:

---

I'm totally shooting for the hip here, but have you tried using a de-esser? I'm sure it won't remove it completely, but it could make it much less offensive without reeking havoc on the rest of your track-- plus you probably already own one.

---

Generally most click track information is much lower in register than the frequencies attacked by a de-esser. You could of course build your own from scratch using a side chain, EQ and compressor - but it'd introduce some audible pumping. Might help a bit though ....

Alternate solutions for future use are:

Don't use a click for drummers - use some other sound, say a drum loop or a percussion track - they tend to be far less noticeable than a click.

For strings - why not use a conductor? - its probably what the players are used to. Only the conductor needs the click and he doesn't need to be near the mics.

--------------------
Daniel Davis
Edinburgh Recording Studio Windmill Sound

How about this -- it's a bit lateral, but I think it might actually help in this instance without you having to set up any further recording. Try finding a section of metronome click on its own, without the instrument playing, slice it out and then polarity-invert it. If you copy and paste that alongside the other clicks in the audio file, and try to match the waveforms as closely as possible, I think you might be able to pull off a useful reduction in level.

Oh, and if it's Cubase's pitched click, then try tuning a super-narrow notch to its resonant frequencies, which shouldn't harm the instrument too much -- certainly less than the sound of the click!

I'd be interested to hear how well this works -- can you post a WAV of the recording somewhere so that I can have a tinker with it?

--------------------
Mixing Secrets for the Small Studio
A complete mixing method based around the techniques of the world's most famous producers.

Will sort out a post somewhere

--------------------
www.myspace.com/shambolic-charm

So this is what should be done to remove click....

Take a sample of the click (from the start of the song for example) and put it into a new track all the way through the song. Reverse the polarity of the new track and.... this should take the volume of the click in the drum track down?

This is how my brain has processed all the information from the previous posts. Is my brain working properly or should I get a new one?

If you have a clean recording of the click and if it is consistent all the way through, then the method suggested above should work. If the click changes due to head movement etc. this this becomes far more problematical.

If you don't have a copy of the click in a quiet part of the track but have the time and patience you can extract the signature from the recording by getting copies of as many different ones along the track as possible and pasting them into other new tracks, then line them all up. While they will add, being the same, the other material will gradually drop. Very roughly, if you add 10 clicks they should rise about 20dB relative to the rest.

Cut the resultant sound patch tight to the click, invert, make appropriately spaced copies (or put it in a timed loop), adjust level and Bob's yer uncle.

... hopefully

--------------------
It wasn't me!
(Well, actually, it probably was)

Hi, also useful is to use soft percussion as the click next time and maybe also a relevant percussion part as was alluded to earlier.Dave

--------------------
My head hurts!

Quote chris...:

---

Quote chew_rocket:

---

I get this problem a lot with drummers as they always need the click to be stupidly high in volume.

---

...which will damage their hearing.

If the click has to be stupidly loud then they're not using the right headphones.

And/or are deaf, due to years of playing drums without, er, wearing the right headphones.

HD25s or IEMs FTW.

---

The best way to use a click is at lower volumes. Yeh sure , when you're in time you can't hear, and that's the point, you'll soon hear it if you drift out of time. It's a mind set.

Riding the click as mentioned earlier is also a great idea. I have had some success too with the phase inversion route.

Cheers

Sean Kenny

--------------------
Sean Kenny - Ten21 Recordings Studios

Forgot to mention that for drums I now tend to use Vic Firth Headphones. Not the most comfortable cans, but the isolation is very good way better than Beyers and the plus is the drummer doesn't have to have the cans as loud either. Double plus and good for his hearing!

Cheers

Sean Kenny

--------------------
Sean Kenny - Ten21 Recordings Studios