About crossfades and Punch ins

  • Thread starter Thread starter undrgrnd studio
  • Start date Start date
undrgrnd studio

undrgrnd studio

New member
I use Adobe Audition for recording, and whenever I need to do a punch in I need to find an area with a bit of silence to allow me to punch in cleanly. I was working with a drummer a few weeks ago who was surprised that I couldn't punch in his drums after completing a perfect verse but screwing up the chorus.

My position is that drums usually don't have silent areas that allow for punch ins. The cymbal will usually be ringing purposely through a pause or there is no pause at all between sections in a song. He said when he recorded at a pro studio he was able to do half a song and screw up, then just immediately recut but keeping the good parts of the original recording without any noticeable difference in the final track.

How is this done? Was the engineer recording two tracks in pro tools and then splicing them together using crossfades.

I could REALLY use this technique for doing difficult drum tracks on prerecorded songs. Sometimes it's just impossible to get through a really complex song on drums without a timing issue, clip, or something causing a problem.
 
Can't really go into too much depth as i was in the band not the engineer but..

Thats pretty much what we did... started playing from a point well before the fuckup so all the cymbals were ringing correctly and then punch in and cross fade the two recorded parts.
 
I find myself punching in on my drum tracks often...with no problem. If you zero in on the overheads, you'll find lots of 0 points to punch into, as long as like you mentioned, there's no cymbal ringing through. Just give the drummer enough pre-roll to fall into the groove before the record light comes on.
 
And be flexable about where to crossfade. More often than not, the place I end up doing the crossfade is not where I first thought it would be. In a good editor, you can drag the cross fade forward or backward while looping over the section, till you find the right spot. You may also have to adjust the crossfade width, and the curve on each side to get it right. Try to get it nearly undetectable while solo'd before you rely on the rest of the mix masking it.
 
Corssfading 101

OK I have to admit I have NEVER used crossfading before. I understand the concept of it and I can see it in mind on an old analog board.

But the problem as I perceive it is, wouldn't the two sections overlap enough that if the drums weren't played exactly the same you would get a double bass hit or phantom crash symbol stutter?

The end section fades out as the new section fades in, right?

So should I, or could I cross fade two tracks in multi track mode so that the two sections overlap each other by a few seconds or milliseconds? I assume that's how it's done, then you bounce to a new track right?

If I'm off base here I could use a little direction.
 
Yea, if you pick the wrong spot it wouldn't line up. The right spot is where several consecutive hits occur simultaneously. Then as it fades out, the second take fades in in perfect time. Just be careful that the area that you fade out and the area you fade in aren't too far apart or else there will be a slight drop in volume. This is automatic in most programs.

But, as a general rule, the drummer should be able to play the song all the way through without any major screw ups before he/she enters the studio. A few minor mistakes give the track "character" and usually that's not a bad thing.
 
hey I haven't done this in years but what I came up with was not a cross fade. I would listen to the track and drop a marker. then I would zoom in and move the marker to a "zero crossing" and split the tracks right at the marker. Drop in the new tracks on top of where the old ones were. Then I would do a very, very fast volume dive over the split point. just a few samples wide and you could never hear it.
 
FALKEN said:
hey I haven't done this in years but what I came up with was not a cross fade. I would listen to the track and drop a marker. then I would zoom in and move the marker to a "zero crossing" and split the tracks right at the marker. Drop in the new tracks on top of where the old ones were. Then I would do a very, very fast volume dive over the split point. just a few samples wide and you could never hear it.
Click to expand...

I don't get it, could you be more detailed? What is a zero crossing?
 
"Zero crossing" is where, on a digital editor, the waveform crosses the zero center line in the timeline display. The idea is to butt joint two clips together by matching up the end of one clip to the beginning of the other, with each happening at their respective zero crossing, and with the waveform traveling in the same direction (up or dowm). The volume dip right at that joint that Falken mentions can help smooth that transition even more.

I have used Falken's method as well. It's a situational call as to whether the butt joint or the crossfade is preferable, IMHO. One of the factors that helps make that call is how much ambience there is in each clip. If either clip has a fairly noticable amount of ambience in it, I find crossfades can be a bit smoother. Even a perfect zero crossing match on a butt joint can be obvious if the overall envelope of the sound is different in the surrounding samples. The crossfade can help provide a smoother transition in the overall ambient sound.

If ambient sound is not an issue, though, sometimes the ol' butt joint can actually be a cleaner, more surgical edit.

BTW, matching of zero crossings is just as important in crossfades as it is in butt joints. You have a little more flexibility in the exact position of the match - somewhere near the middle of the crossfade - but it still should usually be made.

G.
 
SouthSIDE Glen said:
"Zero crossing" is where, on a digital editor, the waveform crosses the zero center line in the timeline display. The idea is to butt joint two clips together by matching up the end of one clip to the beginning of the other, with each happening at their respective zero crossing, and with the waveform traveling in the same direction (up or dowm). The volume dip right at that joint that Falken mentions can help smooth that transition even more.
Click to expand...
Just to add, the zero crossings, while being spots where you'd have the best chance of a click free cold punch, obviously aren't going to line up on a second or more track. Thus your looking for the logical spot in the performance to punch and second, where the energy is lowest on the the tracks. Moving on to sliding or blending the fades on the separate tracks opens up the options where the punch spot is a kludge.
It's amazing to me how much you can get away with on a kit for example, where you'd expect it wouldn't work. But with things happening in ms, the ambiances, different tracks offset a bit perhaps.. good to go. :p
 
Hmm, I do my most of my recording on Saturday nights, I'm going set up my drum mics and see if I can't bang out a drum track using some cross fades. I suck at drums, but I can make it through a verse and a chorus easily enough. I just can;t make it through a whole prerecorded track. It made me feel better to see a pro having trouble on some of the same areas though. Except he eventually got the whole track in one piece within about half an hour. Whereas I would have taken all night, and STILL not gotten it right and even if I did it would sound crappy compared to his skills. I'm going to dive into crossfades and see what kind of magic they can make for me.
 
how bout practicing the part until you get it right? that's the best solution rather than crossfades.

and that's the difference between the pros and the amateurs......the amateur practices until they get it right........the pro practices until they can't get it wrong.


cheers,
wade
 
mrface2112 said:
how bout practicing the part until you get it right? that's the best solution rather than crossfades.

and that's the difference between the pros and the amateurs......the amateur practices until they get it right........the pro practices until they can't get it wrong.


cheers,
wade
Click to expand...
We all see this same response in just about every thread; if everybody could be an expert at everything none of us would be here looking for advice on more efficient ways of home recording. But that's not the case. Maybe I spent my time learning other aspects of music. Not everybody is a pro drummer. Some of us excel at other things like songwriting or rhythm or bass.

If I spent my time playing drums for the last 15 years maybe I would become a ninja drummer that can execute complex precision patterns in a 1 hour session to a song I'm not familiar with. Or maybe I might try to release some stress on myself or the guest musician and get a good scratch track down for demo purposes as quickly and improvisational as possible?

Sometimes something is better than nothing no?
 
The man has a point. The question was about a recording technique.
And who hasn't said 'let me punch it there. I can do it better.'? Pro or no. :cool:
 
Yeah, but the thing is, crossfading and splicing parts together works better with a drummer that can play a part consistently the same way for the most part every time, except they have a mis-hit or something that needs to be fixed. If you are just playing whatever randomly comes to you and hoping you can splice several mediocre takes together to make something good, you are going to have a much harder time of it...
 
Reggie said:
Yeah, but the thing is, crossfading and splicing parts together works better with a drummer that can play a part consistently the same way for the most part every time, except they have a mis-hit or something that needs to be fixed. If you are just playing whatever randomly comes to you and hoping you can splice several mediocre takes together to make something good, you are going to have a much harder time of it...
Click to expand...
+1 +1 +1

thanks Reggie......this is what i was getting at, and i hadn't replied to mixsit b/c i couldn't find a simple way of putting it. you said exactly what i had in mind--have some chicklets! :D

i mean, how much time do we spend talking about "it's all about the song and the performance" around here? this is a case of using the computer to fix the performing power (or lack thereof) of the musician. the best result will ALWAYS be achieved by a good performance.

and there's nothing that says one has to be proficient on everything. that's why we hire folks to play parts we can't......or we have friends who owe us favors. ;)


cheers,
wade
 
I approach recording myself the way you just described

I will spend hours on one vocal track if need be. But drums are a different animal. If I have a good drum line but I'm simply not a good enough drummer to do it perfect through the whole 4 minutes, chances are I won't get anything at all. I'll have an awesome drum track for 3 minutes, and then a timing bobble. So if I could find a way to use crossfades to allow me to punch in drums the way I punch in vocal tracks or guitar solos, well that would make my whole world a hell of a lot easier.

And I don't buy the argument that I'm using the computer to fix my problems, It isn't the computer that wrote the songs, played all the parts, learned about home recording, spent hours reading manuals on complicated gear, and then learned to play drums when noone else was available. The computer didn't do that, I did. All the computer is doing is what I ask of it, I'm asking it smoothly transition a punch in. And then I'll go back and do the drumming. I think a drum machine is cheating a hell of a lot more than a punched in drum track.

I figured it out and did a test track to see if I could hear the punch, I couldn't hear a damn thing. So next time I need to lay down some drums, I have the tools to allow me to do it all by myself instead of relying on someone else. Self reliance is the most efficient recording technique I ever learned.
 
undrgrnd studio said:
And I don't buy the argument that I'm using the computer to fix my problems, It isn't the computer that wrote the songs, played all the parts, learned about home recording, spent hours reading manuals on complicated gear, and then learned to play drums when noone else was available. The computer didn't do that, I did. All the computer is doing is what I ask of it, I'm asking it smoothly transition a punch in. And then I'll go back and do the drumming. I think a drum machine is cheating a hell of a lot more than a punched in drum track.
Click to expand...

Right on Undrgrnd! Punch ins and crossfades are nothing new. They predate computers, and have been used on all manner of professionally recorded tracks of professional musicians over countless albums. Pro's who had no other job than to just play one instrument.
 
I think face and reggie have hit the nail. The issue is not whether punches and edits are a good thing; of course they are a good thing that have been done with positive results ever since we evolved past wax cylinders.

But there comes a point when punching up a worthy performance via punch-ins (good) becomes synthesizing competence from an unworthy performance (not so good). It's one thing to remodel a house, but there comes a point where it cheaper, more efficient, and better results-yielding to just raise the old house and build a new one.

Where is that point? I'm not sure that is so easily quantifiable when it comes to music performance and recording. Nor am I sure just where undergrnd may sit on that spectrum. If he's just punching in a bar and/or or a fill or two - i.e. correcting occasional glitches in an otherwise competent performance, that's just fine. But when it comes to having to assemble and synthesize a workable take because of fundamental incompetence on the part of the musician (unable to, with some modicum of consistancy, execute the fundamentals like keep a beat or to hit a skin or to strum a chord or to tune their instrument with consistancy, etc.) then I think that crosses the line where it's best to just dump the take and either replace it with something at least competent, or to wait on hitting the Big Red Button until some level of competence can be acheived.

Perhaps at least one line in the sand might be put this way: if a track has been so Frankensteined that the resulting recording is as much a document of the engineer's performance as the musician's performance, that should be a red flag that there's something majorly wrong. And no I'm not talking about fancy tricks to the timbre of the sound, I'm talking about corrections for performance.

Some of us long for the day when "direct to disc" actually was a good thing.:D

G.
 
You must log in or register to reply here.
Back
Top