Punch-In Primer

One of the great breakthroughs of the early magnetic tape era was the ability to activate the record function independent of the playback head, thereby allowing performers to isolate and repair intermittent mistakes, rather than starting all over from the top. This technique known as "punching in" remains among the most useful of recording solutions, and here we consider the proper setup for facilitating smooth punch-ins/punch-outs, including where to punch-in during a song (and where not to), using your machine's pre-roll function, plus editing versus punching, and other thoughts.

What is Punching In?
In technical terms, a punch-in involves recording over an errant vocal or instrumental flub with the goal of correcting the mistake without subsequently hearing any stitch marks. To initiate a punch-in, you'd start by rolling the recorded section several seconds from the point that the error occurs, or enough time for the performer to hear the part and get ready to sing or play over the mistake. Once the punch-in is complete, the recording mechanism is turned off, allowing you to immediately roll back and hear if the point of correction sounds right.

Setting the Pre-Roll Function
Though you can facilitate a punch-in manually by simply playing the track and hitting record at the desired spot, letting your machine automatically punch in for you keeps your hands free, allowing you to focus on making the smoothest repair possible. To start, set your recorder's "pre-roll" function-that is, the amount of time prior to the punch-in point that the machine will roll the track without recording. For instance, a full 10-second pre-roll will give you ample time to get into position, particularly if you're tracking in a different room; or you may choose to set a shorter pre-roll time if you're simply patching over while sitting in front of your recorder.

Best Practices for Punching In
When recording a punch in, the idea is to get the "punched" part to blend as much as possible with the previously recorded section. This means looking for a spot just before the errant part to activate the punch-in, as well as a point shortly after the fix to punch-out. If it's a vocal fix, for instance, try to find a place where you're not singing, even if it's just for a second or two, rather than trying to punch in directly over the vocal. The same goes for lead guitar or other solo section-just set your machine to break in at any point where there's a brief pause and take it from there. If you can't find a convenient punch-in spot, try to match the feel of the section you're recording over by playing along with the track for a few seconds leading up to the auto-punch (such as when fixing a flub in a rhythm-guitar track).

Comping Versus Punching
If the goal is to arrive at a flaw-free take, there are other ways to get there besides punching in. Rather than continually recording over the spot where a mistake occurred, you'd simply try another separate take, or several more takes, until you're reasonably satisfied with the performances on each. Using your editing facility, you'd then stitch together the various parts until you have seamless, error-free whole. This method of compiling or "comping" different takes can be a good alternative to punching, especially if the performer is having trouble getting the punch-in parts to match. The downside is that comping means more work for the engineer and can get tedious if there are too many parts to compile; accordingly, limit the number of takes to be comped in order to make the job more manageable.