Jump to content
  • Sign in to follow this  

    Splice Your Way to a Better Performance

    By Anderton |

    Splice Your Way to a Better Performance

    Loop recording may not be the best way to create a perfect take...

     

     

    by Craig Anderton

     

     harmonycentral-spiceyourway-craiganderton-leader-d25e0e4c.jpg.7edd774dbf6ac05ad6083a0062c812ac.jpg

     

    To create the “ultimate take,” a common recording technique is to record multiple takes of a part, then identify the best bits from each take and splice them together (typically with cut and paste operations) to create a single, “perfect” part.

     

    One way of doing this is called cycle recording, loop recording, or composite recording (for more information on this technique, click here). Typically, the performer records several performances on different tracks while listening to a click or rhythmic reference, so that the different performances line up properly in terms of phrasing. Then you simply cut away the bad sections, leave the good sections, and optionally, bounce the remaining sections to a single track.

     

    But what if you’re recording something like solo piano, and not playing to a click? Traditional loop recording is usually not the answer. In this case, it’s often best to record multiple consecutive takes, cut away the sections you want to replace, and find better equivalent replacement sections in other takes. However, you have to pay careful attention to cutting at the right place, and crossfading properly—as shown in the step-by-step screen shots.

     

     

    step1-1c7f4fd0.jpg.8a5ca9f59a71c1be5c5c4f0d2c0e0d71.jpg

    Step 1: Record multiple takes. Make sure that all the takes are recorded under the same conditions: Same mics and mic placement, same tempo, and same dynamics—you don’t want to hear a noticeable difference when you combine sections from different takes. This screen shot shows two complete takes (colored in in blue for emphasis) using four tracks (left and right main mics, and left and right room mics). The orange sections are redos of the beginning, which was particularly difficult.

     

     

    step2-a9c79b95.jpg.279abc7aca5a930ab55264a7ce726726.jpg

    Step 2: Identify sections that need to be replaced, and find suitable replacements. This shot highlights equivalent sections in the music, but in different takes. Inserting markers can help here, as most programs have “shortcuts” that let you navigate instantly to different markers. For example, the markers in the second take define a section that will replace the marked area in the first take.

     

     

    step3-b99f9d4d.jpg.4eb94f59d17f2ce7c8e26c31c1fc5b58.jpg

    Step 3: Find the best cut points. Where you cut a section is crucial. Generally, the best splice points have strong initial transients; cutting at a transient will produce a less noticeable splice than, say, cutting in the middle of a sustained section. Zoom way in to see the exact beginning of the transient. Also note that the best splice point might be before or after the part you want to replace, or you may need to replace an entire phrase (not just the “bad” section) to create a better flow.

     

     

    step4-681b306e.jpg.d18017502e3fbb6487594a4b977d4cf1.jpg

    Step 4: Replace the section you don't want with the better version. When you paste, butt the transient from the replacement section up against where the transient was cut from the original section. Then, to avoid any clicks caused by level discontinuities between the two sections, create a crossfade (a crossfade occurs when two sections of audio overlap; the first piece of audio fades out while the second one fades in, creating a seamless transition between the two). When doing the crossfade, extend the section before the transient so it overlaps with the section leading into the transient. Even a short crossfade (e.g., 10ms) can be sufficient, but try longer fades—they might give an even smoother effect.

     

    However, note that the crossfade curve is important—a linear fade out with a linear fade in will sound like there’s a slight volume dip at the transition point. Most crossfades use “equal power” curves (as shown in Step 4), where the two fades are exponential and complementary. -HC-

     

     

    ___________________________________________

     

    avatar-90ad537b.jpg.5c2d0f473f150a41ae137f61d634f219.jpg

     

     

    Sign in to follow this  


    User Feedback

    Recommended Comments

    There are no comments to display.


×
×
  • Create New...