Editing Audio With Live
By Chad Carrier
Those of you who use Live regularly probably know that the program excels at letting you tweak your sounds into oblivion while leaving the original audio files intact. This is known as nondestructive editing. You can change all sorts of things about a clip, such as transposition, sample offset, volume envelopes and more, all with the reassuring fact in mind that you're never doing irreparable harm to your source files (some recordings are one in a million- especially if you have bad backup skills!). Nondestructive editing is also efficient when it comes to hard drive utilization, since you can have multiple clips in your Set with different settings, all referencing a single source file.
There will be times, however, when nondestructive editing isn't enough- when it's best to change the source files permanently. Perhaps you have a guitar sample that contains audible amp noise before the actual guitar part begins. You could, of course, simply use a volume clip envelope to silence the noise, but this would mean that you would have to do the same thing every time you added this sample to another Live Set. On the other hand, if you remove the noise from the file itself, you'll never have to worry about it again.
In all actuality, Live only performs nondestructive editing. In the following examples, you will see that as you start committing changes to audio, you will always end up creating new audio files rather than overwriting the originals. The benefit of this is that you'll always have access to the originals if you need them again.
Bounce to Track, Resample and Render
Let's say that we have silenced unwanted noise in a clip with a volume clip envelope. This clip now sounds just the way we want it to. In order to commit our change to a sample on disk, we'll need to re-record this clip as a new clip, thereby creating a new sample. There are three ways to do this in Live's Session View- use whichever option suits your music production style and needs:
Bounce to Track
The cleanest way to permanently record changes to a clip is to record the output of the clip's track directly into another track, a process sometimes referred to as bouncing. The figure below shows a setup in which the input of Track 2 is "listening" to the output of Track 1 (notice that "1-Audio" is selected from Track 2's Input Type chooser).
After you've recreated this routing setup for your scenario, launch the clip and make sure the output of Track 1 is not clipping (red will appear in its meter), and then press the Stop button in the Control Bar (or tap the space bar). Live will stop, but the clip will still be set to play the next time Live starts. Arm Track 2 and click one of the round Clip Record buttons in the clip slot grid. A new clip will begin recording, and Live will start playing the original clip at the same time, as shown above. Once you've recorded the duration of the clip, press the space bar or Stop button again.
Resampling
Another way to record a new sample in Live is to resample the Master track, which has the benefit of allowing you to record any effects housed in the return tracks. (With the previous method, the only effects recorded with the new sample are those housed in the original clip's track). The figure below shows this setup; notice that "Resampling" is now chosen from Track 2's Input Type chooser. After setting up the routing as shown, the recording process is the same as described earlier for bouncing.
Render to Disk
A final option for creating a new sample in the Session View is provided by the Render to Disk command found in Live's File menu, which also records the Master output. The benefit of this method is that you can numerically specify the length of your recording- you don't have to stop the recording by hand once you've reached the desired length. However, when the recording is complete, a clip of the rendered file will not be added to your Live Set automatically, so to use the new sample, you'll need to drag the file from the Browser into your Set.
Consolidate Vocal Edits
When I perform, I'll occasionally throw in part of an acapella track for added flavor. But, depending on the occasion, I may need to "tone down" some of the vocals. For example, the track shown in the figure below by Busta Rhymes contains some "offensive language." We can easily create a clean version of this track by reversing the profanity as each word comes along.
To begin with, we move the Busta clip over to the Arrangement View, where we can begin chopping up the vocal and flipping words around. In the figure, I've isolated a word from the vocal track by using CTRL (CMND) + E to split the clip in two places. Now, I have three clips, all referencing the same sample but playing different sections.
I want to flip the second clip around because it contains the profanity, but I will not click the Clip View's Reverse button just yet. Why? Clicking Reverse will cause Live to process the entire length of the sample and will take a while to process.
The trick is to create a new file just for our profanity clip to reference, and we do this with the Edit menu's Consolidate command. For those of you unfamiliar with Consolidate, it takes all the clips within a given range of an Arrangement track and condenses them into one clip. Live does this by recording a new audio clip comprised of all the clips within the selected time range of the track. Once the consolidation is complete, the original clips are removed from the Arrangement (though the original sample will not be deleted from the hard drive) and only the new clip will remain.
Consolidate works on any number of clips from any number of source samples in one track. However, in this case we will use Consolidate on just one clip. If we select the second clip in the acapella track and press CTRL (CMND) + J (Consolidate), a new clip will be recorded that consists of only the audio contained within the length of that clip. Now, there will be two audio files on our hard drive, one that is the original acapella (a long file) and the other that is just a bad word (a short, little file). The end result is that if we now reverse the newly consolidated clip, the reverse will be calculated almost instantly.
After you go through the process of isolating all the "colorful metaphors" in the track, it will probably look something like what is shown below.
There must be about 40 different edits strewn throughout this track. Consolidate will now come in handy again: Select the length of the track and press CTRL (CMND) + J to compile all of these little clips into one large clip, as shown below. The result can now be copied back over to the Session View. As a reminder, we can add the word "clean" to the end of the file name, and the tame acapella clip will be ready to use the next time we need it. Remember that even after all of this editing, the original version remains untouched, so we can still use it from time to time, too!
Unloading the CPU
Rendering new audio files can help keep your CPU usage under control. If you're using a particularly complex virtual instrument in your track, it may be taking up a large portion of your computer's processing power. Oftentimes, especially if you're using effects with the instrument, playing an audio file will be easier for Live as compared to generating the sounds in real time with an instrument. In this case, rendering the part to a file will allow us to remove the instrument and its effects when we're done, freeing up the CPU for other tasks.
The process is the same as re-recording the guitar take in our first example: Create a new audio track and set its input to the track containing your virtual instrument; record the instrument part into a new audio clip as usual; after recording the part into a new track, the devices can be removed from the original.
This trick is really helpful when the part is being generated by more than one instrument. In the figure below, I have a drum part that's being played by two different instruments. Waldorf Attack is being used for the kick and hi-hat, Native Instruments Battery 2 is providing the snares and cymbals, and two audio clips are layered on the top. Resampling all of this into a new clip will provide a significant savings in CPU power. In the figure, you can see that Track 6's input is set to "Resampling," so it will record the entire mix of drums into one clip. For later flexibility, I've recorded multiple versions of the beat, each with a different combination of instruments muted.
Resampling can be used to augment virtual instruments, as there are a number of expressive tweaks and modulations that can only be performed on audio clips. For example, transposing MIDI information up or down sounds different than shifting the pitch of an audio file up or down (see the figure below to see how this might work). Transposing MIDI causes the instrument to play different notes, but shifting audio requires resynthesis, which has its own distinctive sonic character. Once you've created a MIDI part, you can resample it as illustrated in the previous examples, and then edit the audio clip. Try blending the tweaked audio clips with the original MIDI part to create really fat and intricate sounds!
In summary, don't be afraid to render new audio in Live! You'll never damage any of your original files beyond repair, and you may stumble upon some interesting sounds along the way. Furthermore, Live is quite efficient at keeping your hard drive clean- if you render or resample audio that you don't end up using in your Set, Live will ask you if you'd like to save or delete it the next time you save or quit. Most of all, judicious use of consolidating and resampling will keep your CPU usage low, leaving you plenty of extra processing power for additional sound creativity.
Chad Carrier is a Product Manager at M-Audio in California and has just completed a book called Ableton Live 4 Power! with Ableton's own Dave Hill, Jr. The book is loaded with tips, tricks and insights, as well as an in-depth explanation of all the new features in Live 4. It is available now from Thomson Course Technology.