For the last year and a half my preferred way of composing a mix was to use FL Studio (nee Fruity Loops) as my main mixing surface. Unfortunately FL doesn’t have any time or pitch stretching so when I need that functionality I call up ACID, stretch the clips I need, render them out and load them into my FL project. The recent release of FL Studio 4.5 supports ReWire so I decided to take another looks at Live since it is a ReWire host. (ACID is also a ReWire host but the interaction with FL has been tenuous for me. I am being generous. It’s been flaky to the point of unusable.)

In order to compare ACID and Live, I decided to use a pre-existing long original track and try to add a new beat to it. The end result of both trials are extremely similar (strikingly so) which led me to focus the comparison on the productivity and user interface differences between in Live and ACID. What follows is the result of that investigation. (I’ll get to the actual stretching algorithmic differences in another article.)

The Source

A good “tight” band does not mean they play like a metronome. Just the opposite, it means they can flow all over the tempo as a unit as the emotion of the performance requires. The best jazz recordings in the world wouldn’t stand up to a clock tick for two bars. That’s all fine, but if you trying to do a remix you have two choices: 1) become one the greatest drummers in history and tap on your MIDI pads in perfect time with the Modern Jazz Quartet 45 years after they spontaneously grooved it out (minus eye contact) or 2) straighten the damn thing out.

(There are evil implications here, I understand this as well as anyone, so now would be a very good time for you to get over it. Meanwhile, when the track you’re mixing against cooperates it is not difficult to “re-humanize” the track in either tool, albeit more labor intensive in ACID.)

For the purposes of this exercise I didn’t pay attention to a lot things I usually labor over when doing a mix, so please don’t judge the audio clips below from an audio standpoint. This is a test. This is only a test. Not a testament to my mixing sensibilities. In other words I don’t want to hear any of this “I like the original better” stuff.

I could have just as easily used a long instrumental track that I recorded but I chose “Trouble Man” by Marvin Gaye. It is good example of what happens when musicians imply a tempo more than play to it — as opposed to a track I recorded where I meant to play on the beat but missed. Repeatedly. All of which makes “Trouble” a bitch to use in a remix. But I tried anyway. Click on Vincent above to hear the original.

Method in ACID

I loaded the original song into a track in ACID. I used the BeatMapper wizard (turned on by default when you load a large file) to get the approximate tempo of the song which turned out to be 103 BPM. The wizard intended purpose is no use for my track because BeatMapper only works if the entire track is a solid tempo. Any drifting and ACID will get out of sync in a hurry.

Instead, I selected the track and opened the “Chopper” tool. This part takes some ears and experience: I started marking regions that I thought were “straight enough” to serve as fence posts in the new tempo. With each region I would “Chop to new track…” (big screen image here).

For most of the clips ACID created the new track as a one-shot or beatmapped because, well, it couldn’t figure out what else to do with it. In these cases I had to open the clip properties dialog box and change the type to “Loop” and tell it how many beats there are in the loop. If I wanted to use the original again in another ACID project, I would have to save the clip out from this dialog as well.

Then, as you see in the screen image, I would line them up, back to back. I always lay down a hi-hat, bass drum track to make sure the downbeats and backbeats are falling where they should as I preview the work.

Method in Live

Dragging the clip into a Live track gave me a similar Chopper-like view but there is big difference: the beatmapping is done in situ and there is no chopping step (screen shot here).

In Live I can drag the beat markers and then, by double clicking on them, “pin” a marker to a specific place in the sound file. This means that while Live is playing the clip in the course of the arrangement, it will make sure the markers will line up to the context of the set. You could say that it is dynamically speeding up and slowing down the wave to make sure the marker will be hit on time. Note that I can pin individual beats down to 1/16th of a beat granularity.

On playback I noticed that around bar 17 one of the backbeats was slightly off. No problem, I doubled clicked on the “17.2″ marker to make it sticky. Then, with the whole mix looping on bar 17 I nudged it over tad to catch the snare exactly right, using a combination of aural and visual feedback.

(On re-listening the Live version’s vocal track got a little wobbly — this is due more to my inexperience with Live, working with a demo version, not being able to save results, and haste to get this article out. What do you want? I’ve used ACID for six years. Live for six days. Perhaps I’ll get around to fixing that and upload the results. Perhaps. For now, the important part is the sync with the drum track.)

Compare and Contrast

When I sat down to write this piece I approached it with an open mind. In fact, neither experience in the respective applications left me with a bitter taste so I had no leanings one way or the other even after I performed the experiment. Putting it all down on paper however, I realized just how much the whole comparison tilts in favor of Live.

In ACID each segment that deem “straight enough” becomes a wav file and a track in the ACID project. The “tempo map” for the original track is implied in the way I lay the clips out in ACID which means if I ever want to use “Trouble Man” again, I have to load all the files again and lay them out exactly the same way. The smallest realistic granule of work is one bar (four beats) but the finer the grain the more I am punished by this method because each granule costs me a file on disk and a track in the project. If you made a mistake at the boundary between two of the clips (which is something that never happens to me, I’m just saying if it happens to you), you have to go back to the Chopper view and retrace your steps for both clips.

In Live all the work is done non-destructively (or non-constructively as the case seems to be) on a meta-data file that sits next to the original “Trouble Man” file. If ever want to use the track in another project, I simply load the wav and all the tempo mapping data comes auto-magically along with it. Everything is kept in a single track in the Live project (”Set”) or I can break it apart if I chose to spread it out. The smallest granule of mapping is 1/16th of a beat and the entire cost of going that small is a double-click. In ACID the clip segments are inherently loops which makes it easy to extend, which in turns spurns creativity. In Live, however, all I have to do is select a region of the track and either split or copy/paste and place the result anywhere in the arrangement. With a few clicks in the clip’s properties it can set to loop over any area of the original track. (Remember this is a “virtual” clip, which means no audio data has actually been copied to disk.)

And then here comes my all-time favorite feature in Live: you can even drag the clip over the icon and put it into a Track view slot.

Again, I didn’t plan this but on paper, anyway, it sure looks bad for working in ACID. Of course the code behind what Live is doing doesn’t seem that far away from what the BeatMapper wizard in ACID could do and the dynamic tempo changes at playback is not insurmountable. But Live is there now.

[UPDATE]
Matt Hite from (my favorite daily read) BeatMixed originally wrote this in a comment and I thought it was important enough to post back into the article:

As Matt points out there is more than one way to do this in ACID (and I guess I’m on the hook for reading this book — thanks a ton for that, Matt) but the method described above changes the tempo of the project to fit the song. When you are trying to match beats or another mix it gets confusing because you are no longer working in a what-you-hear-is-what-you-get environment. And then there’s redo-ing mistakes. Still, it’s a perfectly legit way to quantize a long track (and heaven knows I love how they animate the tempo shift).