Hi all. I'm trying to figure out how to get started on the recording of a song of mine. It's easy enough to play live; I just slow down and speed up to taste as different sections arrive. Now to record it and add bass, drums, etc. I realize that I need a plan or the recording will likely sound very sloppy since there are frequent tempo changes. There will be many instruments and voices involved, all layered on by me, so it will help if there is some sort of established tempo to follow.
 
I have figured out that I usually do the many verses at about 170, the two bridges at 125 or so and both choruses blaze along at approximately 205.
 
Should I record just the bare bones of each section into one project but not try to place them into their eventual final order? For example:
1. Start at 170 and record the acoustic guitar strumming the chords for all the verses one after the other (later placing markers to separate the different verses, since they're not all the same length) (or maybe leaving a couple of empty measures between verses)
2. Leave a few empty measures and change tempo to 125 and do both of the bridges one after the other (they aren't the same length either, just both the same tempo)
3. Leave more empty measures, change tempo to 205 and do the choruses, intro and coda 
 
 In this case, would it make sense open a different project and try to assemble (by pasting from the first project and adding markers) the verses, bridges and choruses in the order they really belong in?
 
At this point, I'd start adding vocals and other instruments to the second project to try to get the "real" song.
 
Or am I thinking illogically here? It's a very wordy and I hope, comical song that should come in at about six minutes in length. It's hard to get a handle on the right way to get started.
 
Thanks for any help!
LJ
 
 

I would suggest you think in terms of one song.

Start thinking scratch takes to get the structure. At the points you need to speed up or slow down go to View Tempo. You can insert tempo changes as needed at your desired points.

Capture a full song by recording what you're attempting, but capture your scratch rendition which will allow you to start and stop, insert tempo changes, and ultimately markers. You can record your scratch performances on different tracks so you're able to jump right in at your necessary playing style per section.

After you have the structure and tempo captured you can record as appropriate to get the real takes.

I would lean toward doing it as one project.
Use the tempo map to your advantage. MIDI will inherently follow the tempo. Audio, not so much.
I t looks like you have the overall song concept in place, so an audio scratch track might be in order.
Play the song through, and use that to define your tempo map.
Don't worry if it's a little sloppy. Consider this first track disposable. You can refine it as you go.
Do plenty of "save as" versions. If you find the song heading in a direction you don't care for, you can back up to a version you liked.
Don't over-think it. You have some remarkable tools at your disposal.

I don't see any good reason not to put the entire song in a single project at the outset, especially if you already have a structure and arrangement in mind. Sonar can handle tempo changes quite easily. I would only use separate projects if the different parts of the song were intended to sound distinctly different in terms of instrumentation or ambience, e.g. an orchestral section followed by an acoustic section.

I agree with the one project approach. A couple suggestions which will make things easier for you:
 
   Shift-M - "Set Measure/Beat at Now..." - this will automatically insert project tempo changes, which is crucial to mate MIDI work with an audio track. SONAR can readily handle tempo changes without issues for you. Even without using a metronome, you can record an audio track and use Shift-M every couple bars or so to align the bars to the music. When done, if you hit View->Tempo you will see all of your tempo changes.
   M - "Insert Marker..." - this will make it visually easier to find thing and edit (especially years from now).

lawajava, RobertB, Kev99, and metellus, I want to thank you very much for taking the trouble to reply. The information you've given me is invaluable. I never thought of approaching the song that way at all! I have never used Sonar's tempo map to estimate the tempo of a live performance and, though I had heard it mentioned in passing, it never really struck me that it could prove extremely useful in a situation like mine.
 
I will try recording a scratch track as you recommend and inserting shift-m as appropriate and see where it takes me. It will be great if it works, because I'm sure it will sound much more natural and spontaneous than the rigidly pre-planned route I had come up with.
 
One more question, and sorry if it's a dumb one: The song is in 2/4 throughout (it wouldn't work properly in 4/4 at all as several important parts would then come in on "3" instead of on "1"). Would I first set that as the time signature and then start playing the song with no metronome? Or does it not matter what time signature I specify because Sonar will determine the best one automatically? I'm just concerned that Sonar will read the time signature as 4/4.
 
Thanks again for all your help. You have likely turned a real chore into a pleasure :)
 
 

You can update time signatures as needed. There have been a variety of previous threads/answers about that previously.

Here's one:

http://forum.cakewalk.com...ignature-m3096154.aspx

+1 with time signatures... SONAR does quite well with many facets of creation.
 
Without a full understanding of what you are doing, a metronome can be distracting (especially first pass before you set the tempo map). A "lesson learned" from my own experience is not to focus on "mechanical precision" but rather the intended feel of music you are composing. A rough scratch track with main instrument/vocal is worth its weight in gold, and take advantage of it being a scratch track... slice it, move it, stretch/shrink it, create... with the intent of that being its primary function (to be deleted). I have fallen into the pitfall of trying to "perfect" a scratch track, since tools are available (and for some reason I had this hang up with trashing a track and truly forgetting about it), but ultimately this became a futile waste of time. The advantage of unlimited tracks and other tools makes the art of overdubbing takes a breeze with SONAR... essentially "storyboard" the song, overdub the scratch tracks with better versions, mute/delete the scratch tracks.
 
In your situation the scratch tracks will also allow you to insert all of the tempo/time changes for an "overdub" template. Another good idea is to save the project with a new descriptive name at each major stage of production (to the same project folder).

Many thanks again, lawajava and mettelus ! Your answers have really helped me to enjoy recording my song so far instead of feeling stressed.
 
I performed the song in order, singing and playing my guitar into one mic. I let imprecisions go (very hard for me to do!) and tried just to get a pretty solid take rhythmically and characterwise from start to finish. This method of not stopping the whole show for each tiny fault took lots of pressure away as I usually feel like I'm on razor's edge about to screw everything up and don't give a relaxed, natural performance as a result.
 
Next, I added markers and tempo indications as advised.
 
Once this was done, I played along with the guitar (replaying the same part) on a new track, being a little fussier to get it just right.
 
I just finished singing the (same) lead vocal three times and am now taking a break to write to you before going back to see if any of the singing is a keeper.
 
Thanks again for your generosity :)
LJ

Sounds like you have a workable game plan.
 
I have songs with countless time signature changes, and use markers to help me keep sections straight.  I hope you find them helpful as well.
 
I would suggest getting some sort of midi drums together until you are SURE about tempo of any parts of the song, because the midi playback will self-adjust if tempo changes after you record the midi data, but you start to run into trouble quickly when dealing with audio recorded at one speed - and the speed has to be changed.  Just enough of some sort of midi when tempo changes occur, so you can make sure you like the tempo feel.  THEN you can add audio in and will hopefully reduce any need to do things over from tempo-change problems.
 
Hope that made sense.
 
Bob Bone