Sonar Tempo Editing - Care To Discuss?

Some Modest Proposals
 
For me, good tempo editing capabilities are one of the most important functional areas in a DAW.  While for many users this is not the case, hopefully there are a few of you here for whom good tempo editing facilities are also highly important.  If you are one of those folks, please read what follows and contribute your own thoughts.
 
Sonar does have some decent tempo editing tools, but it seems to me they could be vastly better with relatively little work on CW’s part.  If the current capability is like a family sedan, a few modest enhancements would turn that into a ___________ [name high performance or luxury vehicle of your choice].  When I say "relatively little work", I mean that Sonar already offers a variety of methods to add, modify and delete tempo events.  In what follows, I’m going to suggest a few more.  The code in Sonar that is invoked for playback would not need to be touched … thus "relatively little work" assertion.  These suggestions are strictly additional or improved ways to add, modify and delete tempo events.
 
The first two enhancements are positively at the top of my wish list.
 
Tempo View as a First Class Citizen
 
I’d propose that the first order of business should be to stop relegating second class status to the tempo view.  For starters, how about making the Heads Up Display available in the TV just like it is in the PRV?  But more importantly, how about giving us the same kind of editing capabilities we’ve got in controller lanes of the PRV?  In other words, put grab handles in the display and let us adjust tempo events by dragging them.  Unlike the PRV, however, just up/down movement would be appropriate.  Allowing tempo events to drift in the timeline would probably be counterproductive.
 
You could argue that with snap-to-grid enabled, drawing a new event will accomplish the same thing as moving a grab handle.  But, for me anyway, there’s something a little more assured when dragging.  We can do both dragging and drawing for controller events in the PRV.  Both modes of operation definitely have their time and place.  Why not use that same philosophy in the TV?
 
Simple Tempo Recorder
 
This next idea is so obvious and useful that I really can’t understand why it’s not already a standard feature in every DAW that offers a tempo map.  The heading says it all.
 
Let me provide an example usage of one way this capability could work.  Select a unit in snap to grid in the TV (quarter note, eighth note triplet … whatever).  Select a time range in the time bar.  These two selections would define how many tempo events are to be recorded.  If your piece is in four/four, your unit is eighth notes and you’ve selected six measures, you’d be providing eight tempo events per measure or 48 total events to record.  To do that, you’d need to play 49 consecutive MIDI notes, one per beat, for the sequence of beats/tempos to be completely specified.
 
Sonar would capture the durations between note-on events and use that to compute successive tempo values.  The recorder would shut itself off automatically when all events for the selected time range had been acquired.  If the recorder is stopped before the requisite number of notes were all played, the tempo event slots would remain empty for the remainder of the selection.  Tempo events outside the selected range would be unaffected.
 
You could argue that Sonar already has a similar capability in process/fit-to-improv, and to a limited extent that’s a valid point.  But fit-to-improv isn’t a very friendly, efficient way to do this.  First of all, you’re limited to tempo events at the beat boundaries defined by the time signature.  If your piece is 4/4 and you want tempo events on eighth note boundaries, you’d need to play all kinds of games with the MIDI data before that to be able to happen.  Also, you need to remember to lock the MIDI data so that fit-to-improv doesn’t move things around, which is what it needs to do when using the feature for its intended purpose.  Finally, there’s no way to use this for anything but the entire piece.
 
But a simple tempo recorder?  Elegant, powerful, efficient ... what’s not to like?
 
Syncing Views
 
There were a couple of articles in recent issues of Sound On Sound that discussed tempo management in the latest version of Pro Tools.  Although the screen shots were limited, there was a couple from which one could infer that the tempo map and the Pro Tools equivalent of the PRV shared the same space.  In other words, the tempo map was like a lane in the PRV.
 
I initially thought this was a marvelous idea, but upon further consideration, I don’t think there’s enough vertical real estate to make this very practical.  So here’s an alternate idea.  How about providing a view attribute call "Synchronized" for piano roll, tempo and (maybe) track view windows?  It would work like this: an adjustment to horizontal scrolling or horizontal zoom level in a synchronized window would be tracked in all other synchronized windows.  In other words, the viewable time interval in all synchronized windows would always be the same.
 
This might even be useful when multiple PRV windows alone are opened (and the TV isn’t).  Multiple PRVs was a feature I didn’t even know existed until I read the documentation about X1 screen sets.  Speaking of screen sets, it’s probably necessary that Synchronized and Locked would be mutually exclusive window attributes.
 
A Couple More Small Ideas
 
We can mute MIDI tracks without destroying their content.  Why not be able to temporarily disable the tempo map in favor of a constant tempo we could supply in a "mute tempo" dialog?  This would be useful for practice or recording tricky passages.  While we’re at it, maybe also provide a quick and easy way to temporarily specify a "slowdown" percentage, useful again for practice or recording.
 
I recall reading about a feature in another DAW (maybe Logic, but not sure) where one could select a time interval and invoke a dialog with a slider control which could be used to speed up or slow down the tempo in the selected area in a relative manner.  We can do this in Sonar with process/set-time, but it’s not as straight forward as the feature I’m describing here.  It seems reasonable that the tempo view is the place where this sort of capability should exist.  In fact, that same dialog ought to offer a display of the total time of the selected interval and an edit control in which a new value could be entered.  This is the tempo view, after all, so why not allow one-stop-editing for all things tempo?
 
Wrapping Up
 
Thanks if you’ve stuck with me through all this.  I do hope I’ll see some responses here.  I’d love to know if there are other good ideas out there or thoughts on how my feature suggestions could be improved.  Also, if you have experience with another DAW that offers some unique, useful tempo editing capability, please share it.

im terribly sorry no one has chimes in on this article/brilliant idea you have put together.
and its something that is detinetly a ggod topic because i cant agree with you more that tempo is one of the more important pieces of the puzzle.loved your idead about the tempo recorder.
i have always found the natire of a tempo playing without a click track fluctuates heavey on feel and/or dynamics.
this is a great piece youve put together.im sorry  it hasnt been looked at more closely as you have put forth some real effort and valuable information here.

I actually didn’t see this post until you “updated” it.

And first of all.. (since we’re talking about dynamics, timing)..
The headline intrigued me, for sure. But when I opened the thread I just got knocked out by all the text in the first message :) 

Dynamics: Maybe you could have shortened your first message by just giving us the headlines of them (and maybe a little explanation). I think this would have been a better way of getting people’s attention.

Timing: Later in the thread you could have told us more about your ideas and make us go all “woaaah! what a guy!"

Cuz I really like the ideas! I wouldn’t say I need these kind of editing possibilities all the time, but more than once I’ve felt limited by the lack of tempo editing possibilities.

"Tempo View as a First Class Citizen:" I love this idea! Maybe they could implement a “tempo bus”. So you’ll get a separate bus in the track view where you see the tempo the same way you see any envelope. And then, if this bus is selected you can show it as an envelope in the PRV as well.

"Simple Tempo Recorder:" I really like this too. It would be great if it worked with audio the same way too and not just with midi.

"Syncing Views:" I don’t think I get this problem/idea fully. You think that it woldn’t be enough if you could zoom vertically, like you can in the track view by clicking the grey area between track settings and the content and move the mouse up and down?

"A Couple More Small Ideas:" To at least have a “half tempo” button would be great for, as you say, practicing purposes! If there was a “mute tempo” function, what tempo would the project “jump” to when muted? ..Would it calculate the most used tempo and jump to this or is this something you’d set in a “mute tempo dialogue”?

I really like the idea where you get a tempo dialogue when looping an area! Also a great thing for practicing!

I've always wanted full envelope controls for the tempo view and have suggested it several times.
Fred

I was trying to change the Tempo of a project from 78 to 82 the other day. Just a bounced set of instrumental Midi tracks and a preliminary vocal. Changing the project tempo worked good for the instrumental mixdown but the vocals were totally out of sync and would have required extensive Audio Snap manipulation. I really doubt the vocals would have sounded natural after doing so and just changing the tempo made them unnatural.
I have to assume that it is because of the type of events and their durations and nuances that the amount of detail required to match different events is somewhat beyond a simple timing change for expressive events synced to instrument events.

While we are at improving TEMPO editing, let's not leave out saving TEMPO maps and loading them in other projects.  Don't ask me why I like this idea but it would help more than the HALF TEMPO BUTTON, hint.

Jim

The code in Sonar that is invoked for playback would not need to be touched … thus "relatively little work"

 
I'm not a programmer so I don't know, but some of this seems like a lot for not touching any code.
This biggest issue for me (and for many others as I recall) is the ability to globally increase or decrease overall tempo when a tempo map is involved.  Like how "Offset" is used when track gain envelopes are used.
 
 

dmbaer


Well, the response to my earlier post is disappointing to say the least.  I'd hoped there were at least a few folks out there who'd like to see tempo editing capabilities brought up to the standard of other MIDI data editing in Sonar.
 
Musical expressiveness is all about dynamics, timing, phrasing and intonation.  Timing is tempo.  Nobody cares?  I honestly cannot understand this.

!!--it was only 16 hours since you posted the long, detailed post, dmbaer - Gotta give folks a chance to catch up!

As Brundlefly says later in this thread, it's MIDI-oriented Sonar users who are the most involved with complex Tempo mapping.  The problems Crg lists are typical when users want to start changing the tempo map after audio has been recorded - always gets trickier and messier then.

I use the Tempo Map all the time, using the free form drawing tool with grid off so I can draw changes with as much complexity as I want.   I remember not liking the Tempo Map at all when I first started using Cakewalk, but I dug in - now I've long since adapted to the way it is, and it works great for me.  I'll sit in that view for some time, trying out different ritards, fermatas etc, playing back the results over and over until I like them--That works for me well now.

But the concept of having tempo available as an envelope to work with in the TV along with everything else - that'd be very nice.  It doesn't excite me as much as some others here, because I'm picturing it feeling about the same as with the current tempo map window.  I always work with the current track in TV visible above the Tempo view, or PRV, so the synchronization between the two views is always visually displayed for me.

Turning a knob on my control surface to record tempo changes in real time might be nice - I'm picturing needing to try it over and over to control the changes accurately enough--I say that because I find all my real-time automation recording needs quite a bit of editing afterwards.  I like the slow, methodical editing that's possible in not-real-time since it gives me the best results.

I'm happy with the way it works now, but I think it's great you've brought up several ideas that could make it more generally user friendly.  Thanks for the post.

Randy B.

I work the same way as rbowser and do complex tempo maps which is essential for orchestral music.  I've used programs that allowed a controller to record tempo but it was pretty crude.  With drawn in tempo maps I can get excellent rubato effects especially for solo piano and other really exposed instruments. 

It does take practice to get good at tempo maps.  I've done it so many times that most of the time I can get the correct tempo changes first or 2nd time.

Now if I'm recording live guitar, bass and other instruments, then it's essential I get the tempo map worked out before adding the audio tracks.

Jim

haydn12


Now if I'm recording live guitar, bass and other instruments, then it's essential I get the tempo map worked out before adding the audio tracks. 

That has got to be the best way to go. Any alteration of the tempo after audio has been added is going to involve time stretching (or time compressing) of the audio. No matter how good the software does this there are bound to be some distortions introduced.

I'd love to be able to map tempo to the MW in order to "conduct" timing in a realtime pass.

-JT

I can conduct midi time sync with a wii-mote... :) I made a tempo conducting tap tempo in Bome's, then used glove pie for the wii mote and assign down force to trigger the tap tempo.....

Fred Holmes


I've always wanted full envelope controls for the tempo view and have suggested it several times.
Fred

Are you talking about handles to grab with the mouse and drag?  If so, we're in violent agreement.  I can probably train myself to effortlessly use the draw functionality with snap to grid enabled, but dragging seems so much more natural.

brundlefly

Personally, I think trying to "draw" natural tempo changes is a losing proposition. A "Record Tempo" feature would be the most useful. The O.P. mentioned some of the limitations of Fit Improvisation. AudioSnap's Set Project from Clip has none of those limitations, and would work very well on the kind of dedicated "Tempo Recording" track described, where there is one an only one clearly defined transient for every desired tempo event.

It's difficult, there's no doubt about that.  I've was away from MIDI for almost twenty years.  In the early 90s, I was working with a DOS sequencer (on a 286!!!) that had very rudementary tempo change capabilities.  I talked the author of the software to into giving me an exit and I wrote a tempo editor function.  But you had to specify changes one beat at a time (up 5%, down 10%, and so on).  It was a painfully slow process but with perserverance you could come up with a credible result.
 
I'm convinced that the best solution would be to have an easy to use tempo recorder and then be able to fine tune the results in the graphic editor.  The fit-to-improv isn't totally unusable in this capacity, but you have to do the entire piece.  And that's pretty challenging.  To be able to take a piece one section at a time, say in 16 measure chunks would be so much more approachable.

 
 
One of the problems with the Tempo Recording feature as described is that you wouldn't be able to have any skipped beats, because SONAR wouldn't know whether a beat was skipped, or the tempo was halved between two beats. This is currently a limitation of Fit Improvisation, and would be even worse if you wanted to have tempo events on 8ths or higher resolutions because you would have to play an event on every 8th in the the whole piece - not just in the measures where you need them. I suppose you could first record a map at quarters, and then go back and re-record individual measures at higher resolutions. At some point, though, that gets more "fiddly" than AudioSnap. 
  
 

I don't know that there's any way around that.  You probably could do some algorthmic filling in of missing beats provided your tempo never strayed too far from the average.  Maybe also require that the first measure have all beats provided.  But if this feature existed, I'd probably want it to be optional.  On the other hand, you don't normally have a large number of places in a piece where the slow-downs are excessive.  Doing a tempo map that avoids the heavy duty rubatos and then fine tuning those few places you actually want them in the graphic editor would still be a pretty efficient approach.  Certainly, an intellegent "beat detective" feature would be a wonderful enhancement to a tempo recorder.

joetabby


I'd love to be able to map tempo to the MW in order to "conduct" timing in a realtime pass.

-JT

The "MW" would be what?  Sorry, I'm not familar with that abbreviation (athough it'll probably turn out to be something really obvious).
 
The biggest problem with conducting in real time with music playing is that the software cannot know when the next beat is coming.  Given the speed of today's machines though, yeah, it might be doable.  Recording a tempo map while the music is playing would be very cool indeed.  But I'd happily settle for a tempo recorder that sat there silently as you did your conductor thing.