M. Wagener’s and other’s opinions about “varispeed” or “multitempo” tracks

[blueman]

Hi Michael (and everybody).

I’ve seen you posting in the “computers forum” about how to “varispeed” a song in Nuendo in a quantity around 2%, and then I’ve thought I could ask to you in your forum about a dilemma I have on this.

I’m mainly a composer (though I run a project studio with high end weapons), and I’m not sure about if my way to set the time of my songs is right or not. In fact I’ve been discussing with a friend of mine, who has a big experience arranging and producing low budget records, but has very little experience making songs.

What I do:

I take, as a composer, a lot of care about every detail of my songs and work on them without rush till I feel that every part of the song is finished and great.

Then, when ready to record it, I put a mic in the room and play a guitar (or piano) and sing in a free way the song (without click or monitoring). While playing and singing, I try to be sure that I’m giving the “right weight” to the song and the “right intention” with the “right time” according with the emotions that convinced me when composing. I make two or three takes.

Then I cut in my PT those audio tracks into audio regions corresponding to the parts of the song (i.e: Intro, verse 1, bridge 1, chorus 1, verse 2, bridge 2, chorus 2, solo, part C, chorus 3, final) and try to “catch” the tempo of each and every of those parts doing the following:

I place the region at the beginning of a group of bars corresponding to the duration of that part (I mean, if the verse 1 is 16 bars long, I put the verse 1 region at the beginning of a group of 16 bars) and begin to adjust the tempo until the audio region fits perfectly in the 16 bars space. Then, I obtain a tempo, say 118,23.

I do that with every part of the two or three takes. (yes, it’s a lot of work). Then I extract a common or average tempo for each part. (i.e: verse 1-take 1: 118.23; verse 1-take 2: 118.12; verse 1-take 3: 117.65. conclusion verse 1: 118), (and then: chorus 1-take 1: 119.65; chorus 1-take 2: 119.60; chorus 1-take 3: 119.37. conclusion chorus 1: 119.50).

What I usually find is, that if I want to keep the right weight and intention of the song as it was conceived, it has to have slightly different tempos in the different parts (i.e. intro:119, verses: 118; bridges: 119.50; choruses: 119; part C: 120; etc…)

So I set up the session in a “varispeed” way with the different tempos in the different parts (with variations of say 2 - 3%) and then I smooth out every transition from one part to the next. (I mean, i.e: from verse: 118 to bridge 119.50, I use the last two bars of the verse and maybe the first of the bridge to make a “tempo curve” that translates from 118 to 119.50).

That way, I obtain a tempo set up that fits better with the right weight and intention of every part while I try to smooth it for not to be noticeable. Then I can set up a click that drives the musicians into a “multitempo” song without problems.

And the question is:

My friend says that this is stupid, that I should choose just one tempo, preferable the chorus one (cause being the most important) and use it for the whole song, cause a varispeed or multitempo track adds nothing but troubles for me and the musicians and the song has to feel right in a constant tempo.

What do you think? Am I doing it right? Is my friend right and I’m wrong?

In other words: Is such a “sin” to varispeed a song in a 2 – 3% quantity in order to achieve the best weight and intention in every part of the song?


Everybody’s opinion that has carried with this issues is welcome.

Thank you very much.

[joeq]

I'm with you. static tempos can be boring

I do similar micro tempo stuff when I am in a situation where I am overdubbing everything

But it sure is a lot of work to get you back to where a few well rehearsed musicians would get you in 15 minutes playing live

[toolskid]

dude! Unless I've completely misread you're post (which is possible - I'm knackered here)

APPLE+I will be your friend...

Originally Posted by blueman

Hi Michael (and everybody).

I’ve seen you posting in the “computers forum” about how to “varispeed” a song in Nuendo in a quantity around 2%, and then I’ve thought I could ask to you in your forum about a dilemma I have on this.

I’m mainly a composer (though I run a project studio with high end weapons), and I’m not sure about if my way to set the time of my songs is right or not. In fact I’ve been discussing with a friend of mine, who has a big experience arranging and producing low budget records, but has very little experience making songs.

What I do:

I take, as a composer, a lot of care about every detail of my songs and work on them without rush till I feel that every part of the song is finished and great.

Then, when ready to record it, I put a mic in the room and play a guitar (or piano) and sing in a free way the song (without click or monitoring). While playing and singing, I try to be sure that I’m giving the “right weight” to the song and the “right intention” with the “right time” according with the emotions that convinced me when composing. I make two or three takes.

Then I cut in my PT those audio tracks into audio regions corresponding to the parts of the song (i.e: Intro, verse 1, bridge 1, chorus 1, verse 2, bridge 2, chorus 2, solo, part C, chorus 3, final) and try to “catch” the tempo of each and every of those parts doing the following:

I place the region at the beginning of a group of bars corresponding to the duration of that part (I mean, if the verse 1 is 16 bars long, I put the verse 1 region at the beginning of a group of 16 bars) and begin to adjust the tempo until the audio region fits perfectly in the 16 bars space. Then, I obtain a tempo, say 118,23.

I do that with every part of the two or three takes. (yes, it’s a lot of work). Then I extract a common or average tempo for each part. (i.e: verse 1-take 1: 118.23; verse 1-take 2: 118.12; verse 1-take 3: 117.65. conclusion verse 1: 118), (and then: chorus 1-take 1: 119.65; chorus 1-take 2: 119.60; chorus 1-take 3: 119.37. conclusion chorus 1: 119.50).

What I usually find is, that if I want to keep the right weight and intention of the song as it was conceived, it has to have slightly different tempos in the different parts (i.e. intro:119, verses: 118; bridges: 119.50; choruses: 119; part C: 120; etc…)

So I set up the session in a “varispeed” way with the different tempos in the different parts (with variations of say 2 - 3%) and then I smooth out every transition from one part to the next. (I mean, i.e: from verse: 118 to bridge 119.50, I use the last two bars of the verse and maybe the first of the bridge to make a “tempo curve” that translates from 118 to 119.50).

That way, I obtain a tempo set up that fits better with the right weight and intention of every part while I try to smooth it for not to be noticeable. Then I can set up a click that drives the musicians into a “multitempo” song without problems.

And the question is:

My friend says that this is stupid, that I should choose just one tempo, preferable the chorus one (cause being the most important) and use it for the whole song, cause a varispeed or multitempo track adds nothing but troubles for me and the musicians and the song has to feel right in a constant tempo.

What do you think? Am I doing it right? Is my friend right and I’m wrong?

In other words: Is such a “sin” to varispeed a song in a 2 – 3% quantity in order to achieve the best weight and intention in every part of the song?


Everybody’s opinion that has carried with this issues is welcome.

Thank you very much.

[bleen]

Michael's question about getting Nuendo to varispeed has less to do with creating a tempo map for a song (like you are doing), and more with being able to slow down/speed up Nuendo for a pass of a doubled guitar or vocal or piano, etc. like we used to do in the "old days" on a multitrack analog deck. This involves changing the samplerate of the incoming clock rather than creating tempo changes within a song.

And there's nothing wrong with you wanting songs to push/pull/breathe based on the feel of each section, but it certainly sounds like an amazing amount of work to do. And be sure to check out Toolskid's hint - it'll save you hours.thumbsup

[toolskid]

theres faster ways of generating a tempo map with beat detective but apple + I is a good place to start

[bleen]

Originally Posted by toolskid

theres faster ways of generating a tempo map with beat detective but apple + I is a good place to start

True, and either way will be faster than manually trying to figure out the tempo of a region. Talk about an inspiration killer there...

[blueman]

Originally Posted by toolskid

theres faster ways of generating a tempo map with beat detective but apple + I is a good place to start

Beat Detective is not an option cause in a strummed guitar + voice track the beats aren't clear enough

What is apple + I?

[blueman]

Originally Posted by joeq

I'm with you. static tempos can be boring

I do similar micro tempo stuff when I am in a situation where I am overdubbing everything

But it sure is a lot of work to get you back to where a few well rehearsed musicians would get you in 15 minutes playing live

Yes. This is a way to simulate how it would be in a live session with a band of feelingfull musicians

[toolskid]

Originally Posted by blueman

Beat Detective is not an option cause in a strummed guitar + voice track the beats aren't clear enough

What is apple + I?

yes... it is....it works fantastically well for this kind of operation, if you want you can call me and I'll talk you through it. Pm me for my contact or get it from my website

[blueman]

Hi, Emre

I've PMed you

What do you think as a drummer about slightly tempo varied songs? Do you have any problem with it? Do you use to find it often in your jobs?

Thanks, man thumbsup

[octatonic]

I sequence/grid everything first, but then work with a drummer is exceptional.
He likes to work to a click but is very good at drifting around the click.

Then I mute all the sequences and re-record all the guitars/bass etc with click off.
Then manually play the sequenced material into the computer in free time.

It is a lot of work, but the natural drift of the drummer is great for feeding off.

I also give him 'free bars' to do whatever he likes.
Then I'll sometimes write around his parts.
It is an inspiring way to create.

[kats]

My friend says that this is stupid, that I should choose just one tempo, preferable the chorus one (cause being the most important) and use it for the whole song, cause a varispeed or multitempo track adds nothing but troubles for me and the musicians and the song has to feel right in a constant tempo.

There two kinds of dynamics in music - volume and tempo. Putting music to a grid reduces 50% of them. It amazes me how everyone is crying the blues about the loudness wars killing dynamics - yet recorded everything snapped to a grid.

I think the point here is that the use of grids or tempo fluctuation should be used to serve the song... it should NOT be about lack of talent. IE some music will sound better snapped to a grid - depends on the vibe of the song. Other songs (most IMO) sound better and draw the listener in with subtlle dynamic build ups/downs in energy through the use of fluctuating tempo and volume.

I think theres a reason orchestras have conductors and not a big MOFO metronome on stage.

[mwagener]

First off: the varispeed I refer to is not based on different tempos in different parts, it's an overall change in speed (a few %) to give the song a tighter feel. If you have a "gap" between the point where the kick lands and the point where the bass lands on the downbeat, the more you speed it up, the smaller that gap becomes. I am not a friend of "gridding" music and I rather leave a few "mistakes" in there than have everything sounding stale, aligned to a grid. Music has mistakes and unevenness (is that a word?) to make it sound “personal” and I feel we are sometimes chasing after the wrong “perfection”.

The other use of varispeed is during tracking. Let’s say we record a loose, what I call “football choir” (ala Accept’s Balls To The Wall), I would record a bunch of tracks in normal speed, then speed up the song by about 10%, so everybody has to sing a lot higher and faster than normal, and they won’t be on time as much when you slow it down. It creates a bunch of low sounding voices (sounds kinda drunk), organized sloppyness. Do the same by slowing it down 10% during tracking and you add “the girls” to the mix. Sometimes I double an acoustic guitar with a very slightly slowed down or sped up machine to make it sound like a 12 string or give it a chorusy effect. Backing vocals sound great when they are tracked and stacked with slightly different speeds.

Now as to your question: I feel you are making yourself more work than is necessary. You will have a certain feel about the tempo at the time you go into the room and play the song. That feel could change quite a bit if you would play the song at a different time of the day (after dinner, early in the morning etc.), so which one is the right one? I try to get an overall tempo feel for a song during pre-production by having the band play the song at different times in the day without a click and checking the tempo they are playing at. You’d be surprised how much the tempo changes based on how everybody feels at that time, sometimes by more than 6 BPM overall. And it seems like everybody feels the same way at that time. Usually if they play it late at night everybody feels the song should be slower, if they just got out of the shower in the morning, it can’t be fast enough. Sometimes you find out during mixing, when all the instrument fall into place, that the tempo could be different (hopefully faster), hence my wish to be able to varispeed the whole enchilada. Also, If there are ANY doubts about the tempo when tracking time comes around, I would suggest to track a little slower rather than a little faster, it actually sounds good (to me) when you tighten up the song a bit by speeding it up slightly (0.5% – 2% max), but when you slow it down after everything is recorded the “gaps” become larger and it starts sounding sloppy.

The tempo changes you are talking about, from 119.0 to 119.5 i.e. are very small and could be based on a bunch of different factors, like your mood in that particular moment, which is not to say it's a bad thing. A good drummer will do those changes based on the feel of the song anyway. Even if he/she is playing to a click, they wander slightly ahead of the click in the chorus and lay back in the verses, they play in the front of the groove or on the backside of the groove (at least they should anyway). Sometimes it’s just the snare being at the backside of the groove and the kick and hat are in front (listen to “Back In Black or any other AC/DC record).

I would think it is very hard for another person to follow your recorded click changes, if they feel even slightly different about the part. Also, they would have to be listening to and concentrating on the click as a guide much more exactly as when they just play by their own feel, which might take their concentration away from the feel of the song. A click to me is a guide to have the same tempo in the end of the song as in the beginning. That said, I have recorded songs without a click and had the band speed up towards the end on purpose, just to have it sound like more energy building up during the song. If the drummer can play without a click, switch the click off and let him/her go at it, it’s all music.

Now all that said, did you know that radio stations sometimes speed up the music by quite a large amount, so they can fit more music, and more important: more commercials, in the same timeframe. Try checking the tuning of a song you know from a CD when it’s played on the radio. So why are we wrecking our brains about what the exact tempo of a song should be??? I guess because we love what we are doing, right?

[mwagener]

Originally Posted by kats

There two kinds of dynamics in music - volume and tempo. Putting music to a grid reduces 50% of them. It amazes me how everyone is crying the blues about the loudness wars killing dynamics - yet recorded everything snapped to a grid.

I think the point here is that the use of grids or tempo fluctuation should be used to serve the song... it should NOT be about lack of talent. IE some music will sound better snapped to a grid - depends on the vibe of the song. Other songs (most IMO) sound better and draw the listener in with subtlle dynamic build ups/downs in energy through the use of fluctuating tempo and volume.

I think theres a reason orchestras have conductors and not a big MOFO metronome on stage.

thumbsup thumbsup thumbsup

[blueman]

Thanks a lot to everybody.

Michael; I’m sorry about my confusion on your “varispeed in nuendo” thread.

Ok, really, I’m not so sorry cause, that way, I’ve learned one more thing from you.

I can’t stop learning from you, guys, in this forum.

Ok, about the various tempos thing:

Michael:
I, of course, don’t play the same tempo every time I perform a song.
Last time I’ve done this process it has oscillated a couple of points around, but that is not the thing that worries to me. What worries to me is the difference that I always find between two parts of a song. I.e: If I perform, say 2 points faster the chorus than the verse, I’ll always do it around a couple of points, independently on the “general speed of that day”. I mean, if I play the verse around 115, I’ll play the chorus around 117, and if I begin to play the verse in 120, the chorus will be around 122, and never at the same tempo or inferior.

You can call this (and maybe it is) a composition fault, but the thing is:

Setting just one tempo for the whole song: if I choose the one where the verses feel good, the choruses will feel too slow in comparison with the previous and next parts, and if I choose the one where the choruses feel good, the verses will feel too fast in comparison.

The question is:
If you would have this situation, what would you do?

To make an smooth tempo map where every part feels good

or….

Choose just one tempo and renounce to the “best weight/intention on every part” and try to get it the best you can with just one tempo and the musicians.


Thank you very much, Michael and everybody.

[Kiwiburger]

IMO - if you want the chorus to be faster than the verse, simply insert some tempo changes into the Cubase Tempo track. That's what it's there for - by popular demand from the numerous professionals who have influence over Steinberg.

The big trick, imo, is to make the tempo change slightly before they are required. Maybe halfway through the preceding bar, or on the last back beat.

I prefer to use programmed high hats instead of a click/clack. For a tempo change section, you might want to use 16th notes instead of 8ths. It's only a guide, but it will give everyone a better feel.

Let the drummer get slippery with the timing - especially tom fills. They will never feel right if they are perfectly quantised.

[not_so_new]

Originally Posted by kats

There two kinds of dynamics in music - volume and tempo. Putting music to a grid reduces 50% of them. It amazes me how everyone is crying the blues about the loudness wars killing dynamics - yet recorded everything snapped to a grid.

I think the point here is that the use of grids or tempo fluctuation should be used to serve the song... it should NOT be about lack of talent. IE some music will sound better snapped to a grid - depends on the vibe of the song. Other songs (most IMO) sound better and draw the listener in with subtlle dynamic build ups/downs in energy through the use of fluctuating tempo and volume.

I think theres a reason orchestras have conductors and not a big MOFO metronome on stage.

I agree 100%. The thing that I find funny is most people will agree with that yet as soon as they are presented with a band that is not on a grid the same people will say that same band is sloppy.

fuuck

There are folks around here even that will chime in and fall all over themselves to agree yet they sill say that John Bohnman has no groove (yes it has been said around here).

People are funny.....

[mwagener]

Originally Posted by blueman

Thanks a lot to everybody.

Michael; I’m sorry about my confusion on your “varispeed in nuendo” thread.

Ok, really, I’m not so sorry cause, that way, I’ve learned one more thing from you.

I can’t stop learning from you, guys, in this forum.

Ok, about the various tempos thing:

Michael:
I, of course, don’t play the same tempo every time I perform a song.
Last time I’ve done this process it has oscillated a couple of points around, but that is not the thing that worries to me. What worries to me is the difference that I always find between two parts of a song. I.e: If I perform, say 2 points faster the chorus than the verse, I’ll always do it around a couple of points, independently on the “general speed of that day”. I mean, if I play the verse around 115, I’ll play the chorus around 117, and if I begin to play the verse in 120, the chorus will be around 122, and never at the same tempo or inferior.

You can call this (and maybe it is) a composition fault, but the thing is:

Setting just one tempo for the whole song: if I choose the one where the verses feel good, the choruses will feel too slow in comparison with the previous and next parts, and if I choose the one where the choruses feel good, the verses will feel too fast in comparison.

The question is:
If you would have this situation, what would you do?

To make an smooth tempo map where every part feels good

or….

Choose just one tempo and renounce to the “best weight/intention on every part” and try to get it the best you can with just one tempo and the musicians.


Thank you very much, Michael and everybody.

A tempo change of about 2 BPM is proabaly too much for a drummer to make up for if he is playing to a constant click. If your music and the way YOU feel it requires a tempo change of 2 BPM in the chorus then you MUST program a tempo map to acommodate your feel in the music, a straight click won't give you what you need. An alternative would be to play without a click and make the band follow you as a leader. There are no rules, do whatever works to make YOUR music come across the way YOU would like to hear it.

[not_so_new]

Michael thanks for the post! Not to get off subject but... coming from an analog background were I used to do varispeed quite often.. well I miss it.

I have never been comfortable doing this with digital gear. I know that you are a digital guy so what is your take on this? How much will you push or pull a song? What percentage do you do this with before you start to notice?

I know I will have to play with it but I have refrained from doing it for so long now (gotta be 7 or 8 years) that this thread has given me some inspiration to try it again. Pointers to give me a kick in the pants?

Anyone else? Apogee AD/DA-16X with AES16 card here, does anyone do varispeed with hardware using this setup?

Sorry to hijack if that is what I am doing...... go

[mwagener]

Originally Posted by not_so_new

Michael thanks for the post! Not to get off subject but... coming from an analog background were I used to do varispeed quite often.. well I miss it.

I have never been comfortable doing this with digital gear. I know that you are a digital guy so what is your take on this? How much will you push or pull a song? What percentage do you do this with before you start to notice?

I know I will have to play with it but I have refrained from doing it for so long now (gotta be 7 or 8 years) that this thread has given me some inspiration to try it again. Pointers to give me a kick in the pants?

Anyone else? Apogee AD/DA-16X with AES16 card here, does anyone do varispeed with hardware using this setup?

Sorry to hijack if that is what I am doing...... go

Well, so far I didn't have any problems, because the R-1 does a +/- 12.5% varispeed, no problem (have to figure out how it does it) but now with Nuendo (or any other DAW) involved it becomes a problem. In the rare cases where I use varispeed it's normally not more than +2%. Maybe if I could manage to sync Nuendo to SMPTE, right now it's the other way around and the R-1 is following Nuendo via SMPTE, but it would involve another $2,500 syncho box.

[Kiwiburger]

I can do varispeed stuff in Cubase SX, using the offline pitch shifter. This is true varispeed, just like with analog tape - it's a little fiddly, but doesn't require an external variable clock:

Create your guide track - let's say it's in the key of C at 120 bpm.
Bounce a copy to a single audio track.
Mute the original tracks, so you are just hearing the bounced audio version.
Audio process this track with the pitch shifter - e.g. transpose up one tone to D.
Important: turn OFF the time and formant preserving features.
The whole track is now higher and faster - just like tape.
Record your overdubs.
Discard the bounced track and unmute the original.
What you have now is an overdub that is too high and too fast.
Pitch shift the overdub down one tone - still with time and formant preserving OFF.

You can also do this by changing the sample rate - but I find it easier to use the pitch shift. It achieves exactly the same result - a sample rate conversion.

I've pitch shifted up or down several octaves, then reversed it, and compared against the originals and they null very well. Very little damage is done, because the time stretching and formant shifting is disabled.

[Bob Olhsson]

Music is all about how people respond physically to changes in time and pitch.

Great musicians are always conscious of exactly where perfect time and pitch lie but they never play perfect time and pitch because it's really all about the creation of tension against perfect time and pitch.

[blueman]

Originally Posted by mwagener

A tempo change of about 2 BPM is proabaly too much for a drummer to make up for if he is playing to a constant click. If your music and the way YOU feel it requires a tempo change of 2 BPM in the chorus then you MUST program a tempo map to acommodate your feel in the music, a straight click won't give you what you need. An alternative would be to play without a click and make the band follow you as a leader. There are no rules, do whatever works to make YOUR music come across the way YOU would like to hear it.

I never make a straight tempo change of 2 BPM. I always try to do the smoothest progression I can (usually in quarter of BPMs) in a tempo map.

So ok. I can see that your thinking, Michael, is: “It’s not a sin, but it would be preferable to avoid it”. Isn’t it?

Some of you agree with tempo mapping and some others don’t. Of course the best would be a band playing live, but I can’t afford it.
And, of course, all this is music dependant.

Ok I’ll keep thinking about it.

Thanks to everybody.

[Reptil]

Originally Posted by kats

I think theres a reason orchestras have conductors and not a big MOFO metronome on stage.

tik tak tik tak

LOL sorry couldn't help myself seeing a 3 mtr big metronome in the concert hall

excellent posts anyhow
A guy like me learns something every day

[ripper]

I agree with Kats that playing to a click SHOULD be about the music and not about poor musicianship. But often it IS in fact about poor musicianship.

I certainly don't think "perfect" time (btw, Bob O.'s comments about "perfect" time are spot on! But again, I'm not talking about great players in this post, unfortunately) is more important than the natural vibe of a songs progress, slight tempo changes and all.

But MANY times when I'm recording a young band and there's not much time at all in the budget, a click track will in fact make the drummer play more consistently and often times simpler as well (which means more consistency at the skill level these guys are at). At that point, the constrictive use of a click in the drummers cans actually frees up the band to sound more assured in their parts as they're no longer trying to ride a bucking, runaway horse, if you'll pardon the cowboy mentality (we just got the box set of the Sergio Leone spaghetti westerns on DVD at the studio... now THERE's some essential gear!).

Anyways it certainly helps things in a hurry. I've had many of these less-than-brilliant drummers say "wow, i didn't know i could play that consistently" after a few gos with the click. But even with this situation, there will be some of the songs that work best off.

personally, i'm not the biggest fan of clicks in most cases if i'm playing with a good drummer. btw/ i recorded an album for bernard purdie and man, his tempo varied like crazy! but i've never heard anyone complain of this about his drumming because i'm sure that if you're playing with Bernard, you just play to the MAN and all will be right.

[blueman]

I think this has turned a bit into a "click vs. no click" thread, but that was not what I was asking for.

The original question was: Assuming we're going to use a click:

A constant tempo one (although it reduces some emotional dynamics)

vs.

A multitempo map (smoothed) for fitting in the emotional development of the song.

[joeq]

Originally Posted by blueman

I think this has turned a bit into a "click vs. no click" thread, but that was not what I was asking for.

The original question was: Assuming we're going to use a click:

A constant tempo one (although it reduces some emotional dynamics)

vs.

A multitempo map (smoothed) for fitting in the emotional development of the song.

Of course it is song-dependent.

In general I find longer songs need this more. Shorter and 'groove' songs need it less. Many songs are better with an unvarying tempo.

My method is to make a straight click and a midi scratch track and play along on drums. I adjust tempo where I feel the click is inhibiting where I want to go. Keeping stuff midi at this stage saves a lot of time. I rarely move more than a fraction of a BPM at any one shift. I almost always get faster as the song progresses.


Even with very small changes, it is a good idea to point out the location of the tempo shifts to the overdubbing musicians.

[Killahurts]

Originally Posted by blueman

Assuming we're going to use a click:

A constant tempo one (although it reduces some emotional dynamics)

vs.

A multitempo map (smoothed) for fitting in the emotional development of the song.

Hey Blueman I've got another option. As a composer myself, I have also found that some songs do not benefit from a constant tempo. But most of my attempts at creating tempo maps resulted in just another form of unnaturalness, and again, loss of the emotive dynamic.

Here's my method, I use Digital Performer:

I play and record the naked song, usually on a guitar or keyboard with no click (you could, of coarse, record a scratch of the entire band). When the recording reflects the envisioned tempo dynamic, I go to a "record beats" function in DP, and while listening to the take, I literally play the click track to the entire song, using the spacebar or a midi keyboard. I don't have to be perfect with it, because I can always go in after the fact and edit the placement of the beats, just like you would edit midi.

I get the natural feel of a free performance, but also a (somewhat loose) grid so I can do my composer thingies, like midi. And I get a click track that is a slave to the song, and not the other way around.

I have successfully used this method to record a singer/songwriter, and then Overdub the entire band/orchestration around that performance.

I've got to think most of the DAW's out there have this function, DP's had it for many years.

And concerning vari-speed, man do I miss it. As I was reading what Michael wrote about all the things you could do - the stacking tricks, the tightening, the color variations, etc. - I found myself wishing I had a nice slutty hard disk recorder that has that feature.

[blueman]

Hi Killahurts.

Thanks for your comments.

The way you do it is very similar to the way I do it. In fact, I think my way is a bit more accurate (in the way to obtain the tempo that I’ve played “free” or “live”). I begin, like you, recording a “natural” take with a guitar and voice. I don’t have a band to do it, so I have to overdub everything, like you. If I would have a band, I wouldn’t have this problem, cause we would play live all together.

I realize that my fist post in this thread is maybe too heavy to read , but if you dare , you will find that what I’m saying is very similar to your way. One of the things overlooked along the thread is that I don’t have a band to do the real live thing (if I’m lucky, I’ll pay for a drummer and, if I still have money, for a bassist), so I have to do everything overdubbing while trying to achieve the most “natural” “servicing to the song” tempo dynamics.

[darkwater]

Originally Posted by kats

There two kinds of dynamics in music - volume and tempo. Putting music to a grid reduces 50% of them. It amazes me how everyone is crying the blues about the loudness wars killing dynamics - yet recorded everything snapped to a grid.

I think the point here is that the use of grids or tempo fluctuation should be used to serve the song... it should NOT be about lack of talent. IE some music will sound better snapped to a grid - depends on the vibe of the song. Other songs (most IMO) sound better and draw the listener in with subtlle dynamic build ups/downs in energy through the use of fluctuating tempo and volume.

I think theres a reason orchestras have conductors and not a big MOFO metronome on stage.

This might be the most helpful post I've ever read. GOD it's hard to remember this sometimes... Grids make things work so much easier. But you are so right. The real mojo comes from a performance of a human clockwork, not a swiss movement.

props and all that.