Beat Mapping in Logic Pro X

bingefeller

I have noticed that, sometimes, when I am beat mapping a song in Logic Pro X, even if I know for an absolute fact the artist is using a drum machine, the tempo seems to waver and isn't constant as one would expect with a drum machine.  Anyone know why this is?  @Clarky ? 

bingefeller

@Clarky

randomhandclaps

Just because it's a drum machine doesn't mean the drums are programmed square (in fact if the programmer is any good the very rarely are unless it's a deliberate effect) or are not tempo shifted throughout the piece.

bingefeller

randomhandclaps said:

Just because it's a drum machine doesn't mean the drums are programmed square (in fact if the programmer is any good the very rarely are unless it's a deliberate effect) or are not tempo shifted throughout the piece.

So the programmer doesn't programme the drums 100% perfect in time on purpose for a more natural sound?   I don't have any experience with any "real" drum machines, just Ultrabeat, EZDrummer and Logic's built in Drummer.   

Clarky

if the tempo is not 100% spot on I'd guess that it is because:

- the sounds were samples played in via a keyboard / something like V-Drum / some other MIDI instrument

- the song is built out of sampled loops that don't match the song tempo correctly [rap tracks sometimes have this and deliberately so, cos the slight 'wrongness' can sound pretty cool]

- the song's tempo has automation on it. this is not as rare as you'd think.. you can have a small increase in tempo as you head into the chorus to give it a bit of a lift, and then back it off a little as you go back to the verse to drain some energy away.. so you're essentially programming the song to do what a band does naturally

octatonic

A lot of early drum machines had weird clocking and tempo drift was sometimes apparent.

Clarky

haa… or sampling a groove from warped vinyl.. lmao

bingefeller

It's a bit annoying when beat matching to be honest lol.

I've noticed this in songs from Pet Shop Boys, Erasure, early Prince (when he was using the Linn) etc.  

octatonic

The Linn had an analog VCO for its master clock.

It definitely drifts.

It happens a lot with some digital gear as well- I remember doing some recording for Martin McCormick (ex-Therapy?) where he brought along his Electribe drum machine and it was drifting all over the place- very noticeable and unusable.

It tightened up once I had it chasing midi clock from Logic.

bingefeller

octatonic said:

The Linn had an analog VCO for its master clock.

It definitely drifts.

It happens a lot with some digital gear as well- I remember doing some recording for Martin McCormick (ex-Therapy?) where he brought along his Electribe drum machine and it was drifting all over the place- very noticeable and unusable.

It tightened up once I had it chasing midi clock from Logic.

I had automatically assumed that a drum machine would be perfect time :P

octatonic

bingefeller said:

octatonic said:

The Linn had an analog VCO for its master clock.

It definitely drifts.

It happens a lot with some digital gear as well- I remember doing some recording for Martin McCormick (ex-Therapy?) where he brought along his Electribe drum machine and it was drifting all over the place- very noticeable and unusable.

It tightened up once I had it chasing midi clock from Logic.

I had automatically assumed that a drum machine would be perfect time :P

Nope- the best way is to have it chase a sequencer that is known to not drift.

Logic is pretty good but actually the best thing ever for midi timing is the Atari ST, because midi sports are built in and timing is linked to CPU clock timing, where on modern computers they are at a higher level, part of the operating system or via drivers, usually over the USB bus and it isn't as tight.

I miss my old Atari- I wouldn't go back to it because the workflow isn't remotely acceptable compared to what we can do these days but the midi timing was so tight.

bingefeller

octatonic said:

bingefeller said:

octatonic said:

The Linn had an analog VCO for its master clock.

It definitely drifts.

It happens a lot with some digital gear as well- I remember doing some recording for Martin McCormick (ex-Therapy?) where he brought along his Electribe drum machine and it was drifting all over the place- very noticeable and unusable.

It tightened up once I had it chasing midi clock from Logic.

I had automatically assumed that a drum machine would be perfect time :P

Nope- the best way is to have it chase a sequencer that is known to not drift.

Logic is pretty good but actually the best thing ever for midi timing is the Atari ST, because midi sports are built in and timing is linked to CPU clock timing, where on modern computers they are at a higher level, part of the operating system or via drivers, usually over the USB bus and it isn't as tight.

I miss my old Atari- I wouldn't go back to it because the workflow isn't remotely acceptable compared to what we can do these days but the midi timing was so tight.

What Atari did you have?  A pal of mine used to have an ST (I forget the model) but all he did was play games on it

randomhandclaps

bingefeller said:

So the programmer doesn't programme the drums 100% perfect in time on purpose for a more natural sound?

Not so much natural sound as natural 'groove', although it's just semantics depending on what you meant. 

It would help to know the piece.

However a real drummer is never square.  A beat is always pushed or dragged and quite often if you look at the very best drummers only the kick on beat one (and maybe 3) falls completely square on the click.  This means that if you are trying to mimic a real drum part then square is out.

The exceptions that we tend to think of are dance and hip hop (and all the shoot offs).  However these two are ultimately consciously or subconsciously influenced by a 'groove' whether human created or a quirk of technical flaws. 

Hip hop obviously has it's roots in soul and jazz.  It started off very sample driven however early drum machines gave artists the chance to create new beats.  The problem was that some machines like Octa has already mentioned were unreliable tempo wise but just as bad in this arena was machines with ruthlessly accurate quantisation.  With a music that was so beat led (and the beat high in the mix) the squareness of a drum machines stood out like a sore thumb and until DAWs there was no way of shifting beats by small amounts to create grooves.  The solution was the 16th note triplet groove with is over pretty much all the 80 to early 90s stuff.

Dance music embraced the squareness of drum machines initially more so than any other genre.  With most four to the floor house music groove was quite irrelevant because it was about white people jumping around on acid and ecstasy.  This changed when drum and bass arrived by an odd quirk.  Drum and bass pioneers started using real drum hit samples in dance tunes.  To play these back they mainly used the early Akai S-range samplers.  The problem is the moment you started fucking around with these beats to either disguise their origin or create new sounds and then sped it up the samplers became unstable timing-wise on resampling.  This meant not only did drum and bass create that broken beat sound it also (almost by accident) put a groove back into club based dance music.

Now even if you listen to the hardest trance or gabba type tracks the drums are rarely dead square.

A good experiment is to set a tempo of 130bpm (you average club track tempo).  Program a kick drum on every beat and a closed high hat on every other.  Then shift the high hats ahead of the beat by 10-20ms (or try more) and listen to the effect.  The do the same pushing them behind the beat.

octatonic

bingefeller said:

octatonic said:

bingefeller said:

octatonic said:

The Linn had an analog VCO for its master clock.

It definitely drifts.

It happens a lot with some digital gear as well- I remember doing some recording for Martin McCormick (ex-Therapy?) where he brought along his Electribe drum machine and it was drifting all over the place- very noticeable and unusable.

It tightened up once I had it chasing midi clock from Logic.

I had automatically assumed that a drum machine would be perfect time :P

Nope- the best way is to have it chase a sequencer that is known to not drift.

Logic is pretty good but actually the best thing ever for midi timing is the Atari ST, because midi sports are built in and timing is linked to CPU clock timing, where on modern computers they are at a higher level, part of the operating system or via drivers, usually over the USB bus and it isn't as tight.

I miss my old Atari- I wouldn't go back to it because the workflow isn't remotely acceptable compared to what we can do these days but the midi timing was so tight.

What Atari did you have?  A pal of mine used to have an ST (I forget the model) but all he did was play games on it

I had a 1040 ST and later an STE. I also had an Amiga 500 way back when.

bingefeller

randomhandclaps said:

bingefeller said:

So the programmer doesn't programme the drums 100% perfect in time on purpose for a more natural sound?

Not so much natural sound as natural 'groove', although it's just semantics depending on what you meant. 

It would help to know the piece.

However a real drummer is never square.  A beat is always pushed or dragged and quite often if you look at the very best drummers only the kick on beat one (and maybe 3) falls completely square on the click.  This means that if you are trying to mimic a real drum part then square is out.

The exceptions that we tend to think of are dance and hip hop (and all the shoot offs).  However these two are ultimately consciously or subconsciously influenced by a 'groove' whether human created or a quirk of technical flaws. 

Hip hop obviously has it's roots in soul and jazz.  It started off very sample driven however early drum machines gave artists the chance to create new beats.  The problem was that some machines like Octa has already mentioned were unreliable tempo wise but just as bad in this arena was machines with ruthlessly accurate quantisation.  With a music that was so beat led (and the beat high in the mix) the squareness of a drum machines stood out like a sore thumb and until DAWs there was no way of shifting beats by small amounts to create grooves.  The solution was the 16th note triplet groove with is over pretty much all the 80 to early 90s stuff.

Dance music embraced the squareness of drum machines initially more so than any other genre.  With most four to the floor house music groove was quite irrelevant because it was about white people jumping around on acid and ecstasy.  This changed when drum and bass arrived by an odd quirk.  Drum and bass pioneers started using real drum hit samples in dance tunes.  To play these back they mainly used the early Akai S-range samplers.  The problem is the moment you started fucking around with these beats to either disguise their origin or create new sounds and then sped it up the samplers became unstable timing-wise on resampling.  This meant not only did drum and bass create that broken beat sound it also (almost by accident) put a groove back into club based dance music.

Now even if you listen to the hardest trance or gabba type tracks the drums are rarely dead square.

A good experiment is to set a tempo of 130bpm (you average club track tempo).  Program a kick drum on every beat and a closed high hat on every other.  Then shift the high hats ahead of the beat by 10-20ms (or try more) and listen to the effect.  The do the same pushing them behind the beat.

Hmmmm... OK I did what you said.  

Track is 130bpm and first section is everything lined up perfect.  

Second time round I delayed the hihat by 10ms.

Third time it's delayed by 20ms

Fourth time delayed by 50ms.

Fifth time the hihat is 30ms ahead of the beat.  

Interesting.  I'd never considered doing this ever.  I had seen the nudge feature in Logic but never explored it.

bingefeller

octatonic said:

bingefeller said:

octatonic said:

bingefeller said:

octatonic said:

The Linn had an analog VCO for its master clock.

It definitely drifts.

It happens a lot with some digital gear as well- I remember doing some recording for Martin McCormick (ex-Therapy?) where he brought along his Electribe drum machine and it was drifting all over the place- very noticeable and unusable.

It tightened up once I had it chasing midi clock from Logic.

I had automatically assumed that a drum machine would be perfect time :P

Nope- the best way is to have it chase a sequencer that is known to not drift.

Logic is pretty good but actually the best thing ever for midi timing is the Atari ST, because midi sports are built in and timing is linked to CPU clock timing, where on modern computers they are at a higher level, part of the operating system or via drivers, usually over the USB bus and it isn't as tight.

I miss my old Atari- I wouldn't go back to it because the workflow isn't remotely acceptable compared to what we can do these days but the midi timing was so tight.

What Atari did you have?  A pal of mine used to have an ST (I forget the model) but all he did was play games on it

I had a 1040 ST and later an STE. I also had an Amiga 500 way back when.

Nice.  I had an Amiga 1200 but, again, never recorded any music on it. Those Ataris were the some of the best computers for midi and recording music back then.

Clarky

bingefeller said:

randomhandclaps said:

bingefeller said:

So the programmer doesn't programme the drums 100% perfect in time on purpose for a more natural sound?

Not so much natural sound as natural 'groove', although it's just semantics depending on what you meant. 

It would help to know the piece.

However a real drummer is never square.  A beat is always pushed or dragged and quite often if you look at the very best drummers only the kick on beat one (and maybe 3) falls completely square on the click.  This means that if you are trying to mimic a real drum part then square is out.

The exceptions that we tend to think of are dance and hip hop (and all the shoot offs).  However these two are ultimately consciously or subconsciously influenced by a 'groove' whether human created or a quirk of technical flaws. 

Hip hop obviously has it's roots in soul and jazz.  It started off very sample driven however early drum machines gave artists the chance to create new beats.  The problem was that some machines like Octa has already mentioned were unreliable tempo wise but just as bad in this arena was machines with ruthlessly accurate quantisation.  With a music that was so beat led (and the beat high in the mix) the squareness of a drum machines stood out like a sore thumb and until DAWs there was no way of shifting beats by small amounts to create grooves.  The solution was the 16th note triplet groove with is over pretty much all the 80 to early 90s stuff.

Dance music embraced the squareness of drum machines initially more so than any other genre.  With most four to the floor house music groove was quite irrelevant because it was about white people jumping around on acid and ecstasy.  This changed when drum and bass arrived by an odd quirk.  Drum and bass pioneers started using real drum hit samples in dance tunes.  To play these back they mainly used the early Akai S-range samplers.  The problem is the moment you started fucking around with these beats to either disguise their origin or create new sounds and then sped it up the samplers became unstable timing-wise on resampling.  This meant not only did drum and bass create that broken beat sound it also (almost by accident) put a groove back into club based dance music.

Now even if you listen to the hardest trance or gabba type tracks the drums are rarely dead square.

A good experiment is to set a tempo of 130bpm (you average club track tempo).  Program a kick drum on every beat and a closed high hat on every other.  Then shift the high hats ahead of the beat by 10-20ms (or try more) and listen to the effect.  The do the same pushing them behind the beat.

Hmmmm... OK I did what you said.  

Track is 130bpm and first section is everything lined up perfect.  

Second time round I delayed the hihat by 10ms.

Third time it's delayed by 20ms

Fourth time delayed by 50ms.

Fifth time the hihat is 30ms ahead of the beat.  

Interesting.  I'd never considered doing this ever.  I had seen the nudge feature in Logic but never explored it.

if you need it bang on with the click in Logic you can use the flex tool to fix it

painstaking stuff.. but it works pretty good